2010-11-21T00:01:06  <burny> I said intelligably.. not flawlessly :P
2010-11-21T00:02:02  <burny> like.. around that point at which, a good human player, would not always beat the bot, even with it's awesome reaction time
2010-11-21T00:02:05  <antimatroid1> there are bots from people at around 1000 that haven't been touched since the beginning of the contest
2010-11-21T00:02:21  <antimatroid1> so i'd say once you're below 1000 you will start seeing somewhat intelligent bots
2010-11-21T00:02:37  <burny> ffeyou mean rank 1000?
2010-11-21T00:02:48  <antimatroid1> yeah, but they're still not very good there
2010-11-21T00:03:00  <antimatroid1> but you'll start getting some bots people have spent a bit of time on i think
2010-11-21T00:03:39  <burny> the one I have right now, is my under 5minutes bot
2010-11-21T00:18:29  <contestbot> Top 10 players: bocsimacko(3442), Raschi(3184), _iouri_(3167), bix0r4ever(3156), _Astek_(3143), dmj111(3094), wagstaff(3078), davidjliu(3068), Manwe56(3068), protocolocon(3064)
2010-11-21T00:22:14  *** delt0r <delt0r!~delt0r@188-22-168-225.adsl.highway.telekom.at> has joined #aichallenge
2010-11-21T00:22:40  *** delt0r___ <delt0r___!~delt0r@188-22-30-222.adsl.highway.telekom.at> has quit IRC (Ping timeout: 265 seconds)
2010-11-21T00:47:19  *** yasith_ <yasith_!~yasith@112.134.122.23> has quit IRC (Ping timeout: 265 seconds)
2010-11-21T00:56:06  *** davidjliu <davidjliu!473a6d0e@gateway/web/freenode/ip.71.58.109.14> has quit IRC (Quit: Page closed)
2010-11-21T00:59:43  *** yasith_ <yasith_!~yasith@112.134.127.210> has joined #aichallenge
2010-11-21T01:08:13  *** amstan <amstan!~AndChat@76-10-183-61.dsl.teksavvy.com> has joined #aichallenge
2010-11-21T01:17:52  *** amstan_ <amstan_!~alex@76-10-183-61.dsl.teksavvy.com> has joined #aichallenge
2010-11-21T01:17:52  *** ChanServ sets mode: +o amstan_
2010-11-21T01:18:27  <contestbot> Top 10 players: 
2010-11-21T01:19:56  *** amstan <amstan!~AndChat@76-10-183-61.dsl.teksavvy.com> has quit IRC (Quit: Bye)
2010-11-21T01:20:16  *** amstan_ is now known as amstan
2010-11-21T01:28:39  *** McLeopold <McLeopold!~scott@c-98-247-248-39.hsd1.wa.comcast.net> has joined #aichallenge
2010-11-21T01:38:28  *** amriedle <amriedle!~amriedle@CPE00d009e408a1-CM001225401352.cpe.net.cable.rogers.com> has quit IRC (Quit: leaving)
2010-11-21T01:41:52  *** McLeopold <McLeopold!~scott@c-98-247-248-39.hsd1.wa.comcast.net> has quit IRC (Ping timeout: 240 seconds)
2010-11-21T01:43:32  *** McLeopold <McLeopold!~scott@c-98-247-248-39.hsd1.wa.comcast.net> has joined #aichallenge
2010-11-21T01:44:25  <janzert> wow,  14 hours and 33 games without a loss for bocsimacko
2010-11-21T01:47:15  <Zannick> er, that was a helpful ranking contestbot
2010-11-21T01:47:21  *** Rubicon-|-Cross <Rubicon-|-Cross!~RubiconCr@adsl-99-30-224-80.dsl.pltn13.sbcglobal.net> has joined #aichallenge
2010-11-21T01:48:02  *** CryBaby <CryBaby!~chatzilla@54.sub-75-220-192.myvzw.com> has joined #aichallenge
2010-11-21T01:52:53  *** Rubicon-|-Cross <Rubicon-|-Cross!~RubiconCr@adsl-99-30-224-80.dsl.pltn13.sbcglobal.net> has quit IRC (Quit: Leaving)
2010-11-21T01:53:41  <foucist> can the whoel game be classified a knapsack problem?
2010-11-21T01:57:29  *** amstan <amstan!~alex@76-10-183-61.dsl.teksavvy.com> has quit IRC (Read error: Connection reset by peer)
2010-11-21T01:58:19  *** amstan <amstan!~alex@76-10-183-61.dsl.teksavvy.com> has joined #aichallenge
2010-11-21T01:58:19  *** ChanServ sets mode: +o amstan
2010-11-21T02:05:26  <burny> I've got 20 in a row :P
2010-11-21T02:05:40  <burny> and I wrote the thing in about 5mins
2010-11-21T02:12:33  *** McLeopold <McLeopold!~scott@c-98-247-248-39.hsd1.wa.comcast.net> has quit IRC (Ping timeout: 240 seconds)
2010-11-21T02:18:28  *** CryBaby <CryBaby!~chatzilla@54.sub-75-220-192.myvzw.com> has left #aichallenge
2010-11-21T02:18:31  <contestbot> Top 10 players: bocsimacko(3440), Raschi(3162), bix0r4ever(3161), _iouri_(3154), dmj111(3113), _Astek_(3110), Manwe56(3095), oldman(3091), protocolocon(3080), wagstaff(3072)
2010-11-21T02:19:32  <burny> first palce seems to be a lock, unless someone makes some changes to their bots
2010-11-21T02:20:07  <burny> woot, cracked the 2k ELO mark
2010-11-21T02:21:40  <burny> it's around the 2200-2500 ELO mark, that games start being intelligable?
2010-11-21T02:29:52  <amstan> woot, dmj111 is getting strong
2010-11-21T02:32:49  <burny> that yours?
2010-11-21T02:43:55  *** smloh1 <smloh1!4786e1ca@gateway/web/freenode/ip.71.134.225.202> has joined #aichallenge
2010-11-21T02:46:33  <smloh1> finally the streak is broken
2010-11-21T02:47:22  <burny> which? bocs?
2010-11-21T02:47:47  <smloh1> yeah
2010-11-21T02:48:39  <burny> ya, 30something is pretty nice
2010-11-21T02:50:51  <burny> how can you find out what your +/- numbers are, and the likes?
2010-11-21T02:51:07  <janzert> basically you can't
2010-11-21T02:51:29  <burny> bah.. just had a game with someoen 600ranks below me
2010-11-21T02:52:39  <burny> guess that helps the streak though lol
2010-11-21T02:52:50  <janzert> :)
2010-11-21T02:53:04  <burny> not that this streak will hold up much longer anyway
2010-11-21T02:54:17  <burny> and broke the 1k rank mark
2010-11-21T02:54:43  <janzert> nice
2010-11-21T02:55:15  <burny> starting to look like I'm gona have to actually enter this thing proper some time this week
2010-11-21T02:55:48  <burny> to get from 2200elo to 3200elo, means basically making it be 1000 times better?
2010-11-21T02:56:47  <burny> aww.. and the streak is broken, by rank 691
2010-11-21T02:57:17  <janzert> 1000 elo means it should beat the lower ranked one about 99.7% of the time
2010-11-21T02:57:17  <smloh1> climbing the ranks sure is fun, isn't it
2010-11-21T02:58:20  <smloh1> i wonder, if we run mega1's bot exhaustively against everyone else on every map...
2010-11-21T02:58:57  <burny> so looks like my minimalist bot is around 2100-2200 elo
2010-11-21T02:59:06  <burny> I was expecting more like 1400-1600
2010-11-21T03:02:23  <janzert> hmm, I think you're underestimating where it will end up
2010-11-21T03:02:43  <janzert> your 1400-1600 my not be as far off as you think
2010-11-21T03:03:17  *** tapwater <tapwater!~tapwater@pool-72-64-104-180.dllstx.fios.verizon.net> has quit IRC (Quit: tapwater)
2010-11-21T03:03:20  <janzert> oh oops
2010-11-21T03:03:29  <janzert> I thought you were say ranking not elo
2010-11-21T03:03:51  <burny> oh lol
2010-11-21T03:03:58  <burny> na, it's already 900ish rank
2010-11-21T03:04:02  <janzert> 1400-1600 ranking is still going to be low
2010-11-21T03:04:05  <janzert> yeah
2010-11-21T03:08:32  <burny> hm, my loss(I can't see the viewer.. jus tlooking at the replay text file).. was it cause I attaked middle too soon? or was there other bad stuff early on?
2010-11-21T03:09:51  <lavalamp_> burny, link?
2010-11-21T03:09:58  <burny> http://ai-contest.com/visualizer.php?game_id=8172977
2010-11-21T03:10:36  <janzert> you expanded to slow
2010-11-21T03:10:41  <lavalamp_> I think you lost because of your poor planet choices
2010-11-21T03:11:36  <burny> on first turn you mean lava?
2010-11-21T03:11:42  <burny> or a bit later?
2010-11-21T03:11:52  <lavalamp_> yeah, real early on
2010-11-21T03:11:59  <lavalamp_> first few turns
2010-11-21T03:12:30  <lavalamp_> if you look at the graph the other guy was always leading in production
2010-11-21T03:12:39  <burny> graph?
2010-11-21T03:13:11  <burny> hm, maybe I'm looking at the replay text wrong.. I see me having 3planets before they have 2
2010-11-21T03:13:50  <lavalamp_> # of planets doesn't matter, it's how much they produce
2010-11-21T03:14:10  <burny> ya
2010-11-21T03:14:40  <lavalamp_> ...and yeah you must be looking at the wrong side ^^
2010-11-21T03:15:37  <burny> ah, so turn 1 they go for 2planets, and I go for 1?
2010-11-21T03:16:10  <lavalamp_> well, that and their pick was just a better value
2010-11-21T03:16:13  <janzert> burny: not near as nice as the default visualizer but does the java viz not work either?
2010-11-21T03:16:16  <janzert> http://ai-contest.com/visualizer_java.php?game_id=8173511
2010-11-21T03:16:27  <burny> I don't have java
2010-11-21T03:16:54  <janzert> and you can't install something other than IE for the browser either?
2010-11-21T03:18:29  <contestbot> Top 10 players: bocsimacko(3411), Raschi(3159), bix0r4ever(3154), _iouri_(3140), dmj111(3109), Manwe56(3099), _Astek_(3084), protocolocon(3080), wagstaff(3067), felixcoto(3067)
2010-11-21T03:20:56  <burny> k, fixing hte thing now that made me pick wrong first planet on that map
2010-11-21T03:21:30  <janzert> looking at http://ai-contest.com/visualizer.php?game_id=8173511
2010-11-21T03:21:58  <janzert> you seem to prone to attack opponent instead of expanding to neutrals, especially when the opponent is still far away
2010-11-21T03:22:48  <burny> do the attacks on opponent do any good? or are they defended against, 100% neutralizing them?
2010-11-21T03:23:08  <janzert> in this case they were defended against
2010-11-21T03:23:13  <burny> k
2010-11-21T03:24:12  <burny> but ya, in all cases, distance was being under valued
2010-11-21T03:30:17  <burny> heh, ya.. the one I liked is horible.. take the second best planet on turn 1.. and then do nothing for like 20turns
2010-11-21T03:38:11  *** McLeopold <McLeopold!~scott@c-98-247-248-39.hsd1.wa.comcast.net> has joined #aichallenge
2010-11-21T03:49:49  <burny> gona have to teach my dumb bot how to defend, how to snipe, and how to escort troops to the front lines
2010-11-21T03:50:01  <burny> then it should be top 100 proly
2010-11-21T03:54:52  <delt0r> burny: you wanted a alternative to twister for rnd number generation?
2010-11-21T03:56:34  <delt0r> :wq
2010-11-21T03:56:38  <delt0r> doh
2010-11-21T03:56:58  <burny> ya,
2010-11-21T03:57:10  <burny> or rather, wanted to know of some
2010-11-21T03:57:19  <burny> twisters the only one I've ever seen mentione
2010-11-21T03:57:24  <delt0r> so there are a few--I don't know why twister is so popular as its not very fast
2010-11-21T03:57:37  <burny> it's reasonably fast
2010-11-21T03:57:42  <delt0r> let me find the refer
2010-11-21T03:57:52  <burny> I made my own before..
2010-11-21T03:58:02  <delt0r> well my one is much faster than a lcg
2010-11-21T03:58:04  <janzert> delt0r: probably mostly because it's been very well studied and found to have good characteristics
2010-11-21T03:58:19  <delt0r> and it passes all the die hard and die harder tests
2010-11-21T03:58:23  <burny> piles faster than twister, but not super good repeat length.. still pretty good randomness though
2010-11-21T03:58:26  <delt0r> where twister fails one.
2010-11-21T03:58:49  <burny> which test is that?
2010-11-21T03:58:51  <delt0r> has a sequnece lenght of 2^128
2010-11-21T03:59:03  <burny> ah, mine's around 2^100
2010-11-21T03:59:21  <delt0r> the die hard and die harder tests are "test for randomness"
2010-11-21T03:59:29  <delt0r> lcg fails epic like
2010-11-21T03:59:29  <burny> oh, ok..
2010-11-21T03:59:35  <burny> though you meant a specific single test
2010-11-21T03:59:42  <burny> lcg?
2010-11-21T04:00:06  <delt0r> linear congurinal generator
2010-11-21T04:00:11  <burny> rotfl..
2010-11-21T04:00:13  <burny> that's not an rng :P
2010-11-21T04:00:16  <delt0r> my spelling is shocking sorry
2010-11-21T04:00:18  <janzert> linear congruential generator
2010-11-21T04:00:22  <janzert> :)
2010-11-21T04:00:25  <delt0r> thx
2010-11-21T04:00:34  <delt0r> well its used a *lot*
2010-11-21T04:00:41  <delt0r> as a rng
2010-11-21T04:00:43  <burny> ya..
2010-11-21T04:00:59  <burny> by people that don't want random numbers.. just want numbers that are named 'rand'
2010-11-21T04:01:02  <delt0r> also both that and twister has the hyperplanes problem
2010-11-21T04:01:11  <burny> hyperplanes?
2010-11-21T04:01:27  <delt0r> twister just has an insane number of dimesions/hyperplanes
2010-11-21T04:01:59  <burny> isn't the only rbg that's worse then lcg like..
2010-11-21T04:02:10  <burny> int rand() {   return 0; }    //?
2010-11-21T04:02:23  <burny> or...
2010-11-21T04:02:37  <burny> int rand() { static int value=0  return value++; }    //?
2010-11-21T04:03:19  <burny> what is the hyperplane problem though?
2010-11-21T04:03:27  <delt0r> not quite
2010-11-21T04:03:42  <delt0r> the old lcg in the IBM C was really bad
2010-11-21T04:04:00  <delt0r> but the current ones are not so bad. They pass many tests of randomness
2010-11-21T04:04:24  <delt0r> but the hyperplanes problem is best show in a wiki animation
2010-11-21T04:05:22  <delt0r> http://en.wikipedia.org/wiki/Linear_congruential_generator
2010-11-21T04:05:50  <delt0r> http://en.wikipedia.org/wiki/File:Lcg_3d.gif
2010-11-21T04:05:56  <burny> k..
2010-11-21T04:05:57  <delt0r> So what do i use...
2010-11-21T04:06:06  <delt0r> I will past bin it --brb
2010-11-21T04:06:09  <burny> so just that it's extremely patterened, in some numbe rof dimmensions?
2010-11-21T04:07:07  <burny> wouldn't lcg have the hyperplane problem in 2d also?
2010-11-21T04:08:44  <delt0r> http://pastebin.com/3y2cSgUc
2010-11-21T04:09:02  <delt0r> burny: no--its has some min number of dimesions
2010-11-21T04:09:06  *** Mistmanov <Mistmanov!4da38f1c@gateway/web/freenode/ip.77.163.143.28> has joined #aichallenge
2010-11-21T04:10:04  <delt0r> burny: the paste bin is java
2010-11-21T04:10:23  <delt0r> shouldn't be too hard to follow however
2010-11-21T04:10:49  <delt0r> step is the main method
2010-11-21T04:11:05  <delt0r> #
2010-11-21T04:11:05  <delt0r> counter ^=(counter << 21);
2010-11-21T04:11:05  <delt0r> #
2010-11-21T04:11:05  <delt0r>                 counter ^=(counter >>> 35);
2010-11-21T04:11:05  <delt0r> #
2010-11-21T04:11:06  <delt0r>                 counter ^=(counter << 4);
2010-11-21T04:11:12  <delt0r> bugger... sorry...
2010-11-21T04:11:34  <delt0r> thats not what it looked like before i hit enter
2010-11-21T04:12:02  <burny> how doe sthat have a repeat of 2^128?
2010-11-21T04:12:12  <delt0r> anyway --this gives an m-sequence that has very nice pretend random properties
2010-11-21T04:12:23  <delt0r> with a period of 2^64-1
2010-11-21T04:12:34  <lavalamp_> is that xorshift?
2010-11-21T04:12:41  <lavalamp_> that's my fave rng
2010-11-21T04:12:46  <burny> k
2010-11-21T04:12:52  <delt0r> the stepping var has a large prime added each step
2010-11-21T04:13:03  <delt0r> that has a period of 2^64
2010-11-21T04:13:27  *** smloh1 <smloh1!4786e1ca@gateway/web/freenode/ip.71.134.225.202> has quit IRC (Quit: Page closed)
2010-11-21T04:13:32  <burny> isn't that actualyl slower than mersene?
2010-11-21T04:13:33  <delt0r> my output is the higher order bits of (counter+stepped)
2010-11-21T04:13:44  <delt0r> the low order bit is not very random
2010-11-21T04:13:49  <delt0r> burny: no
2010-11-21T04:13:57  <delt0r> well not in my tests
2010-11-21T04:14:01  <delt0r> and its non linear
2010-11-21T04:14:26  <burny> oh, ya, nm.. I was looking at the 'double' stuff..
2010-11-21T04:14:36  <delt0r> a really good SSE imp may be
2010-11-21T04:14:39  <burny> so ya, overall it should be pretty equal to mersense I think
2010-11-21T04:14:43  <burny> for performane
2010-11-21T04:14:53  <delt0r> but then a really good SSE imp of this would be very very fast
2010-11-21T04:15:16  <burny> is it a common thing for prng's to have low bit's not be good?
2010-11-21T04:15:29  <delt0r> yes
2010-11-21T04:15:32  <burny> cause mine doesn't have that problem,
2010-11-21T04:15:43  <delt0r> when you use the stepped+=largePrime
2010-11-21T04:15:55  <burny> I tested each bit independantly, through a variety of rng tests, and they all passed all of them
2010-11-21T04:16:06  <delt0r> also using add and masking makes it more non-linear
2010-11-21T04:16:20  <delt0r> for my simulations it was important that it was non linear
2010-11-21T04:16:29  <burny> non-linear?
2010-11-21T04:16:38  <delt0r> burny: all but the last bit does
2010-11-21T04:16:42  <delt0r> for that one
2010-11-21T04:17:01  <delt0r> however i cant find the ref for where i got this... it had lots of other generators
2010-11-21T04:17:15  <delt0r> like lagged Fibbonachi generators etc
2010-11-21T04:17:26  <delt0r> and the constants that works well with them
2010-11-21T04:17:50  <delt0r> but i my experieance twister was never fast.
2010-11-21T04:17:54  <burny> wish I knew math, so I could prove my rng is fit/unfit for stuff
2010-11-21T04:18:08  <delt0r> what do you do?
2010-11-21T04:18:19  <burny> crazy xor tricks
2010-11-21T04:18:29  <contestbot> Top 10 players: bocsimacko(3392), Raschi(3167), _iouri_(3154), bix0r4ever(3146), dabino(3112), dmj111(3109), Manwe56(3097), _Astek_(3090), protocolocon(3082), davidjliu(3077)
2010-11-21T04:20:31  <burny> when I tested it before about 5years ago, found about 8 things on internet(chi square, entropy, etc.), and it passed all those
2010-11-21T04:22:34  <burny> and I ran it brute forcing for repeats, and it generated the same 32bit value like 50times in the first 200billion values, and when I masked off various bits, the number of occurances of repeats in the pattern was appropriate(and the distance between repeats, passed the tests also)
2010-11-21T04:24:07  <burny> eg. if the low 8bits of the first numbers is 13, 28, 182, 97, 56.. the distance between occurances of '13, 28', is random, but occurs roughly 1 in 65536 numbers(for larger sets)
2010-11-21T04:25:25  <delt0r> burny: try the die harder tests
2010-11-21T04:25:41  <burny> plural?
2010-11-21T04:25:45  <delt0r> but you know--for the most part it probably won't matter...
2010-11-21T04:25:55  <delt0r> die hard (first movie )
2010-11-21T04:26:03  <delt0r> die harder (die hard 2)
2010-11-21T04:26:11  <delt0r> second set of tests
2010-11-21T04:26:20  <burny> ahhh
2010-11-21T04:26:21  <burny> k
2010-11-21T04:26:26  <delt0r> oh year  --its lots of tests
2010-11-21T04:26:34  <delt0r> a lcg passes many of them
2010-11-21T04:26:43  <delt0r> twister pass more
2010-11-21T04:27:40  <delt0r> for some things it really matters... for most even a lcg is probably good enough
2010-11-21T04:27:59  <burny> lcg only passed one test, when I was checking them
2010-11-21T04:28:23  <delt0r> one with well chosen constant passes quite a  few
2010-11-21T04:28:45  <delt0r> ie java.util.random is not bad
2010-11-21T04:28:54  <delt0r> just slow with the mod operation
2010-11-21T04:29:07  <delt0r> for a linear rng anyway
2010-11-21T04:32:13  <burny> the one that I found, made the most rng's fail.
2010-11-21T04:32:38  <burny> was to simply render the stream as pixels, in a x*y image
2010-11-21T04:33:00  <burny> and then look at it, and see if there's any obvious patterns
2010-11-21T04:33:53  *** antimatroid1 <antimatroid1!~Nick@125.0395.dsl.mel.iprimus.net.au> has quit IRC (Ping timeout: 240 seconds)
2010-11-21T04:34:11  <burny> some specific x,y values make much more obvious patterns.. but pretty much any value makes the patterns rather apparent
2010-11-21T04:35:56  <burny> sleepy times
2010-11-21T04:35:58  <burny> later all
2010-11-21T04:46:49  *** Mjothvitnir <Mjothvitnir!432ace74@gateway/web/freenode/ip.67.42.206.116> has quit IRC (Quit: Page closed)
2010-11-21T04:46:53  *** antimatroid <antimatroid!~Nick@188.153.dsl.mel.iprimus.net.au> has joined #aichallenge
2010-11-21T04:56:53  *** McLeopold <McLeopold!~scott@c-98-247-248-39.hsd1.wa.comcast.net> has quit IRC (Quit: Leaving)
2010-11-21T04:59:04  <antimatroid> someones username on zerovis is my last name aha
2010-11-21T05:13:52  *** Appleman1234 <Appleman1234!~Appleman1@CPE-60-226-176-19.qld.bigpond.net.au> has quit IRC (Ping timeout: 240 seconds)
2010-11-21T05:15:07  *** espes <espes!espes@123-243-221-248.tpgi.com.au> has quit IRC (Read error: Connection reset by peer)
2010-11-21T05:15:51  <delt0r> burny: http://www.jstatsoft.org/v08/i14/paper and his other papers...
2010-11-21T05:18:31  <contestbot> Top 10 players: bocsimacko(3392), Raschi(3167), bix0r4ever(3145), _iouri_(3144), oldman(3123), dmj111(3107), davidjliu(3099), Manwe56(3093), protocolocon(3080), dabino(3064)
2010-11-21T05:20:16  *** yasith_ <yasith_!~yasith@112.134.127.210> has quit IRC (Ping timeout: 240 seconds)
2010-11-21T05:20:38  *** yasith_ <yasith_!~yasith@112.134.121.77> has joined #aichallenge
2010-11-21T05:21:37  <delt0r> burny: http://www.javagaming.org/index.php/topic,18426.0.html
2010-11-21T05:22:54  <delt0r> ops
2010-11-21T05:22:59  <delt0r> thats a worng link
2010-11-21T05:23:28  <delt0r> burny: http://wwwmaths.anu.edu.au/~brent/pd/CTAC06t4.pdf
2010-11-21T05:37:06  *** phreeza <phreeza!~phreeza@f054166104.adsl.alicedsl.de> has joined #aichallenge
2010-11-21T05:47:45  *** Appleman1234 <Appleman1234!~Appleman1@CPE-60-226-176-19.qld.bigpond.net.au> has joined #aichallenge
2010-11-21T05:58:00  *** jaspervdj <jaspervdj!~jaspervdj@zeus.ugent.be> has joined #aichallenge
2010-11-21T06:18:30  <contestbot> Top 10 players: bocsimacko(3369), Raschi(3159), _iouri_(3135), bix0r4ever(3130), oldman(3097), dmj111(3094), Manwe56(3086), protocolocon(3072), dabino(3064), wagstaff(3057)
2010-11-21T06:19:40  *** Mistmanov <Mistmanov!4da38f1c@gateway/web/freenode/ip.77.163.143.28> has quit IRC (Quit: Page closed)
2010-11-21T06:22:11  <Error323> is the website down or something?
2010-11-21T06:27:10  <lavalamp_> appears so
2010-11-21T06:31:12  *** medrimonia <medrimonia!~Medrimoni@ABordeaux-158-1-116-7.w90-60.abo.wanadoo.fr> has joined #aichallenge
2010-11-21T06:31:19  <medrimonia> Hello everyone
2010-11-21T06:32:39  *** bobng <bobng!~nigg_free@2607:f358:1:fed5:20:0:1767:1> has joined #aichallenge
2010-11-21T06:33:30  <antimatroid> evening
2010-11-21T06:35:07  <medrimonia> Grr I almost halves of my games which are time out
2010-11-21T06:39:57  *** bobng <bobng!~nigg_free@2607:f358:1:fed5:20:0:1767:1> has quit IRC (Quit: ZNC - http://znc.sourceforge.net)
2010-11-21T06:44:20  <medrimonia> Error323: You just resubmitted?
2010-11-21T06:48:31  <antimatroid> @time
2010-11-21T06:48:31  <contestbot> antimatroid: 06:48 AM, November 21, 2010
2010-11-21T06:48:43  <antimatroid> is that time an hour out?
2010-11-21T06:48:54  <antimatroid> i think it is
2010-11-21T06:48:57  <antimatroid> in that case, then yes
2010-11-21T06:52:23  <medrimonia> Does the compiling has an issue now?
2010-11-21T06:52:36  <medrimonia> Nov 21st 05:49:03  ready to be compiled               Unknown
2010-11-21T06:52:58  <antimatroid> it still hasn't compiled for error
2010-11-21T06:53:05  <antimatroid> it looks like yeah maybe
2010-11-21T06:53:50  <medrimonia> He has the same message as me
2010-11-21T06:54:10  <medrimonia> And I'm pretty sure I did put all the files in the tar.gz
2010-11-21T06:54:49  <Error323> ye same here
2010-11-21T06:55:13  <antimatroid> jaznert: you about :)
2010-11-21T06:55:19  <antimatroid> janzert**
2010-11-21T06:57:47  <medrimonia> I'm unlucky today^^
2010-11-21T07:03:47  *** bobng <bobng!~nigg_free@2607:f358:1:fed5:20:0:1767:1> has joined #aichallenge
2010-11-21T07:05:10  *** choas <choas!~lars@p578F6A62.dip.t-dialin.net> has joined #aichallenge
2010-11-21T07:07:31  *** greghaynes <greghaynes!~irc@2607:f2f8:a490::2> has quit IRC (Read error: Operation timed out)
2010-11-21T07:07:46  *** p4p4 <p4p4!~chatzilla@82.113.106.28> has joined #aichallenge
2010-11-21T07:15:47  *** mceier <mceier!~mceier@chello089077208118.chello.pl> has joined #aichallenge
2010-11-21T07:18:30  <contestbot> Top 10 players: bocsimacko(3369), Raschi(3159), _iouri_(3135), bix0r4ever(3130), oldman(3097), dmj111(3094), Manwe56(3086), protocolocon(3072), dabino(3064), wagstaff(3057)
2010-11-21T07:20:25  *** yasith_ <yasith_!~yasith@112.134.121.77> has quit IRC (Ping timeout: 252 seconds)
2010-11-21T07:25:31  *** Appleman1234 <Appleman1234!~Appleman1@CPE-60-226-176-19.qld.bigpond.net.au> has quit IRC (Quit: Leaving)
2010-11-21T07:32:46  *** yasith_ <yasith_!~yasith@112.134.122.33> has joined #aichallenge
2010-11-21T07:52:04  *** antimatroid1 <antimatroid1!~Nick@103.221.dsl.mel.iprimus.net.au> has joined #aichallenge
2010-11-21T07:55:17  *** antimatroid <antimatroid!~Nick@188.153.dsl.mel.iprimus.net.au> has quit IRC (Ping timeout: 276 seconds)
2010-11-21T08:02:03  *** Palmik <Palmik!~Palmik@palmikpc.sks6.muni.cz> has joined #aichallenge
2010-11-21T08:18:31  <contestbot> Top 10 players: bocsimacko(3369), Raschi(3159), _iouri_(3135), bix0r4ever(3130), oldman(3097), dmj111(3094), Manwe56(3086), protocolocon(3072), dabino(3064), wagstaff(3057)
2010-11-21T08:21:06  *** delt0r___ <delt0r___!~delt0r@178-191-73-252.adsl.highway.telekom.at> has joined #aichallenge
2010-11-21T08:22:35  *** delt0r <delt0r!~delt0r@188-22-168-225.adsl.highway.telekom.at> has quit IRC (Ping timeout: 245 seconds)
2010-11-21T08:44:23  <medrimonia> Is it me or the ranking isn't updated anymore
2010-11-21T09:08:01  *** greghaynes <greghaynes!~irc@2607:f2f8:a490::2> has joined #aichallenge
2010-11-21T09:18:32  <contestbot> Top 10 players: bocsimacko(3369), Raschi(3159), _iouri_(3135), bix0r4ever(3130), oldman(3097), dmj111(3094), Manwe56(3086), protocolocon(3072), dabino(3064), wagstaff(3057)
2010-11-21T09:32:30  <janzert> did compilation finally happen?
2010-11-21T09:34:54  <medrimonia> yep no problem
2010-11-21T09:35:08  <medrimonia> But there isn't anymore update of the rankings it seems
2010-11-21T09:36:25  <janzert> yeah, just fixing that. shut it off for some database work and forgot to re-enable it.
2010-11-21T09:37:20  *** bhasker <bhasker!~bhasker@adsl-99-25-118-235.dsl.pltn13.sbcglobal.net> has joined #aichallenge
2010-11-21T09:42:42  <Zoost> oh man... this is so wrong. I win agaisnt +ELO bots and lose constantly to 2...well...not so good bots which drop my ELO quite well :P
2010-11-21T09:48:20  <Error323> ah
2010-11-21T09:51:07  <Error323> janzert: has the server gotten more pounding since the countdown estimate? =)
2010-11-21T09:53:12  <janzert> I'm not really sure, it does seem a bit more but I haven't looked at any numbers
2010-11-21T10:00:06  *** choas <choas!~lars@p578F6A62.dip.t-dialin.net> has quit IRC (Quit: leaving)
2010-11-21T10:01:33  <janzert> http://ai-contest.com/visualizer.php?game_id=8199614
2010-11-21T10:02:05  <janzert> a rare top rated draw
2010-11-21T10:02:45  <antimatroid1> god damnit
2010-11-21T10:02:56  <antimatroid1> brb, my internet wont load pages, but now you have made me :P
2010-11-21T10:04:33  *** antimatroid <antimatroid!~Nick@116.78-27-211.dynamic.dsl.mel.iprimus.net.au> has joined #aichallenge
2010-11-21T10:05:02  <antimatroid> it's good productivity wise when i can't load websites
2010-11-21T10:07:07  *** antimatroid1 <antimatroid1!~Nick@103.221.dsl.mel.iprimus.net.au> has quit IRC (Ping timeout: 245 seconds)
2010-11-21T10:09:53  <Error323> lol
2010-11-21T10:09:56  <Error323> hehe
2010-11-21T10:18:31  <contestbot> Top 10 players: bocsimacko(3353), oldman(3175), medrimonia(3153), Raschi(3150), dmj111(3146), bix0r4ever(3135), _iouri_(3127), Manwe56(3109), protocolocon(3065), davidjliu(3064)
2010-11-21T10:24:02  <burny> woah.. my dumb bot is ranked top700
2010-11-21T10:25:04  <burny> totally gona have to enter this thing for real now :\
2010-11-21T10:26:38  *** Naktibalda <Naktibalda!~quassel@cpc5-heme9-2-0-cust144.9-1.cable.virginmedia.com> has joined #aichallenge
2010-11-21T10:27:52  *** davidjliu <davidjliu!473a6d0e@gateway/web/freenode/ip.71.58.109.14> has joined #aichallenge
2010-11-21T10:28:12  <Error323> who owns that GA bot on tcp?
2010-11-21T10:28:14  <Error323> its insane
2010-11-21T10:28:30  *** phreeza <phreeza!~phreeza@f054166104.adsl.alicedsl.de> has quit IRC (Quit: Computer has gone to sleep)
2010-11-21T10:28:50  *** Blkt <Blkt!~user@net-93-145-62-81.cust.dsl.teletu.it> has joined #aichallenge
2010-11-21T10:30:29  <antimatroid> no idea
2010-11-21T10:30:53  <antimatroid> they were shit against me before
2010-11-21T10:31:04  <antimatroid> and i still haven't added everything into my code and it was buggy as before
2010-11-21T10:31:08  <antimatroid> http://72.44.46.68/getplayer?player=s
2010-11-21T10:31:13  *** yasith__ <yasith__!~yasith@112.134.120.183> has joined #aichallenge
2010-11-21T10:32:26  <antimatroid> http://72.44.46.68/canvas?game_id=1124293
2010-11-21T10:32:26  *** yasith_ <yasith_!~yasith@112.134.122.33> has quit IRC (Ping timeout: 276 seconds)
2010-11-21T10:32:30  <antimatroid> that game was actually kinda cool
2010-11-21T10:34:25  <antimatroid> i'm working on updating my future states as they're used rather than the whole lot every node now, hopefully that will speed things up quite a bit and i can make a bigger game tree
2010-11-21T10:39:25  *** fawek <fawek!~faw@chello089079043119.chello.pl> has joined #aichallenge
2010-11-21T10:39:50  *** nann <nann!~apophis@80-219-211-230.dclient.hispeed.ch> has joined #aichallenge
2010-11-21T10:43:00  *** Apophis <Apophis!~apophis@80-219-211-230.dclient.hispeed.ch> has quit IRC (Ping timeout: 245 seconds)
2010-11-21T10:47:59  *** Blkt <Blkt!~user@net-93-145-62-81.cust.dsl.teletu.it> has quit IRC (Ping timeout: 264 seconds)
2010-11-21T10:56:16  *** phreeza <phreeza!~phreeza@68-076.eduroam.RWTH-Aachen.DE> has joined #aichallenge
2010-11-21T10:57:01  <foucist> Error323: what, a genetic algorithm bot? is it any good?
2010-11-21T10:57:48  <Error323> don't know if its ga or sth but that's what its called on tcp
2010-11-21T10:59:42  <foucist> Error323: where can i see it is there avisualizer for that
2010-11-21T11:00:01  <Error323> http://72.44.46.68/getratings just scroll to the bottom spam
2010-11-21T11:00:06  <Error323> and click the name
2010-11-21T11:00:53  <foucist> thanks!
2010-11-21T11:01:21  <antimatroid> my bot finally seems to be coming together
2010-11-21T11:01:29  <antimatroid> it's not going to win but it should do alright
2010-11-21T11:01:31  <Error323> mine is falling apart :/
2010-11-21T11:01:41  <burny> what elo anti?
2010-11-21T11:01:51  <antimatroid> burny: haven't got one on the main site
2010-11-21T11:01:56  <burny> ah
2010-11-21T11:01:59  <antimatroid> i've mostly been testing locally
2010-11-21T11:02:07  <antimatroid> http://72.44.46.68/getplayer?player=s
2010-11-21T11:02:34  <antimatroid> i ran some tcp games before, but it hasn't got all my featurs running and had some pretty awful bugs sending ships at their neutral claims before it was smart to
2010-11-21T11:02:47  <burny> 3 turn0 losses?.. win!
2010-11-21T11:03:10  <antimatroid> :P
2010-11-21T11:03:41  <antimatroid> i also haven't got it "leaving" ships at home that leave the planet vulnerable atm
2010-11-21T11:03:45  <antimatroid> that often ruins a game
2010-11-21T11:03:52  <antimatroid> but that'll be added again before the end of the week
2010-11-21T11:04:04  <antimatroid> redistribution is desperately needed too
2010-11-21T11:04:24  <burny> ya, I'm just now, starting to do stuff that isn't doable in under 10mins :\
2010-11-21T11:05:54  <antimatroid> yeah, i had to throw some ideas out
2010-11-21T11:06:02  <antimatroid> at the beginning i had the hopes of using simplex
2010-11-21T11:06:07  <antimatroid> never going to happen
2010-11-21T11:06:13  <burny> what's that?
2010-11-21T11:06:24  <antimatroid> algorithm used for solving linear programs
2010-11-21T11:06:31  <burny> ahh
2010-11-21T11:06:56  <antimatroid> you can use it for finding mixed strategy nash equilibrium in 2 player zero sum games
2010-11-21T11:06:57  <foucist> antimatroid: just make it attack its own planets!
2010-11-21T11:07:22  <antimatroid> foucist: that doesn't work properly
2010-11-21T11:07:44  <antimatroid> say i consider taking the best planet first, it will then consider sending fleets from the second best planet, even if i don't want to actually use them
2010-11-21T11:08:14  <antimatroid> i have a method of having them mock send all their fleets at me, then clearing it once i've picked makable moves before moving onto the next node in the game tree
2010-11-21T11:11:20  <antimatroid> dmj111: looking better again :P
2010-11-21T11:11:34  <foucist> antimatroid: do you need to go down the nodes in the tree, then reverse direction and go back up all the nodes and re-normalize the makable moves relative to all the moves the other nodes are making?
2010-11-21T11:11:41  <burny> hrm.. can't decide.. if I'm winning by 50ships, and there's 40 turns left, and we have equal growth.. should I attack, or just wait it out?
2010-11-21T11:12:14  <antimatroid> foucist: no
2010-11-21T11:12:17  *** narnach <narnach!~Narnach@195-241-27-75.ip.telfort.nl> has joined #aichallenge
2010-11-21T11:12:17  *** narnach is now known as narnach_mb
2010-11-21T11:12:30  <foucist> burny: doesn't that depend on distance, likelihood of other bot attacking, etc
2010-11-21T11:12:31  <j3camero> antimatroid: what's stopping you from using Simplex?
2010-11-21T11:12:38  <antimatroid> j3camero: not enough time
2010-11-21T11:12:43  <antimatroid> too much information
2010-11-21T11:12:51  <burny> foucist, to some extent.. but not really
2010-11-21T11:13:08  <antimatroid> burny: yeah, i don't make decisions like that
2010-11-21T11:14:05  <burny> ya, me either..
2010-11-21T11:14:24  <burny> but I'm wanting to add code to prevent it from turning 80ships into 2growth, if there's 40turns remaining
2010-11-21T11:14:47  <foucist> what about a weight that's affected by the number of turns left
2010-11-21T11:14:59  <burny> figure attacking the opponent is basically the same logic
2010-11-21T11:15:00  <foucist> maybe an exponential weight
2010-11-21T11:15:15  <foucist> so that it's fairly consistent in the first 120 turns
2010-11-21T11:15:26  <antimatroid> j3camero: one of my ideas was to use decision trees to generate pure strategies for each player, then make a 2 player zero sum simultaneous game from that, then find pure strategy nash equilibria, pick the most desirable one i could enforce and play that, if none existed, calculate the mixed strategy one using simplex, with the bonus thing being that simplex goes towards the optimal solution while maintaining feasible so
2010-11-21T11:16:12  <foucist> antimatroid: i'm trying to figure out how the heck to encode these strategies you speak of..
2010-11-21T11:16:22  <foucist> i wnat to apply GA to it
2010-11-21T11:16:22  <burny> what exactly is pure vs mixed strategy?
2010-11-21T11:16:24  <antimatroid> i have code that sets up said zero sum game though, but doesn't seem to work well as I can't consider very many pure strategies for each player
2010-11-21T11:17:01  <antimatroid> burny: a pure strategy in rps is either playing rock, paper, or scissors, a mixed strategy is saying you might play any of them with equal probability
2010-11-21T11:17:01  *** Mistmanov <Mistmanov!4da38f1c@gateway/web/freenode/ip.77.163.143.28> has joined #aichallenge
2010-11-21T11:17:24  <antimatroid> then when you play, you use some random process to pick your strategy based off the probability distrbution you assigned
2010-11-21T11:18:33  <contestbot> Top 10 players: bocsimacko(3347), dmj111(3258), oldman(3179), Raschi(3150), medrimonia(3143), bix0r4ever(3143), _iouri_(3125), Manwe56(3118), davidjliu(3093), protocolocon(3069)
2010-11-21T11:19:13  <foucist> antimatroid: how do you generate the pure strategies for this ai challenge..
2010-11-21T11:19:39  <antimatroid> for me, a single "move" consists of a vector of fleets, all fleets have the same destination, but have different source planets and send times
2010-11-21T11:19:50  <antimatroid> a move is a pure strategy
2010-11-21T11:19:58  <antimatroid> although you could consider a pure strategy to be a vector of moves
2010-11-21T11:20:32  <antimatroid> or just allow a move to have multiple destinations, but that's the same as the second option just implemented differently
2010-11-21T11:21:05  <foucist> but how does that 'strategy' have the concept of what the enemy is doing, enemy's relative strength and location etc
2010-11-21T11:21:38  *** rebelxt_ <rebelxt_!181c0b18@gateway/web/freenode/ip.24.28.11.24> has joined #aichallenge
2010-11-21T11:22:09  <antimatroid> at each node in my game tree, i pick out the best x moves as described above, then create children, one where each move is implemented, then i just alternate between players on each layer
2010-11-21T11:22:35  <antimatroid> i rank terminal nodes based off who won otherwise the expected difference in ships at the end of the game
2010-11-21T11:22:35  *** amstan <amstan!~alex@76-10-183-61.dsl.teksavvy.com> has quit IRC (Ping timeout: 245 seconds)
2010-11-21T11:22:40  <antimatroid> then run minimax on that baby
2010-11-21T11:23:56  <lavalamp_> antimatroid: you can't have two targets at once?
2010-11-21T11:24:01  <antimatroid> to do the zero sum one, i did something like the game tree, but two of them, with each player just making moves for themselves, then i'd take the combinations of moves to get to each terminal node, and create a pure strategy out of each of those, then my game matrix consisted of gamestates where i then implemented both players moves and evaluated the states for payoffs
2010-11-21T11:24:32  <antimatroid> lavalamp_: I consider doing nothing, so say moves a and b are feasible, then i have that in my game tree as (a, nothing, b)
2010-11-21T11:24:42  <foucist> lavalamp_: btw, if you're building a payoff table and treating this like prisoner's dilemma, do you encode tit for tat? ;)
2010-11-21T11:24:58  <antimatroid> i don't do anything like that
2010-11-21T11:25:05  <antimatroid> i originally wanted to, but have never got around to it
2010-11-21T11:25:25  <antimatroid> i have ideas of using bayesian probability to update a model of the opponent
2010-11-21T11:25:28  <lavalamp_> foucist: unfortunately I won't have time to start identifying the strategies my opponent is using... :'(
2010-11-21T11:25:33  <antimatroid> but even my above ideas were too over the top
2010-11-21T11:25:37  <lavalamp_> it was my original plan
2010-11-21T11:25:46  <antimatroid> had* ideas
2010-11-21T11:25:47  <j3camero> antimatroid where are you ranked now? I expected to find you in the top 100. Tell me that you're not >100 with that sort of sophisticated strategy...
2010-11-21T11:26:00  <foucist> j3camero: it's not in place
2010-11-21T11:26:01  <antimatroid> j3camero: no bot is currently uploaded
2010-11-21T11:26:09  <antimatroid> i'd be disappointed not to be in the top 50 at least
2010-11-21T11:26:20  <j3camero> Ah that makes more sense.
2010-11-21T11:26:26  <antimatroid> http://ai-contest.com/profile.php?user_id=3945
2010-11-21T11:26:28  <antimatroid> that's my profile
2010-11-21T11:26:34  <j3camero> Have you ever had a bot in place, or are you waiting until the last day to submit?
2010-11-21T11:26:34  <lavalamp_> I'm barely staying in the 200's
2010-11-21T11:26:39  <antimatroid> j3camero: want to clear my about me for me? i'm working by myself in the end
2010-11-21T11:26:49  <antimatroid> i did ages ago, was usually top 30 somewhere
2010-11-21T11:26:52  <antimatroid> but that was a long time ago
2010-11-21T11:27:31  <lavalamp_> I need more time :(
2010-11-21T11:27:43  <j3camero> You should submit something so you can see your ranking!
2010-11-21T11:27:46  <antimatroid> one could always use more time on this i think
2010-11-21T11:27:48  <lavalamp_> why did I not find out about this for a whole month :(
2010-11-21T11:28:01  <antimatroid> j3camero: i'm working on a final copy, it hasn't got everything added in yet
2010-11-21T11:28:11  <antimatroid> i will have it uploaded a couple of days before to fix issues
2010-11-21T11:28:55  <j3camero> lavalamp_ you dont need more time, you just need more focus. You need to reconsider your priorities. Do you have a wife and kids now? Turf 'em. That will give you an extra couple hours each day to work on your AI contest entry right there.
2010-11-21T11:29:29  <lavalamp_> j3camero: already did. and thinking about taking time off work, too
2010-11-21T11:29:55  <j3camero> That's what we like to see, man.
2010-11-21T11:29:59  * antimatroid hasn't worked in months, i should do something about that
2010-11-21T11:30:25  <antimatroid> it can wait another week
2010-11-21T11:30:26  <foucist> lavalamp_: well considering it's just a week left..  ;)
2010-11-21T11:30:26  <lavalamp_> this contest is worse than MMO's
2010-11-21T11:30:33  <j3camero> I didnt work for 12 months last year. I was soo poor. So glad to have an income once again. It's quite nice.
2010-11-21T11:30:57  <antimatroid> the only video games i play are minesweeper and flash games
2010-11-21T11:31:05  <antimatroid> although i play a lot of flash games
2010-11-21T11:31:11  <antimatroid> tga motocross 3 <3
2010-11-21T11:31:13  <j3camero> Speaking of, I should play Galcon now.
2010-11-21T11:32:04  <lavalamp_> interesting thing about this game, even a horrible bot is impossible to beat in human mode in jbotman
2010-11-21T11:32:15  <antimatroid> nah
2010-11-21T11:32:18  <burny> human mode?
2010-11-21T11:32:22  <burny> and, what's jbotman?
2010-11-21T11:32:28  <antimatroid> i can still beat my bot if i watch what it does on the first few moves and ruin that
2010-11-21T11:32:32  <lavalamp_> jbotmanager has a human vs bot mode
2010-11-21T11:32:37  <antimatroid> depending on the map
2010-11-21T11:32:49  <antimatroid> some maps my bot manages to get it right
2010-11-21T11:32:57  <lavalamp_> you must be faster with your mouse than me ^^
2010-11-21T11:33:06  <antimatroid> you can up the time limit to whatever you want :P
2010-11-21T11:33:11  <antimatroid> then hit enter when you're done
2010-11-21T11:33:19  <lavalamp_> oh ok that would help
2010-11-21T11:33:34  <antimatroid> curny: http://ai-contest.com/forum/viewtopic.php?f=18&t=751
2010-11-21T11:33:40  <antimatroid> burny*
2010-11-21T11:34:58  <rebelxt_> j3camero: <3 galcon. i'm also about to play some 1on1
2010-11-21T11:35:23  *** amstan <amstan!~AndChat@207.61.84.210> has joined #aichallenge
2010-11-21T11:35:57  <burny> I'd be down for some 1on1's
2010-11-21T11:36:32  <rebelxt_> Burny, jump in. i'm "rebelxt" in Theta3 room at the moment.
2010-11-21T11:36:34  <foucist> my galcon strategy is typically to attack the enemy right away
2010-11-21T11:36:40  <foucist> seems to be the only way for me to stay ahead
2010-11-21T11:37:19  <foucist> altho i do try to colonize a couple big planets nearby first
2010-11-21T11:38:48  <burny> I don't see a theta3
2010-11-21T11:39:18  <rebelxt_> burny, are you in multiplayer lobby?
2010-11-21T11:39:22  <burny> ya
2010-11-21T11:39:45  <j3camero> How could we scale the Galcon concept up into a much bigger persistent world?
2010-11-21T11:40:10  <Zoost> man... you should just let the servers run after the contest has ended and call it a "ladder". This has so many possibilities and a huge learning curve that its a shame it ends now :(
2010-11-21T11:40:28  <antimatroid> j3camero: what do you mean?
2010-11-21T11:40:36  <rebelxt_> burny: which rooms do you see then? are you using iphone or android?
2010-11-21T11:40:37  <antimatroid> i'm pretty keen for tryin n-player and massive maps later
2010-11-21T11:40:42  <antimatroid> i think mcleopold will be keen too
2010-11-21T11:40:46  <burny> flash
2010-11-21T11:41:05  <Zoost> Im pretty certain magic will happen once people see what bocsimacko does
2010-11-21T11:41:07  <burny> lol, maybe that's the problem?
2010-11-21T11:41:09  <rebelxt_> n-player FFA (free for all) won't work
2010-11-21T11:41:41  <rebelxt_> however cooperative like 2on2, 3on3 would work great.. but then you need to team up with other bots
2010-11-21T11:41:49  <antimatroid> i see ff and think fast fourier
2010-11-21T11:42:13  <antimatroid> why wont ffa work though?
2010-11-21T11:42:25  <antimatroid> should work fine if people set up to take idiots out properly
2010-11-21T11:42:28  <burny> because ffa has a TOTALLY different strategy
2010-11-21T11:42:33  <rebelxt_> antimatroid: play some 4-player FFA in galcon and you will immediately see stand-offs
2010-11-21T11:42:43  <rebelxt_> whoever attacks first, always loses
2010-11-21T11:43:10  <burny> not always: but most often
2010-11-21T11:43:35  <burny> I've had a couple 'proper' wins.. but those are always pure position
2010-11-21T11:43:37  <antimatroid> i think the best thing would be to put them on fairly big maps and big starting distances
2010-11-21T11:43:45  <antimatroid> most bots don't expand similarly
2010-11-21T11:44:16  <burny> 4vs4 ffa, could be 'ok', if there's 2 on one side, and 2 on the other, with a larger distance between them
2010-11-21T11:44:26  <burny> so it's basically 2 games of 2vs2
2010-11-21T11:44:35  <burny> but then.. why bother making it 4?
2010-11-21T11:44:43  <rebelxt_> antimatroid: it does not matter, players expand first, then there is a standoff, since as soon as you attack, you are vulnerable to someone backstabbing you with very little army ;)
2010-11-21T11:45:16  <burny> FFA particularly falls apart because of..
2010-11-21T11:45:25  <antimatroid> i don't think there would be that many standoffs
2010-11-21T11:45:31  <burny> if 3 players gain 1 planet, and the other is the by far best player, gainst 3 planets..
2010-11-21T11:45:34  <rebelxt_> j3camero: what do you mean by persistent world?
2010-11-21T11:45:44  <burny> then it quickly becomes a 4planets vs 6planets game
2010-11-21T11:46:03  <amstan> hi
2010-11-21T11:46:09  <antimatroid> hi
2010-11-21T11:46:31  <rebelxt_> antimatroid: i've player 35,000+ galcon games, so i've seen thousands of standoffs in FFA games except for 1v1 of course.. you can watch some 4-player games online to see what i mean
2010-11-21T11:46:37  <lavalamp_> you'd have to add some dynamic like you only get to keep a neutral for a certain amount of time and then it reverts to being neutral
2010-11-21T11:47:03  <burny> the easiest/best strat I can think of for ffa games..
2010-11-21T11:47:18  <burny> is that planets growth decreases as it's population increases
2010-11-21T11:47:32  <antimatroid> what about proportional growth
2010-11-21T11:47:40  <antimatroid> that'd be hard as
2010-11-21T11:47:41  <lavalamp_> and maybe fog of war and allow fleets to change destination, and to travel to arbitrary points instead of just planets
2010-11-21T11:47:43  <burny> and, when a player dies, 100population is distributed amongst the people who killed it(based on ships killed)
2010-11-21T11:47:57  <antimatroid> changing direction of fleets is messy with the ceiling integer calc
2010-11-21T11:48:32  <amstan> @rankings
2010-11-21T11:48:34  <contestbot> amstan: Top 10 players: bocsimacko(3343), dmj111(3285), oldman(3169), medrimonia(3163), Raschi(3131), bix0r4ever(3129), Manwe56(3118), _iouri_(3117), davidjliu(3099), protocolocon(3070)
2010-11-21T11:48:50  <lavalamp_> whoa dmj must've figured something out
2010-11-21T11:48:55  <j3camero> What I mean by persistent world, is to create a game where there is just "one big world" where all the players connect to and play. An MMO basically. It would be cool if we could create a nice strategic MMO to use for an AI contest.
2010-11-21T11:49:16  <antimatroid> i think it's going to be quite a toss up as to who finishes where from 2-40 or so
2010-11-21T11:49:21  <rebelxt_> if fleet redirection were allowed in this challenge like in Galcon, we would be seeing some crazy games from top bots
2010-11-21T11:49:59  <antimatroid> it's hard to rank that
2010-11-21T11:50:08  <antimatroid> if people are playing for different periods of time etc.
2010-11-21T11:50:10  *** amstan <amstan!~AndChat@207.61.84.210> has quit IRC (Quit: Bye)
2010-11-21T11:50:16  <antimatroid> i guess you can reset once the final tournament runs
2010-11-21T11:50:39  <burny> if fleet redirection were allowed, it'd be pretty dumb
2010-11-21T11:50:48  <burny> there'd be thousands of fleets flying around near the middle
2010-11-21T11:51:09  <antimatroid> burny: actually, probably nice voronoi diagrams around the planets :)
2010-11-21T11:51:45  <antimatroid> i think it'd be kind of cool
2010-11-21T11:52:06  <antimatroid> what you would probably end up doing is "creating your own planets" each move, which are your "ideal locations" to have fleets, and aim for those points
2010-11-21T11:52:52  <antimatroid> actually, i think it could maybe work fairly easily
2010-11-21T11:53:10  <antimatroid> you need to have the engine pass back fleets with coordinates though
2010-11-21T11:53:30  <burny> performance would become a way more critical issue
2010-11-21T11:54:15  <antimatroid> it should also be the case that if fleets are within one distance of enemy fleets, they battle each other
2010-11-21T11:54:50  <antimatroid> that'd make it even harder
2010-11-21T11:57:47  <foucist> j3camero: how about bots that play the VGA Planets game, it's an old turn based strategy game
2010-11-21T11:58:17  <foucist> well there's different versions of it though, but i liked the original one
2010-11-21T11:58:52  <foucist> altho i guess it'd be a bit too close to galcon in some ways (the whole space/planet/ships part)
2010-11-21T12:00:02  <rebelxt_> j3camero: that's an interesting idea. will have to think about it
2010-11-21T12:01:16  <foucist> i like that mmo idea defo though
2010-11-21T12:01:55  <rebelxt_> antimatroid: yeah, hard to make ranking work unless there is an open MMO aspect as well as controlled battlegrounds for 1v1 combat
2010-11-21T12:05:31  *** rebelxt_ <rebelxt_!181c0b18@gateway/web/freenode/ip.24.28.11.24> has quit IRC (Quit: Page closed)
2010-11-21T12:06:42  <j3camero> Sweet yeah let me know.
2010-11-21T12:07:12  <j3camero> I am thinking with an MMO type of game ranking would have to work differently. It would be whoever holds the most planets or territory currently.
2010-11-21T12:07:19  <j3camero> Or you just don't have rankings at all.
2010-11-21T12:08:37  *** amstan <amstan!~alex@76-10-183-61.dsl.teksavvy.com> has joined #aichallenge
2010-11-21T12:08:37  *** ChanServ sets mode: +o amstan
2010-11-21T12:08:49  <j3camero> I might take a crack at implementing a big MMO RTS game engine today. I won't care about balance for now, I just want to experiment with the idea.
2010-11-21T12:09:58  <foucist> j3camero: territory, makes it sound like a bit of a Go game there eh
2010-11-21T12:10:19  <j3camero> Yeah. Territory would work well for an RTS-type game.
2010-11-21T12:10:24  <j3camero> Like a big MMO starcraft.
2010-11-21T12:10:31  <foucist> yeah
2010-11-21T12:11:38  <foucist> i never really classified starcraft and other "RTS" games as being strategy games in my mind, i typically consider them to be more tactical than strategic..
2010-11-21T12:11:58  <foucist> in fact the most strategic SC players seem to get is figuring out optimal build orders and that type of thing
2010-11-21T12:12:17  *** bhasker <bhasker!~bhasker@adsl-99-25-118-235.dsl.pltn13.sbcglobal.net> has quit IRC (Quit: bhasker)
2010-11-21T12:12:27  <j3camero> I had this idea that what I could do is create a human-playable simplified knockoff of Starcraft, except make it so there is one huge map where all the players get spawned. This world would be persistent for the life of the contest. There would be an API so you could eventually start to automate the things you're doing in the world.
2010-11-21T12:13:06  <foucist> j3camero: yeah that sounds like a good idea, almost like a MUD
2010-11-21T12:13:23  <foucist> altho i imagine you'd set it up so it'd be easy to have a GUI frontend for it anyways
2010-11-21T12:13:40  <j3camero> There would be fog of war and unexplored territory, so you would not have any idea where on the map you are or how big it is. Those would be closely guarded secrets of the game admins. The players themselves would be left to share information and discover the world.
2010-11-21T12:14:20  <j3camero> Yeah the people would start playing using the GUI, much like playing starcraft. Except that it would also an easy-to-use API so you could bot it eventually.
2010-11-21T12:16:58  <j3camero> I was thinking I could make the edges of the map wrap around to the other side, just to fuck with people. People would be going around looking for the edges. When they fail to find any edges, they would assume the map is infinite and self-generating. But then they would start mapping it and eventually notice how the world works.
2010-11-21T12:17:58  <foucist> pssh, why would anybody assume there'd be edges! that's retarded thinking ;)
2010-11-21T12:18:14  <foucist> edges would make it too easy to setup defensible spots
2010-11-21T12:18:18  <foucist> people would want corners
2010-11-21T12:18:33  <j3camero> Oh yeah. Those would be highly valuable.
2010-11-21T12:18:34  <contestbot> Top 10 players: bocsimacko(3350), dmj111(3252), oldman(3148), medrimonia(3147), bix0r4ever(3134), Raschi(3133), _iouri_(3127), Manwe56(3126), davidjliu(3085), protocolocon(3068)
2010-11-21T12:18:44  <amstan>  heheh, epic
2010-11-21T12:19:26  <foucist> j3camero: actually one of my fave strategy games is this: http://www.windowsgames.co.uk/slay.html
2010-11-21T12:19:31  <foucist> what about a huge version of that
2010-11-21T12:19:40  <foucist> multiplayer etc
2010-11-21T12:20:33  <foucist> there are some very similar games for the iphone
2010-11-21T12:20:41  <amstan> foucist: that looks like wesnoth
2010-11-21T12:20:53  <foucist> amstan: it's nothing like wesnoth
2010-11-21T12:21:02  <foucist> it's nothing like any hex based game out there afaik
2010-11-21T12:21:10  <foucist> it's really simple
2010-11-21T12:21:16  <foucist> and endlessly addictive
2010-11-21T12:21:40  <foucist> give the demo a try if you guys dare! ;)
2010-11-21T12:22:02  <j3camero> foucist: oo I like this Slay game. I should buy it.
2010-11-21T12:22:16  <foucist> j3camero: you're already trying the demo?
2010-11-21T12:22:29  <j3camero> No
2010-11-21T12:22:38  <j3camero> I will though. Just gotta duck out for a second.
2010-11-21T12:23:30  <foucist> slay & galcon are my fave games on my ipod touch heh
2010-11-21T12:25:06  <amstan> j3camero: have you tried minecraft?
2010-11-21T12:25:23  <foucist> tactica is also a knockoff of slay (for iphone..  uniwar is anoter potential clone i haven't tried)
2010-11-21T12:25:41  <foucist> amstan: thinking of minecraft for an ai challenge?
2010-11-21T12:25:53  <amstan> foucist: that would be epic
2010-11-21T12:26:07  <amstan> foucist: a lot of work would be required for map storage, api and visualizer
2010-11-21T12:28:42  <amstan> how about a 2d minecraft like game?
2010-11-21T12:28:55  <amstan> you have to still build, you'll have huge maps
2010-11-21T12:29:09  <amstan> fighting?
2010-11-21T12:29:17  *** perror <perror!~fleury@aldebaran.labri.fr> has joined #aichallenge
2010-11-21T12:30:04  <foucist> i haven't tried minecraft yet, i briefly tried dwarf fortress but the interface was a bit complicated so i didn't get into it
2010-11-21T12:31:44  <amstan> j3camero: the epic thing about minecraft, is that is has very simple physics and stuff, but they can be put together to make very interesting contraptions
2010-11-21T12:32:10  <phreeza> does the server have scipy?
2010-11-21T12:32:22  <phreeza> i think i've asked this a million times, sorry
2010-11-21T12:32:44  <amstan> phreeza: ImportError: No module named scipy
2010-11-21T12:32:50  *** albert_ <albert_!868204f3@gateway/web/freenode/ip.134.130.4.243> has joined #aichallenge
2010-11-21T12:32:58  <phreeza> dang
2010-11-21T12:32:59  <phreeza> thanks
2010-11-21T12:33:57  <antimatroid> amstan: have you played "everybody edits"?
2010-11-21T12:34:09  <amstan> antimatroid: i don't think so, what's that?
2010-11-21T12:34:23  <antimatroid> http://www.kongregate.com/games/QRious/everybody-edits
2010-11-21T12:34:30  <antimatroid> it's like 2d minecraft in a way
2010-11-21T12:36:35  <amstan> antimatroid: interesting
2010-11-21T12:36:47  <amstan> antimatroid: ya, something like that, though i was thinking a from up top view
2010-11-21T12:36:59  <amstan> antimatroid: like a horizontal slice of minecraft
2010-11-21T12:37:39  <antimatroid> what is really needed is some kind of "objective"
2010-11-21T12:38:11  <amstan> a score based thing could work
2010-11-21T12:43:18  <j3camero> Minecraft wont work on my shitty netbook. I still dont really know... what is the "objective" in minecraft?
2010-11-21T12:43:27  <antimatroid> to be a kid
2010-11-21T12:43:31  <amstan> j3camero: there isn't one
2010-11-21T12:43:38  <j3camero> Interesting.
2010-11-21T12:43:40  <antimatroid> it's lego for adults
2010-11-21T12:43:41  <amstan> j3camero: you build/explore/fight mobs
2010-11-21T12:43:51  *** narnach <narnach!~Narnach@195-241-27-75.ip.telfort.nl> has joined #aichallenge
2010-11-21T12:43:59  <j3camero> It sounds like Minecraft could be improved with the addition of weapons.
2010-11-21T12:44:05  <amstan> j3camero: there are
2010-11-21T12:44:11  <j3camero> Oh.
2010-11-21T12:44:19  <amstan> j3camero: there's 2 versions btw, there's classic, which is only building
2010-11-21T12:44:20  <j3camero> So you can go around killing other players?
2010-11-21T12:44:23  <amstan> and alpha
2010-11-21T12:44:27  <antimatroid> there's zombies dude
2010-11-21T12:44:29  <antimatroid> what more do you want?
2010-11-21T12:44:37  <antimatroid> i guess i could use some unicorns and rainbows
2010-11-21T12:44:44  <amstan> which is best played singleplayer, you have to survive monsters, there's infinitely sized maps
2010-11-21T12:45:05  <amstan> j3camero: http://www.youtube.com/watch?v=NszGurDUlHc
2010-11-21T12:45:20  <j3camero> I would really like to be able to fight the other people PVP. Killing NPCs is fine, but the real fun of a game like that would be the ability to kick down somebody else's sandcastle.
2010-11-21T12:45:28  <amstan> j3camero: i agree
2010-11-21T12:46:51  <antimatroid> i'd be surprised if it doesn't come
2010-11-21T12:47:00  <antimatroid> it's something i expect most people would like
2010-11-21T12:47:12  <j3camero> Yo so I want to try to make a simple MMO RTS today. I gotta figure this out. Basically the top-down minecraft idea... but with weapons.
2010-11-21T12:47:26  <amstan> actually, yeah, minecraft multiplayer is going to have pvp soon
2010-11-21T12:47:40  <amstan> he just implemented multiplayer health, till now they were all invincible
2010-11-21T12:47:49  <j3camero> Oh neat. I am going to buy a real computer some time soon. I will have to get addicted to minecraft pretty sure.
2010-11-21T12:48:18  <amstan> j3camero: yeah, do that
2010-11-21T12:48:20  <antimatroid> i've been holding off till the end of this to buy it
2010-11-21T12:48:27  <antimatroid> i've only ever played the crappy free version
2010-11-21T12:48:38  <amstan> antimatroid: they don't compare, lol
2010-11-21T12:48:45  <antimatroid> i know
2010-11-21T12:48:53  <antimatroid> that's the point, i've been able to not play
2010-11-21T12:50:29  <antimatroid> what would be really awesome is a very general game that people can add components to
2010-11-21T12:50:39  <antimatroid> so people could be playing like ea skate and gta in the same world
2010-11-21T12:52:08  <amstan> the idea is to try to make simple rules on things that can have many uses
2010-11-21T12:52:17  <amstan> that not even admins tought about
2010-11-21T12:53:55  *** Accoun <Accoun!~1@static-213-88-2-78.chebnet.ru> has quit IRC ()
2010-11-21T13:01:47  <amstan> j3camero: k.. so.. you go buy minecraft then play it for a week or so, and we think about how to implement it, lol
2010-11-21T13:05:39  *** Accoun <Accoun!~1@static-213-88-2-78.chebnet.ru> has joined #aichallenge
2010-11-21T13:06:46  <antimatroid> that would make for an awesome contest
2010-11-21T13:06:55  <antimatroid> you get to build your own defense systems and shit
2010-11-21T13:06:58  * lavalamp_ has bloody fight with eval function
2010-11-21T13:07:09  <antimatroid> lavalamp_: it's a losing battle :P
2010-11-21T13:07:23  <lavalamp_> :(
2010-11-21T13:07:34  <antimatroid> i go for simplicity with my evaluation functions
2010-11-21T13:09:16  *** mega1 <mega1!~quassel@2001:470:1f05:548:215:58ff:fe7d:773b> has joined #aichallenge
2010-11-21T13:18:33  <contestbot> Top 10 players: bocsimacko(3341), dmj111(3247), oldman(3148), _iouri_(3138), bix0r4ever(3135), Manwe56(3126), Raschi(3112), davidjliu(3069), protocolocon(3067), wagstaff(3065)
2010-11-21T13:24:48  <j3camero> amstan: my computer is too shitty to run minecraft.
2010-11-21T13:24:59  <amstan> j3camero: oh man... come on..
2010-11-21T13:25:11  <amstan> don't you have a desktop or something?
2010-11-21T13:29:48  <lavalamp_> antimatroid: it certainly helps if you don't accidentally give a reward for losing production
2010-11-21T13:30:55  *** amstan <amstan!~alex@76-10-183-61.dsl.teksavvy.com> has quit IRC (Ping timeout: 245 seconds)
2010-11-21T13:31:03  <antimatroid> haha, that is helpful
2010-11-21T13:31:25  <burny> a 1on1 rts, where there's a few unit types, and randomish starting positions could be kinda cool
2010-11-21T13:31:35  <antimatroid> the worst is when a bug fix makes your bot worse
2010-11-21T13:31:39  <burny> and fog of war
2010-11-21T13:32:22  <antimatroid> dmj is holding onto second for the moment
2010-11-21T13:34:05  <lavalamp_> yeah my current submission has a bug that counts defendeders twice... and it's my best so far
2010-11-21T13:37:05  <lavalamp_> grr damn assertions
2010-11-21T13:38:04  <antimatroid> i don't have any error checks for my current bot
2010-11-21T13:38:37  <antimatroid> i will fix it for sending invalid orders at the end, but it shouldn't be needed
2010-11-21T13:38:59  *** phreeza <phreeza!~phreeza@68-076.eduroam.RWTH-Aachen.DE> has quit IRC (Quit: Computer has gone to sleep)
2010-11-21T13:42:03  <lavalamp_> I have assertions if an impossible order is attempted
2010-11-21T13:42:38  *** Utkarsh <Utkarsh!~admin@117.201.81.139> has joined #aichallenge
2010-11-21T13:45:55  <antimatroid> bah, i hate errors that make no sense
2010-11-21T13:46:24  <lavalamp_> are you using c++?
2010-11-21T13:46:26  <antimatroid> i'm trying to separate a function up so i can do it individually for each planet, and it crashes even thought i copied the code :\
2010-11-21T13:46:28  <antimatroid> yep
2010-11-21T13:46:34  <antimatroid> you?
2010-11-21T13:47:06  <antimatroid> i might give up for tonight
2010-11-21T13:47:09  <antimatroid> it's almost 6am
2010-11-21T13:47:41  <lavalamp_> golang.org
2010-11-21T13:48:49  <lavalamp_> errors that make no sense ----> language must be C++ ^^
2010-11-21T13:49:04  <lavalamp_> (my day job is C++ so I'm allowed to say that)
2010-11-21T13:50:50  <burny> lol lava
2010-11-21T13:58:26  *** amriedle <amriedle!~amriedle@CPE00d009e408a1-CM001225401352.cpe.net.cable.rogers.com> has joined #aichallenge
2010-11-21T14:01:18  *** phreeza <phreeza!~phreeza@ip233.hermann.RWTH-Aachen.DE> has joined #aichallenge
2010-11-21T14:10:21  *** boegel <boegel!~boegel@91.182.156.200> has joined #aichallenge
2010-11-21T14:11:17  *** Rubicon-|-Cross <Rubicon-|-Cross!~RubiconCr@adsl-99-30-224-80.dsl.pltn13.sbcglobal.net> has joined #aichallenge
2010-11-21T14:11:20  <dmj111> antimatroid: thanks!  we'll see though, it hasn't played many games.
2010-11-21T14:11:23  *** phreeza <phreeza!~phreeza@ip233.hermann.RWTH-Aachen.DE> has quit IRC (Ping timeout: 265 seconds)
2010-11-21T14:11:25  <boegel> LAST WEEK!!!
2010-11-21T14:11:26  <boegel> w00!
2010-11-21T14:12:09  <dmj111> j3camero: have you looked at the facebook puzzles?  one is a persistent dinosaur island type thing, where you bot your breed to try to stay alive.
2010-11-21T14:12:11  <mega1> dmj111: that's quite an improvement. You are making me nervous :-)
2010-11-21T14:12:48  <dmj111> http://www.facebook.com/careers/puzzles.php?puzzle_id=19
2010-11-21T14:12:50  *** phreeza <phreeza!~phreeza@68-076.eduroam.RWTH-Aachen.DE> has joined #aichallenge
2010-11-21T14:12:55  <antimatroid> mega1: stop trying to flatter people :P
2010-11-21T14:13:22  <antimatroid> your bot has no worries atm :P
2010-11-21T14:13:36  <dmj111> mega1: that is kind of you.  but I think you will be ok :)
2010-11-21T14:13:37  <mega1> that's a tactic. It is though a sizeable jump.
2010-11-21T14:13:59  <mega1> I'm still trying to get alpha beta working.
2010-11-21T14:14:14  <mega1> it beat my current bot, it only fails against the rest.
2010-11-21T14:14:32  <dmj111> i tried, and gave up, but fixed a pretty bad bug yesterday that probably was part of the issue.
2010-11-21T14:14:41  <mega1> maybe if you all could run my old bot?
2010-11-21T14:14:44  <dmj111> (back when I tried alpha beta)
2010-11-21T14:14:54  <dmj111> ha :)
2010-11-21T14:15:01  <Error323> dmj111: is ur bot doing better?
2010-11-21T14:15:11  <antimatroid> Error323: it's second
2010-11-21T14:15:16  <Error323> oh nice
2010-11-21T14:15:16  <Error323> !
2010-11-21T14:15:18  <Error323> well
2010-11-21T14:15:22  <Error323> depending on how long :)
2010-11-21T14:15:25  <Error323> lemme check
2010-11-21T14:15:25  <dmj111> Error323: for now.  it has only a few games though.  yours was the one beating it the most on tcp i think
2010-11-21T14:15:30  <mega1> has few games, but judging by its tcp performance it's for real
2010-11-21T14:15:53  <antimatroid> mega1: i'm not sure your bot playing my current version would give you much information
2010-11-21T14:15:57  <Error323> dmj111: cool, gratz, is it alphabeta?
2010-11-21T14:16:09  <antimatroid> i'm not sure i trust alpha beta
2010-11-21T14:16:12  <dmj111> no, a hacked up 2 ply search.
2010-11-21T14:16:22  <Error323> cuz i've got fully functional code of alphabeta lying here
2010-11-21T14:16:30  <Error323> which is a bit frustrating :P
2010-11-21T14:16:43  <antimatroid> alphha beta should be pretty easy to implement?
2010-11-21T14:16:49  <mega1> yes, it is
2010-11-21T14:16:57  <antimatroid> i shall try it once i'm more polished in a day or two and see if it does better or worse
2010-11-21T14:16:57  <mega1> but to make it work well is another thing
2010-11-21T14:17:06  <Error323> well, i had a bit of a problem with it
2010-11-21T14:17:18  <Error323> the cutoff gave me nightmares
2010-11-21T14:17:24  <Error323> inconsistencies
2010-11-21T14:17:27  <Error323> but its fixed
2010-11-21T14:17:30  <antimatroid> what do you mean?
2010-11-21T14:17:33  <Error323> however working like crap
2010-11-21T14:18:19  <Error323> well because of the cutoff i had some recursive issues with state restoring
2010-11-21T14:18:26  <Error323> and action restoring
2010-11-21T14:18:34  <contestbot> Top 10 players: bocsimacko(3316), bix0r4ever(3300), dmj111(3228), _iouri_(3112), oldman(3105), Manwe56(3091), davidjliu(3084), Raschi(3077), protocolocon(3061), wagstaff(3058)
2010-11-21T14:19:23  <antimatroid> i take that back, he's third :P
2010-11-21T14:19:35  <Error323> yeah
2010-11-21T14:19:39  <antimatroid> bix0r only just submitted though
2010-11-21T14:19:47  <Error323> top 10 is as stable as to stacked marbles
2010-11-21T14:19:52  <Error323> except for one constant :P
2010-11-21T14:20:06  <Error323> of whom we won't speak
2010-11-21T14:20:07  <Error323> :P
2010-11-21T14:20:13  <antimatroid> yep
2010-11-21T14:20:24  <antimatroid> i said before, i think it's a real toss up for 2-30 atm
2010-11-21T14:20:30  <Error323> ye
2010-11-21T14:20:37  <dmj111> he who must not be named.
2010-11-21T14:20:38  <Error323> i hope i'll visit it aswell again :P
2010-11-21T14:20:43  <Error323> dmj111: exactly ;)
2010-11-21T14:21:00  <antimatroid> there's a sub contest, for second place
2010-11-21T14:21:25  <mega1> I'm pretty sure that bot will be beaten by several in the last week.
2010-11-21T14:21:28  *** p4p4_ <p4p4_!~chatzilla@82.113.106.28> has joined #aichallenge
2010-11-21T14:21:58  <mega1> the pace has picked up visibly
2010-11-21T14:22:35  <Error323> true
2010-11-21T14:22:53  <antimatroid> but you also have a week
2010-11-21T14:23:00  *** p4p4 <p4p4!~chatzilla@82.113.106.28> has quit IRC (Ping timeout: 245 seconds)
2010-11-21T14:23:01  <antimatroid> but yeah, it's certainly got a lot more competitive of late
2010-11-21T14:25:35  <antimatroid> i miss country/organisation rankings
2010-11-21T14:25:45  <dmj111> so, I have had both felixcoto and davidjliu crash in the middle games against me in the last two days.  anyone else seeing that much?
2010-11-21T14:26:01  <antimatroid> haven't really been playing people
2010-11-21T14:26:03  <dmj111> and davidjliu was winning at the time.
2010-11-21T14:26:23  <mega1> really? I'm not check my wins.
2010-11-21T14:26:29  <mega1> *checking
2010-11-21T14:26:54  <Error323> mmm haven't faced them yet
2010-11-21T14:26:59  <Error323> with new submission
2010-11-21T14:27:16  <dmj111> http://ai-contest.com/visualizer.php?game_id=8209081 that was felix
2010-11-21T14:27:47  <Error323> hmm
2010-11-21T14:27:55  <Error323> he's probably doing cpu intensive stuff
2010-11-21T14:28:04  <dmj111> i had one crash before because the opp. took a neutral down to 0 planets, and I divided by it :/
2010-11-21T14:28:22  <Error323> oh yeah
2010-11-21T14:28:25  <dmj111> ah, maybe a time out, I see.
2010-11-21T14:28:26  <antimatroid> yeah, i still need to sort my tree out to stop itself form doing intensive stuff past where i call generation off
2010-11-21T14:28:48  <antimatroid> i think i got most of it before
2010-11-21T14:29:51  <Error323> and there are also some that haven't resubmitted after mapchange
2010-11-21T14:29:53  <davidjliu> oh hello
2010-11-21T14:29:57  <Error323> they will be in for a surprise i think
2010-11-21T14:30:13  <mega1> look out for george
2010-11-21T14:30:28  <mega1> he's 11th with an oct 30 submission
2010-11-21T14:30:33  <antimatroid> mega1: i think there's a lot of people to look out for
2010-11-21T14:30:35  <antimatroid> shangas too
2010-11-21T14:31:03  <Error323> ye, and manwe, accoun, flanders
2010-11-21T14:31:11  <Error323> they are all still better than me >_<
2010-11-21T14:31:29  *** perror <perror!~fleury@aldebaran.labri.fr> has quit IRC (Quit: Bye all !)
2010-11-21T14:34:27  <dmj111> hey davidjliu, I was trying to figure out that game with the timeout, but see you have resubmitted :)
2010-11-21T14:34:54  <davidjliu> it was probably just me being slow
2010-11-21T14:35:09  <dmj111> ahh.  python bots unite!
2010-11-21T14:35:12  <davidjliu> my new bot shouldn't time out as much
2010-11-21T14:35:40  <Error323> it shouldn't timeout at all :P
2010-11-21T14:36:54  <antimatroid> i managed to secure that the leaderboard will be live during the contest too
2010-11-21T14:36:55  *** Naktibalda <Naktibalda!~quassel@cpc5-heme9-2-0-cust144.9-1.cable.virginmedia.com> has quit IRC (Ping timeout: 272 seconds)
2010-11-21T14:37:04  <davidjliu> Apparently the timeisup exception in the unofficial starter pack doesn't work, and I actually don't know anything about python, so I don't know about other time packages
2010-11-21T14:37:04  <antimatroid> i still can't believe that's even a question
2010-11-21T14:38:38  <dmj111> I just put a check in my code to update the games state.  it is called a bunch, so I am hoping that works.
2010-11-21T14:39:09  <Error323> bixor4ever has also resubmitted
2010-11-21T14:39:22  <Error323> damn, theres a lot of action :P
2010-11-21T14:40:04  *** bhasker <bhasker!~bhasker@69-170-40-2.static-ip.telepacific.net> has joined #aichallenge
2010-11-21T14:40:09  <Error323> hey bhasker
2010-11-21T14:40:20  <bhasker> hey
2010-11-21T14:40:31  <Error323> got some more new cool stuff? =)
2010-11-21T14:40:52  *** fmeyer <fmeyer!~fmeyer@189.100.228.254> has joined #aichallenge
2010-11-21T14:41:04  <Error323> whois flanders btw
2010-11-21T14:41:09  <bhasker> now idea came out of the blue
2010-11-21T14:41:12  <bhasker> no* idea
2010-11-21T14:41:19  <bhasker> as for new ideas, nope
2010-11-21T14:41:29  <Error323> it scares me when epic bots come *out of the blue*
2010-11-21T14:41:42  * boegel is currently evolving a tuned version of his bot, and hopes to get into the top 100 that way
2010-11-21T14:42:02  <Error323> :)
2010-11-21T14:42:10  <Error323> and once ur in ur hoping to be in the top 50
2010-11-21T14:42:12  <Error323> top 20
2010-11-21T14:42:13  <Error323> top 10
2010-11-21T14:42:14  <Error323> top 3
2010-11-21T14:42:16  <Error323> 1
2010-11-21T14:42:18  <Error323> ^_^
2010-11-21T14:42:32  <Error323> top 10 will be most fascinating
2010-11-21T14:42:43  <Error323> i don't think i'll make it :/
2010-11-21T14:42:55  <Error323> perhaps top 20
2010-11-21T14:44:23  <medrimonia> With the numbers of good new bots incoming I already hope to be in the top 50^^
2010-11-21T14:45:13  <Error323> well my bot appears to be doing well against high ranked players on tcp
2010-11-21T14:45:32  <Error323> though still losing from lowerranked also >_<
2010-11-21T14:45:36  <antimatroid> yeah, i always wanted top 10
2010-11-21T14:45:47  <antimatroid> but now i'll just be happy for top 50
2010-11-21T14:45:53  <medrimonia> mine too, but I had already lost two games
2010-11-21T14:46:35  <Error323> it beats oldman, dmj, wagstaff, accoun and flanders on regular bases
2010-11-21T14:46:38  <Error323> basis
2010-11-21T14:47:33  <medrimonia> E323-test
2010-11-21T14:47:39  <medrimonia> is this your new bot?
2010-11-21T14:48:05  <Error323> no 18.0-4 and 18.0-5 are
2010-11-21T14:48:13  <Error323> the former being on official
2010-11-21T14:48:45  <Error323> 379 elo on tcp
2010-11-21T14:52:19  <antimatroid> who's crapbot?
2010-11-21T14:52:26  <Error323> good question
2010-11-21T14:52:38  <Error323> its name doesn't give it justice though
2010-11-21T14:52:41  <Error323> its quite good :)
2010-11-21T14:52:51  <antimatroid> i played him twice before
2010-11-21T14:52:53  <antimatroid> won one, lost one
2010-11-21T14:53:08  <antimatroid> i might go to bed now
2010-11-21T14:53:13  <Error323> i'm 26 at official atm
2010-11-21T14:53:14  <Error323> :D
2010-11-21T14:53:15  <antimatroid> i'll try to add the rest in tomorrow :|
2010-11-21T14:53:17  <antimatroid> nice
2010-11-21T14:53:23  <bhasker> heh mine is at 49 holding
2010-11-21T14:53:34  <Error323> i'm still climbing i think
2010-11-21T14:53:38  <Error323> (i hope) :P
2010-11-21T14:53:38  <bhasker> which tcp server are you guys playing on
2010-11-21T14:54:08  <bhasker> oh i went to 51:-\
2010-11-21T14:55:03  <antimatroid> i'm not playing any atm
2010-11-21T14:55:06  <bhasker> dmj has a new version
2010-11-21T14:55:09  <Mistmanov> The matchmaking on the official server seems to be a bit off
2010-11-21T14:55:10  <antimatroid> played 30 or so games a few hours ago
2010-11-21T14:55:17  <antimatroid> found lots of juicy bugs
2010-11-21T14:55:28  <Error323> Mistmanov: ur doing well also
2010-11-21T14:55:31  <Error323> which ain't good
2010-11-21T14:55:33  <Mistmanov> every few games you get paired with someone who is wildly outside of your rank (like.. I'm 32, and my last game was against 1000)
2010-11-21T14:55:39  * Error323 wants to be #1 of his country :P
2010-11-21T14:55:45  <bhasker> hehe all the best e323
2010-11-21T14:56:00  <Mistmanov> I think you're going to win this one error =)
2010-11-21T14:56:01  <antimatroid> Mistmanov: you can't just play people with the same rank as you
2010-11-21T14:56:10  <Error323> Mistmanov: put your bot on tcp for a night
2010-11-21T14:56:13  <Error323> how so?
2010-11-21T14:56:18  <Mistmanov> but they should at least hgave kinda similar ranking
2010-11-21T14:56:25  <Mistmanov> so top 100 only play top 100
2010-11-21T14:56:31  <bhasker> now who is swann?
2010-11-21T14:56:32  <Mistmanov> 30 vs 1000 is pointless for us both I think
2010-11-21T14:56:45  <Error323> Mistmanov: no that may cause some local optima
2010-11-21T14:56:54  <antimatroid> Mistmanov: bocsi lost to 600+ rank the other day
2010-11-21T14:56:55  <Error323> you have to perform some "kruisbestuiving"
2010-11-21T14:57:04  <antimatroid> http://ai-contest.com/visualizer.php?game_id=8224517
2010-11-21T14:57:04  <Mistmanov> hmm -_-
2010-11-21T14:57:38  <antimatroid> you can't just have a bot that does well against other top bots
2010-11-21T14:57:50  <Mistmanov> yeah ok I agree
2010-11-21T14:57:54  <Mistmanov> but still :p
2010-11-21T14:58:00  <antimatroid> but nothing :P
2010-11-21T14:58:02  <bhasker> who the heck is GA.bot?
2010-11-21T14:58:05  * Mistmanov doesn't have any more arguments
2010-11-21T14:58:07  <bhasker> so many versions on dhartmei
2010-11-21T14:58:32  <Error323> indeed
2010-11-21T14:58:41  <bhasker> wtf
2010-11-21T14:58:44  <bhasker> like 30+
2010-11-21T14:58:48  <lavalamp_> this competition has taught me to swear like a sailor
2010-11-21T14:58:59  <Error323> well i've been spamming tcp aswell with quite some versions :$
2010-11-21T14:59:10  <Mistmanov> I believe GA is a genetic thing
2010-11-21T14:59:18  <Mistmanov> so all the bots are various genomes etc
2010-11-21T14:59:23  <bhasker> aah
2010-11-21T14:59:27  <Error323> genetic algorithms
2010-11-21T14:59:34  <Error323> yes, well its failing hard :P
2010-11-21T14:59:39  <bhasker> maybe the space invaders bot
2010-11-21T14:59:54  <Mistmanov> I personally find people who run multiple bots under the same name more annoying -_-
2010-11-21T15:00:09  <Mistmanov> yesterday, I had something like 20 games against only 2 competitors
2010-11-21T15:00:10  <Mistmanov> :/
2010-11-21T15:00:25  <Error323> put your bot on tcp for a night
2010-11-21T15:00:34  <Mistmanov> you mean like overnight?
2010-11-21T15:00:36  <Error323> because this is exactly the reason why you need a lot of daa
2010-11-21T15:00:41  <Error323> data*
2010-11-21T15:00:45  <Error323> i.e. games
2010-11-21T15:00:52  <Error323> yes
2010-11-21T15:00:52  <bhasker> looks like i went up against the GA bot now
2010-11-21T15:01:19  <Mistmanov> I like to give my PC some rest at night :p
2010-11-21T15:01:25  <Mistmanov> poor thing is already running the entire day ;)
2010-11-21T15:01:34  <bhasker> computers don't need no rest
2010-11-21T15:01:39  <Mistmanov> mine does -_-
2010-11-21T15:01:43  <bhasker> i just leave my laptop on running games all night
2010-11-21T15:01:52  <Error323> my pc is always on
2010-11-21T15:02:01  <bhasker> i can't keep my desktop on, cause its in the bedroom
2010-11-21T15:02:05  <bhasker> and the fan noise is annoying
2010-11-21T15:02:10  <Mistmanov> besides, any idea how much money that costs :p
2010-11-21T15:02:11  <Error323> mine is passive
2010-11-21T15:02:15  <Mistmanov> 11 cents per kwh or something
2010-11-21T15:02:19  <Mistmanov> so that's like.. a euro!
2010-11-21T15:02:23  <bhasker> mey a computer barely uses up any electricity
2010-11-21T15:02:31  <bhasker> especially if your monitor is not on
2010-11-21T15:02:44  <bhasker> the fan/ac in your house probably consumes a ton more
2010-11-21T15:02:58  *** fmeyer <fmeyer!~fmeyer@189.100.228.254> has quit IRC (Quit: leaving)
2010-11-21T15:03:08  <antimatroid> my computer is ALWAYS on and in my room
2010-11-21T15:03:08  <Error323> i assembled mine to be very silent and powerefficient
2010-11-21T15:03:28  <Mistmanov> aha, another argument: if I leave my pc on my AC will need to work harder to keep the house cool..
2010-11-21T15:03:36  <Mistmanov> (not that I have an ac, buteh..)
2010-11-21T15:03:44  <Error323> its winter :P
2010-11-21T15:03:55  <Mistmanov> I can keep this up all day (Y)
2010-11-21T15:03:57  <medrimonia> My laptop is running games during all night generaly
2010-11-21T15:04:24  <medrimonia> And even if my desktop is in the room, I keep it on sometimes
2010-11-21T15:04:26  <bhasker> mistmanov we have a game on tcp
2010-11-21T15:04:26  <bhasker> going
2010-11-21T15:04:37  <Mistmanov> are you malazanbot?
2010-11-21T15:04:40  <bhasker> yes
2010-11-21T15:04:44  <Mistmanov> ah k =)
2010-11-21T15:04:47  <bhasker> this is my latest hacked up scoring version
2010-11-21T15:04:51  <bhasker> it scores order combinations
2010-11-21T15:04:59  <bhasker> and executes the combination that seems to be the bets
2010-11-21T15:05:01  <bhasker> best*
2010-11-21T15:05:09  <bhasker> rather than executing the highest value order first
2010-11-21T15:05:15  <Error323> oh thats what mega1 does
2010-11-21T15:05:16  <bhasker> so going to see how it does
2010-11-21T15:05:28  <bhasker> sadly my heuristics to calculate aren't as good
2010-11-21T15:05:29  <Mistmanov> my bot is currently logging everything to file and not showing it in the console, so I can't see "live" how the game is going
2010-11-21T15:05:32  <bhasker> nor is my eval function
2010-11-21T15:05:33  <Error323> but defining "best" is hard :D
2010-11-21T15:05:36  <bhasker> oh i just tail the log:p
2010-11-21T15:05:36  <lavalamp_> I do that too ^^
2010-11-21T15:05:45  <bhasker> rather grep it
2010-11-21T15:05:52  <bhasker> looks like you will win
2010-11-21T15:05:58  <Error323> me too
2010-11-21T15:05:58  <bhasker> my bot blew it midway
2010-11-21T15:06:00  <bhasker> it was winning
2010-11-21T15:06:01  <Error323> tail -f
2010-11-21T15:06:05  <Error323> =)
2010-11-21T15:06:16  <bhasker> 2010-11-21 12:05:13,960-v0.1-INFO- Turn:108            updateStateObj() - My Value of current state:137.206162
2010-11-21T15:06:16  <bhasker> 2010-11-21 12:05:13,960-v0.1-INFO- Turn:108            updateStateObj() - Enemy Value of current state:63.064370
2010-11-21T15:06:16  <bhasker> 2010-11-21 12:05:14,630-v0.1-INFO- Turn:109            updateStateObj() - My Value of current state:72.613095
2010-11-21T15:06:16  <bhasker> 2010-11-21 12:05:14,630-v0.1-INFO- Turn:109            updateStateObj() - Enemy Value of current state:119.114839
2010-11-21T15:06:17  <bhasker> 2
2010-11-21T15:06:19  <bhasker> turn 108
2010-11-21T15:06:20  <Mistmanov> yeah well, I'm an utter noob at cygwin etc, so I'm just happy I got something that works :p
2010-11-21T15:06:24  <bhasker> looks like it lost a critical planet
2010-11-21T15:06:25  <bhasker> ^
2010-11-21T15:06:27  <bhasker> :-\
2010-11-21T15:07:07  <Error323> value being?
2010-11-21T15:07:34  <Error323> i'm also gonna try one last thing
2010-11-21T15:07:36  <albert_> j3camero: hey, is it possible to install scipy on the server?
2010-11-21T15:07:40  <Error323> wrt neutral capping
2010-11-21T15:07:44  <bhasker> an approximate value of the current state in terms of ships/growth
2010-11-21T15:07:45  <bhasker> in the future
2010-11-21T15:07:51  <Error323> ah ok
2010-11-21T15:08:06  <bhasker> terminal conditions have a 99999/-99999 value
2010-11-21T15:08:17  <bhasker> like this
2010-11-21T15:08:17  <bhasker> 2010-11-21 12:06:07,189-v0.1-INFO- Turn:168            updateStateObj() - Enemy Value of current state:219.739051
2010-11-21T15:08:18  <bhasker> 2010-11-21 12:06:07,879-v0.1-INFO- Turn:169            updateStateObj() - My Value of current state:-999999.000000
2010-11-21T15:08:18  <bhasker> 2010-11-21 12:06:07,879-v0.1-INFO- Turn:169            updateStateObj() - Enemy Value of current state:999999.000000
2010-11-21T15:08:22  <bhasker> :(
2010-11-21T15:08:43  <Error323> ye that's not looking good :P
2010-11-21T15:09:04  <lavalamp_> I say it as a pair: Eval: (0.23625449172704346, 133.37697429050039)
2010-11-21T15:09:19  <bhasker> yours and enemies?
2010-11-21T15:09:23  <lavalamp_> yeah
2010-11-21T15:09:29  <lavalamp_> also not good
2010-11-21T15:09:34  <bhasker> hehe
2010-11-21T15:09:40  *** p4p4_ <p4p4_!~chatzilla@82.113.106.28> has quit IRC (Ping timeout: 245 seconds)
2010-11-21T15:09:47  <bhasker> whats your eval function like
2010-11-21T15:09:57  <lavalamp_> hairy and subpar ^^
2010-11-21T15:10:05  <lavalamp_> I have a bunch of factors
2010-11-21T15:10:10  <bhasker> mine is a very simple one
2010-11-21T15:10:26  <bhasker> myTotalShips, enemyTotalShips, my growth,enemy growth weighted
2010-11-21T15:10:31  <lavalamp_> unfortified planets and snipable planets hurt, extra growth and ships help, etc
2010-11-21T15:10:34  <bhasker> at start it weights growth
2010-11-21T15:10:39  <bhasker> as game proceeds it weights ships more
2010-11-21T15:10:50  <bhasker> so towards the end it will prefer moves that keep ship counts higher
2010-11-21T15:10:54  <lavalamp_> I do the same
2010-11-21T15:11:00  <bhasker> at least i hope it does that
2010-11-21T15:11:27  <lavalamp_> if it gets near the end I give a huge bonus to having more ships at the last turn
2010-11-21T15:11:35  <bhasker> ah
2010-11-21T15:11:57  *** p4p4 <p4p4!~chatzilla@82.113.106.28> has joined #aichallenge
2010-11-21T15:12:50  <Error323> dmj111: did you happen to keep record how often i beated ur bot
2010-11-21T15:12:51  <Error323> ?
2010-11-21T15:15:14  <Error323> You WIN against Mistmanov
2010-11-21T15:15:15  <Error323> ^_^
2010-11-21T15:15:21  <Mistmanov> -_-
2010-11-21T15:15:42  *** Ice_Harley <Ice_Harley!5eb2ba62@gateway/web/freenode/ip.94.178.186.98> has joined #aichallenge
2010-11-21T15:15:51  *** bhasker <bhasker!~bhasker@69-170-40-2.static-ip.telepacific.net> has quit IRC (Quit: bhasker)
2010-11-21T15:16:20  *** fawek <fawek!~faw@chello089079043119.chello.pl> has quit IRC (Ping timeout: 245 seconds)
2010-11-21T15:18:36  <contestbot> Top 10 players: bocsimacko(3332), dmj111(3268), bix0r4ever(3260), oldman(3123), _iouri_(3100), Manwe56(3091), Raschi(3085), krokkrok(3058), george(3053), protocolocon(3052)
2010-11-21T15:22:28  *** tapwater <tapwater!~tapwater@pool-72-64-104-180.dllstx.fios.verizon.net> has joined #aichallenge
2010-11-21T15:25:47  <lavalamp_> aaaaaa I'm making it worse :(
2010-11-21T15:26:08  <lavalamp_> Eval: (2.7141186436641073e-07, 1813.8742795780845)
2010-11-21T15:29:52  <Mistmanov> http://72.44.46.68/canvas?game_id=1129234
2010-11-21T15:30:06  <Mistmanov> hmmm.. took me 109 turns to defeat an enemy that doesn't do anything
2010-11-21T15:31:42  <boegel> Mistmanov: hey, that's my bot :)
2010-11-21T15:32:00  <Mistmanov> ah ok =)
2010-11-21T15:32:12  <Mistmanov> we were already wondering who made the GA bots
2010-11-21T15:32:28  <Mistmanov> still some programming/evolving left I guess? :p
2010-11-21T15:33:23  <boegel> Mistmanov: yes
2010-11-21T15:33:30  <boegel> Mistmanov: but I'm working on it :)
2010-11-21T15:33:46  <boegel> shit, shouldn't have revealed myself :P
2010-11-21T15:33:46  <Mistmanov> it's a genetic algorithm thing right?
2010-11-21T15:33:53  * boegel shuts up
2010-11-21T15:33:57  <boegel> Mistmanov: yeah :)
2010-11-21T15:34:13  <Mistmanov> are you using the tcp games to evolve your bots?
2010-11-21T15:34:19  <boegel> Mistmanov: I'm tired of trying to figure out good tuned parameters for my bots
2010-11-21T15:34:21  <Mistmanov> (I would think that doesn't go fast enough?)
2010-11-21T15:34:30  <boegel> Mistmanov: depends on how you do it ^_^
2010-11-21T15:34:48  <boegel> Mistmanov: I have some experience with GA's from my research
2010-11-21T15:34:56  <Mistmanov> ah ok
2010-11-21T15:35:05  <boegel> Mistmanov: and I believe it may at least help me to figure out how to finetune my current bot
2010-11-21T15:35:23  <boegel> I won't have time to implement additional stuff in it, so I figure I'll try it this way :)
2010-11-21T15:35:40  <Mistmanov> =) yeah I tried evolving some parameters by just playing locally
2010-11-21T15:35:53  <Mistmanov> but after 10 hours of 100% cpu load the bot was actually worse than it was before
2010-11-21T15:36:13  <Mistmanov> but then again, I've never in my life even seen a proper GA
2010-11-21T15:36:24  <Error323> i programmed it a few times
2010-11-21T15:36:28  <Error323> for various project
2010-11-21T15:36:30  <Error323> s
2010-11-21T15:37:38  <medrimonia> Error323: Beat you on official^^
2010-11-21T15:37:45  <Mistmanov> yeah I really start noticing my non-programming background now... all the stuff about eval functions and minimax and I go "eh...?"
2010-11-21T15:37:56  <Error323> medrimonia: yea ur screwing with my ranking damn you ! ;)
2010-11-21T15:37:56  <boegel> Mistmanov: having experience with a GA makes a hell of lot of difference :)
2010-11-21T15:38:14  <boegel> Mistmanov: also, it makes sense you win against that bot, it literally never sends anything
2010-11-21T15:38:26  <Mistmanov> yeah, that's why I posted the game :p
2010-11-21T15:38:28  <boegel> Mistmanov: the GA doesn't know what that'll result in, so it tries :)
2010-11-21T15:39:11  <Mistmanov> well, there was this post on the forum of a bot that only fled whenever it's single planet was being attacked... it eventually won after ~100 turns because the other guy crashed
2010-11-21T15:39:16  <Mistmanov> so it's worth a try :p
2010-11-21T15:39:36  <Error323> I hope you'll kick ass boegel
2010-11-21T15:39:43  <Error323> I like it when ppl try new stuff
2010-11-21T15:39:48  <boegel> Error323: thx :)
2010-11-21T15:40:02  <boegel> Error323: my best bot (running on official now) is decent
2010-11-21T15:40:11  <Error323> also ga?
2010-11-21T15:40:22  *** samulihs <samulihs!~samuli@cs181069183.pp.htv.fi> has quit IRC (Ping timeout: 252 seconds)
2010-11-21T15:40:23  <boegel> Error323: and I fully parameterized it, allowing me to finetune it with a GA
2010-11-21T15:40:28  <boegel> Error323: no, all heuristics
2010-11-21T15:40:51  <Error323> ah ok
2010-11-21T15:40:58  <boegel> Error323: I'm trying to finetune different thresholds, and enable/disable certain substrategies
2010-11-21T15:41:21  <Error323> i now understand why the tcp server is spammed with ga bots
2010-11-21T15:41:39  <boegel> hehe :D
2010-11-21T15:41:42  <boegel> sorry about that
2010-11-21T15:41:45  <Error323> np
2010-11-21T15:41:48  <boegel> I'm working on something now
2010-11-21T15:41:52  <Error323> just make it awesome :P
2010-11-21T15:42:02  <boegel> which will hopefully result in way better GA bots
2010-11-21T15:42:04  <boegel> hopefully
2010-11-21T15:42:24  <Error323> i've had many ideas with hope in them :P
2010-11-21T15:42:33  <boegel> I'll let it run tonight, let's see what it comes up with :)
2010-11-21T15:42:33  <Error323> which resulted in the almighty trashcan
2010-11-21T15:42:43  <boegel> same here
2010-11-21T15:42:47  <boegel> but this is fun to do
2010-11-21T15:42:56  <Error323> yep
2010-11-21T15:43:01  <boegel> so whether it'll work out or not, it's worthwhile
2010-11-21T15:44:32  <Error323> if it finds out when to cap what neutral let me know :P
2010-11-21T15:45:13  <dmj111> Error323: I don't have a record, but I have the output from the games. I can let you know when I get a few minutes to grep...
2010-11-21T15:45:35  <Error323> dmj111: cool, if you would I'd be curious
2010-11-21T15:45:40  <boegel> Error323: it's not that kind of GA :)
2010-11-21T15:45:49  <boegel> Error323: it's not using in GA during playtime
2010-11-21T15:45:59  <boegel> s/in GA/a GA/
2010-11-21T15:46:34  <Error323> I've no idea how you modelled it
2010-11-21T15:46:35  <Error323> hehe
2010-11-21T15:46:44  <dmj111> Error323: ok, that was easier than I thought, and more disconcerting :/
2010-11-21T15:47:00  <dmj111> I lost 2 of 2 to 18.0.6, and
2010-11-21T15:47:09  <dmj111> won 3 of 3 against 18.0.4
2010-11-21T15:47:19  <Error323> hmmm
2010-11-21T15:47:29  <dmj111> didn't realize we played so few on that version.
2010-11-21T15:47:31  <Error323> that's hardly significant unfortunatly
2010-11-21T15:47:48  <Error323> why are you all working in python
2010-11-21T15:47:56  <Error323> i can't exchange binaries with python guys :/
2010-11-21T15:48:03  <dmj111> ha.  that is a downside.
2010-11-21T15:48:46  <Error323> i can send you mine but i'm not sure yet :P
2010-11-21T15:48:52  <Error323> dmj111: got linux?
2010-11-21T15:49:15  <dmj111> got linux.
2010-11-21T15:50:03  <dmj111> back later...
2010-11-21T15:50:03  <Error323> interested in my binary?
2010-11-21T15:50:06  <Error323> ah ok :P
2010-11-21T15:50:34  <boegel> are players exchanging binary version of their bots?
2010-11-21T15:50:39  <dmj111> possibly, but not sure I will have time to use it :/  I'll talk to you later if you are around.
2010-11-21T15:50:48  <Error323> dmj111: rgr
2010-11-21T15:50:48  <boegel> I'd love to get my hands on some, to finetune my bot against :)
2010-11-21T15:50:59  <Error323> oh i'll send you mine boegel
2010-11-21T15:51:09  <boegel> Error323: nah, nvm :
2010-11-21T15:51:25  <Error323> heh ok :P
2010-11-21T15:51:29  <boegel> Error323: I'll try and get some decent results out of what I have now, let's start there
2010-11-21T15:53:09  <Ice_Harley> boegel
2010-11-21T15:53:25  <Ice_Harley> what is your bot current ranking?
2010-11-21T15:53:44  *** fawek <fawek!~faw@chello089079043119.chello.pl> has joined #aichallenge
2010-11-21T15:53:54  <Ice_Harley> i can give you my binary... if your binary will be useful for me also )
2010-11-21T15:54:15  <Ice_Harley> mine is 30
2010-11-21T15:55:09  *** narnach <narnach!~Narnach@195-241-27-75.ip.telfort.nl> has quit IRC (Read error: Operation timed out)
2010-11-21T15:56:05  *** medrimonia1 <medrimonia1!~Medrimoni@ABordeaux-158-1-56-107.w90-16.abo.wanadoo.fr> has joined #aichallenge
2010-11-21T15:56:24  <boegel> Ice_Harley: oh, my best bot is ~200 or so
2010-11-21T15:56:53  *** medrimonia <medrimonia!~Medrimoni@ABordeaux-158-1-116-7.w90-60.abo.wanadoo.fr> has quit IRC (Ping timeout: 240 seconds)
2010-11-21T15:57:10  *** narnach_mb <narnach_mb!~Narnach@195-241-27-75.ip.telfort.nl> has quit IRC (Ping timeout: 245 seconds)
2010-11-21T15:57:31  <Ice_Harley> mm, ok. I can give you my old version
2010-11-21T15:57:38  *** Naktibalda <Naktibalda!~quassel@cpc5-heme9-2-0-cust144.9-1.cable.virginmedia.com> has joined #aichallenge
2010-11-21T15:58:20  <boegel> Ice_Harley: not for now, I can't put it to good use (yet)
2010-11-21T15:58:26  <boegel> I might in the near future though :)
2010-11-21T16:02:23  *** narnach <narnach!~Narnach@195-241-27-75.ip.telfort.nl> has joined #aichallenge
2010-11-21T16:02:23  *** narnach is now known as narnach_mb
2010-11-21T16:02:29  *** narnach <narnach!~Narnach@195-241-27-75.ip.telfort.nl> has joined #aichallenge
2010-11-21T16:03:33  <Ice_Harley> Error323, what about you?
2010-11-21T16:04:03  *** McLeopold <McLeopold!~scott@c-98-247-248-39.hsd1.wa.comcast.net> has joined #aichallenge
2010-11-21T16:04:35  <boegel> Ice_Harley: how are you using the binaries? so kind of tuning proces? or just testing?
2010-11-21T16:05:36  <Ice_Harley> I am now on the level when I can hardly say if my changes make my bot stronger or not. So I make series of tests against different bots
2010-11-21T16:05:46  <Ice_Harley> tuning, yes )
2010-11-21T16:07:14  <Error323> hmm?
2010-11-21T16:07:37  <Error323> Ice_Harley: you wanna trade binaries?
2010-11-21T16:07:51  <Ice_Harley> yes
2010-11-21T16:08:36  <Error323> Ice_Harley: got linux?
2010-11-21T16:08:43  <Ice_Harley> no
2010-11-21T16:08:54  <Error323> :(
2010-11-21T16:08:54  <boegel> :)
2010-11-21T16:09:00  <Error323> C#
2010-11-21T16:09:02  <Error323> yuck :P
2010-11-21T16:09:06  <boegel> haha
2010-11-21T16:09:09  <Ice_Harley> C# )
2010-11-21T16:09:19  <Ice_Harley> am i only windows user here?
2010-11-21T16:09:44  <Mistmanov> I originally wanted to code in windows..
2010-11-21T16:09:57  <Ice_Harley> but?
2010-11-21T16:09:58  <boegel> coding in Windows is bleh
2010-11-21T16:10:02  <Mistmanov> but I ended up using cygwin since windows didn't work :p
2010-11-21T16:10:04  <boegel> so is coding in C#
2010-11-21T16:10:41  <Mistmanov> it seriously took something like 3 hours of changing stuff before I had the first example bot running..
2010-11-21T16:10:56  <Mistmanov> <<<--- not used to programming :(
2010-11-21T16:11:54  <Ice_Harley> by the way, thanks for your guide, Mistmanov
2010-11-21T16:11:58  *** mceier <mceier!~mceier@chello089077208118.chello.pl> has quit IRC (Quit: leaving)
2010-11-21T16:12:26  <Mistmanov> I wondered whether anyone found them useful..
2010-11-21T16:12:27  <Error323> i should read it
2010-11-21T16:12:34  <Error323> i'll do that tomorrow
2010-11-21T16:12:36  <Ice_Harley> when i read it, i understand that i am on right way ) It was very similar to my own thoughts
2010-11-21T16:12:38  <Mistmanov> it's really basic stuff I think
2010-11-21T16:13:11  <Ice_Harley> mm, my bot is still basic, i think )
2010-11-21T16:13:20  <Mistmanov> probably not really useful for people who've already made a top ~50 bot
2010-11-21T16:13:32  <Mistmanov> make that certainly not really useful :p
2010-11-21T16:13:35  <boegel> which guide?
2010-11-21T16:14:02  <Mistmanov> don't know actually :p
2010-11-21T16:14:10  <boegel> heh
2010-11-21T16:14:10  <Mistmanov> I wrote two threads in the strategy forum
2010-11-21T16:14:16  <boegel> ah, k
2010-11-21T16:14:19  <Mistmanov> dunno which one harley read
2010-11-21T16:14:20  <Ice_Harley> first one
2010-11-21T16:14:28  * boegel doesn't have time to look into that in depth
2010-11-21T16:14:57  <Ice_Harley> Complete overview of my script
2010-11-21T16:15:06  <Ice_Harley> your script :)
2010-11-21T16:16:32  <Mistmanov> currently stuck on my algorithm for determining which planets are the "front"
2010-11-21T16:16:39  <Mistmanov> (still not really sure if it's helpful btw)
2010-11-21T16:17:05  <Mistmanov> front at a specific point in time is easy, but for reinforcements you want to know the front when reinforcements arrive
2010-11-21T16:17:22  <Mistmanov> but the current front is also still relevant.. and then things start getting weird
2010-11-21T16:17:37  <Ice_Harley> if it's interesting for you i can tell my way
2010-11-21T16:17:46  <Mistmanov> please do =)
2010-11-21T16:18:08  <Mistmanov> maybe I missed something obvious.
2010-11-21T16:18:22  <Error323> Mistmanov: are you compiling linux binaries or *.exe?
2010-11-21T16:18:34  <Error323> i don't know how that works with cygwin
2010-11-21T16:18:35  <contestbot> Top 10 players: bocsimacko(3320), bix0r4ever(3257), dmj111(3239), oldman(3128), GreenTea(3118), Manwe56(3109), Raschi(3095), _iouri_(3088), netwalker2000(3076), davidjliu(3057)
2010-11-21T16:18:38  <Ice_Harley> i am using 4 sectors: NE, SE, NW, SW. Closest my planets in each sectors for each enemy planets are front
2010-11-21T16:18:39  <Mistmanov> ,exe
2010-11-21T16:19:07  <Error323> mkay
2010-11-21T16:19:08  *** medrimonia <medrimonia!~medrimoni@ABordeaux-158-1-56-107.w90-16.abo.wanadoo.fr> has joined #aichallenge
2010-11-21T16:19:20  <Mistmanov> hm
2010-11-21T16:19:24  <Error323> goddamnit lost again @ official>_<
2010-11-21T16:20:42  <Mistmanov> hah! I regained the dutch crown
2010-11-21T16:20:52  * Ice_Harley defeated _iouri_ and very proud
2010-11-21T16:21:02  <Error323> Mistmanov: grrr
2010-11-21T16:21:21  <Mistmanov> did you just make some bad adjustment that caused you to lose the top10 position? or is this it's naturally position?
2010-11-21T16:21:43  <Mistmanov> natural -_-
2010-11-21T16:22:22  *** delt0r <delt0r!~delt0r@188-22-173-47.adsl.highway.telekom.at> has joined #aichallenge
2010-11-21T16:22:25  <Mistmanov> (for some reason I often think of the correct word in english, but then type a related word in correct spelling :/)
2010-11-21T16:22:34  <Error323> Mistmanov: no just not enough data
2010-11-21T16:22:58  <Ice_Harley> Mistmanov, can your binaries run under windows?
2010-11-21T16:23:06  <Mistmanov> I have no idea
2010-11-21T16:23:20  *** delt0r___ <delt0r___!~delt0r@178-191-73-252.adsl.highway.telekom.at> has quit IRC (Ping timeout: 265 seconds)
2010-11-21T16:23:25  <Mistmanov> as mentioned, getting the contest to run was a bit of a disaster over here
2010-11-21T16:24:06  <Mistmanov> I'd gladly give you my source, but I believe people are opposed to that -_-
2010-11-21T16:24:12  <Ice_Harley> no-no
2010-11-21T16:25:20  <Mistmanov> (oh, if I double click the .exe in windows it gives an error message saying that it needs cygwin.dll So I'd say that it doesn't work under windows =))
2010-11-21T16:25:20  <Ice_Harley> one week later )
2010-11-21T16:25:48  <Ice_Harley> cygwin is command line, isn't it?
2010-11-21T16:25:51  <Error323> giving your source is giving everyone who's got it a higher spot
2010-11-21T16:25:58  <Mistmanov> yeah I know
2010-11-21T16:26:16  <Ice_Harley> why it needs command line to work? )
2010-11-21T16:26:32  <Mistmanov> oh no cygwin is some kind of linux emulator
2010-11-21T16:26:42  <Mistmanov> not entirely sure what it does
2010-11-21T16:26:45  *** smloh1 <smloh1!4786e1ca@gateway/web/freenode/ip.71.134.225.202> has joined #aichallenge
2010-11-21T16:27:06  <Ice_Harley> i think i understand
2010-11-21T16:29:26  <Ice_Harley> hmm, ELO is surprising: i was 39. Lose to 41 and win on 7 - and i am 109 now!
2010-11-21T16:32:06  <Error323> how many games?
2010-11-21T16:32:25  <Error323> also note that ranks shown aren't always up to date
2010-11-21T16:32:31  <Error323> he may have dropped himself
2010-11-21T16:33:04  *** p4p4 <p4p4!~chatzilla@82.113.106.28> has quit IRC (Remote host closed the connection)
2010-11-21T16:33:57  <smloh1> so i'm testing two slightly different tweaks
2010-11-21T16:34:13  <smloh1> ver1 beats ver2 75% of the time
2010-11-21T16:34:44  <smloh1> but testing both against an older, more different version of my bot, ver2 actually performs slightly better
2010-11-21T16:34:58  <smloh1> i guess it's a good reminder about getting stuck in local maxima when tuning
2010-11-21T16:36:02  <Error323> playing against urself is dangerous
2010-11-21T16:36:14  <Error323> smloh1: got linux and using a compiling language? :P
2010-11-21T16:37:04  <smloh1> nope, windows but i do .jar
2010-11-21T16:38:14  <Error323> ah
2010-11-21T16:38:45  <Ice_Harley> Error323, only 2 games.  Now I win against 107 and come to 23 place ) You are right - it needs some time to become stable
2010-11-21T16:39:19  <Error323> it needs quite a lot of time
2010-11-21T16:39:29  <Error323> :(
2010-11-21T16:39:56  <Error323> imo its stable-ish after about 100 games
2010-11-21T16:40:02  <Ice_Harley> so i can go sleep :) Good night
2010-11-21T16:40:32  <davidjliu> does anybody have any idea what Rufes Band is?
2010-11-21T16:40:38  <Error323> nope
2010-11-21T16:40:48  <Error323> the spanish guys right?
2010-11-21T16:41:03  <davidjliu> yeah
2010-11-21T16:41:07  *** Ice_Harley <Ice_Harley!5eb2ba62@gateway/web/freenode/ip.94.178.186.98> has quit IRC (Quit: Page closed)
2010-11-21T16:41:07  <Mistmanov> google doesn't know em, so I'm guessing it's a nickname for a group of friends or something?
2010-11-21T16:43:07  <Error323> crap 5 losses in a row on official :/
2010-11-21T16:43:13  <Error323> that just hurts
2010-11-21T16:43:43  *** narnach <narnach!~Narnach@195-241-27-75.ip.telfort.nl> has quit IRC (Quit: Leaving.)
2010-11-21T16:44:03  *** narnach_mb <narnach_mb!~Narnach@195-241-27-75.ip.telfort.nl> has quit IRC (Quit: Leaving.)
2010-11-21T16:45:27  *** yasith__ <yasith__!~yasith@112.134.120.183> has quit IRC (Ping timeout: 245 seconds)
2010-11-21T16:45:49  *** bhasker <bhasker!~bhasker@adsl-99-25-118-235.dsl.pltn13.sbcglobal.net> has joined #aichallenge
2010-11-21T16:51:28  *** Palmik <Palmik!~Palmik@palmikpc.sks6.muni.cz> has quit IRC (Remote host closed the connection)
2010-11-21T16:55:23  *** sigh <sigh!~sigh@lithium.mailguard.com.au> has joined #aichallenge
2010-11-21T16:56:22  *** boegel <boegel!~boegel@91.182.156.200> has quit IRC (Quit: Leaving)
2010-11-21T16:58:51  *** yasith__ <yasith__!~yasith@112.134.125.49> has joined #aichallenge
2010-11-21T17:00:29  *** Utkarsh <Utkarsh!~admin@117.201.81.139> has quit IRC (Ping timeout: 276 seconds)
2010-11-21T17:06:15  *** Utkarsh <Utkarsh!~admin@117.201.81.92> has joined #aichallenge
2010-11-21T17:06:44  *** smloh1 <smloh1!4786e1ca@gateway/web/freenode/ip.71.134.225.202> has quit IRC (Quit: Page closed)
2010-11-21T17:06:52  *** Rubicon-|-Cross <Rubicon-|-Cross!~RubiconCr@adsl-99-30-224-80.dsl.pltn13.sbcglobal.net> has quit IRC (Quit: Leaving)
2010-11-21T17:08:40  *** McLeopold <McLeopold!~scott@c-98-247-248-39.hsd1.wa.comcast.net> has quit IRC (Quit: Leaving)
2010-11-21T17:10:27  *** Mistmanov <Mistmanov!4da38f1c@gateway/web/freenode/ip.77.163.143.28> has quit IRC (Quit: Page closed)
2010-11-21T17:15:33  *** albert_ <albert_!868204f3@gateway/web/freenode/ip.134.130.4.243> has quit IRC (Quit: Page closed)
2010-11-21T17:18:36  <contestbot> Top 10 players: bocsimacko(3305), GreenTea(3301), dmj111(3235), Manwe56(3111), oldman(3110), Raschi(3096), _iouri_(3077), davidjliu(3053), george(3051), protocolocon(3050)
2010-11-21T17:20:36  *** Azrathud <Azrathud!~Azrathud@c-67-166-43-227.hsd1.co.comcast.net> has joined #aichallenge
2010-11-21T17:27:01  *** mceier <mceier!~mceier@chello089077208118.chello.pl> has joined #aichallenge
2010-11-21T17:27:50  *** Utkarsh <Utkarsh!~admin@117.201.81.92> has quit IRC (Ping timeout: 250 seconds)
2010-11-21T17:34:19  *** Utkarsh <Utkarsh!~admin@117.201.80.223> has joined #aichallenge
2010-11-21T17:40:30  <lavalamp_> arghh just shoot me now
2010-11-21T17:43:05  *** Xrillo <Xrillo!~christian@h85-8-26-145.static.se.alltele.net> has left #aichallenge ("Ex-Chat")
2010-11-21T17:49:03  *** sorpha <sorpha!803d2384@gateway/web/freenode/ip.128.61.35.132> has joined #aichallenge
2010-11-21T17:51:59  *** sorpha <sorpha!803d2384@gateway/web/freenode/ip.128.61.35.132> has quit IRC (Client Quit)
2010-11-21T17:56:34  <lavalamp_> gahhhhhh bugssss
2010-11-21T17:58:10  <pgpaskar> :|
2010-11-21T18:10:49  <medrimonia> hmmm it seems to me that I've to check better the first turn
2010-11-21T18:10:51  <medrimonia> http://ai-contest.com/visualizer.php?game_id=8246957
2010-11-21T18:15:29  <mega1> medrimonia: looks fine tome
2010-11-21T18:15:50  <mega1> might be the second turn that's suspect
2010-11-21T18:16:45  <phreeza> ulope: i'm trying to rewrite my bot with your python starter kit, but i keep timing out after a few turns, and i've traced it back to the readline() before the new turn starts just giving an empty string. any ideas?
2010-11-21T18:17:54  <phreeza> seems like the pipe is broken but it doesn't detect it...
2010-11-21T18:18:37  <contestbot> Top 10 players: bocsimacko(3320), dmj111(3226), GreenTea(3194), oldman(3128), Manwe56(3111), Raschi(3099), bix0r4ever(3070), _iouri_(3070), protocolocon(3059), george(3052)
2010-11-21T18:19:48  <medrimonia> I don't know if I could have snipe him the second turn mega1
2010-11-21T18:20:56  <mega1> is taking the 81 on right a bad move on the 2nd turn?
2010-11-21T18:22:10  <medrimonia> it's all right too, but my scoring of expansion seems to have preferred the two others
2010-11-21T18:23:17  <medrimonia> Yeah, that's it, didn't snipe because my scoring was too bad for the interception
2010-11-21T18:23:29  <medrimonia> I've maybe to correct some things
2010-11-21T18:31:19  <Error323> wuut new bot is pwning so far
2010-11-21T18:31:22  <Error323> on tcp
2010-11-21T18:31:56  *** Cold-Phoenix <Cold-Phoenix!~Cold-Phoe@87.113.149.249> has joined #aichallenge
2010-11-21T18:33:29  *** Appleman1234 <Appleman1234!~Appleman1@CPE-60-226-176-19.qld.bigpond.net.au> has joined #aichallenge
2010-11-21T18:40:22  <j3camero> Error323 what's your username in the official rankings?
2010-11-21T18:40:40  <Error323> j3camero: Error323
2010-11-21T18:40:49  <j3camero> Seems easy enough.
2010-11-21T18:43:42  <burny> anyoen for some 1on1 or 2on2 glacon flash?
2010-11-21T18:53:44  <Error323> j3camero: but that's not the tcp version!
2010-11-21T19:00:50  <Error323> manwe is my nemesis >_<
2010-11-21T19:08:32  *** Cold-Phoenix <Cold-Phoenix!~Cold-Phoe@87.113.149.249> has quit IRC ()
2010-11-21T19:11:12  *** Utkarsh <Utkarsh!~admin@117.201.80.223> has quit IRC (Ping timeout: 260 seconds)
2010-11-21T19:14:06  <j3camero> I found this game called Boundless Planet. I am having quite a nice time with it so far.
2010-11-21T19:14:17  <j3camero> It's like Starcraft, except on a virtually unlimited map.
2010-11-21T19:14:23  *** fawek <fawek!~faw@chello089079043119.chello.pl> has quit IRC (Ping timeout: 264 seconds)
2010-11-21T19:15:14  *** amriedle <amriedle!~amriedle@CPE00d009e408a1-CM001225401352.cpe.net.cable.rogers.com> has quit IRC (Quit: leaving)
2010-11-21T19:15:58  *** Utkarsh <Utkarsh!~admin@117.201.82.162> has joined #aichallenge
2010-11-21T19:18:38  <contestbot> Top 10 players: bocsimacko(3328), GreenTea(3186), bix0r4ever(3175), dmj111(3169), Raschi(3116), Manwe56(3088), oldman(3087), _iouri_(3073), smloh1(3061), protocolocon(3056)
2010-11-21T19:22:05  *** Utkarsh <Utkarsh!~admin@117.201.82.162> has quit IRC (Ping timeout: 255 seconds)
2010-11-21T19:22:38  *** Prillicy <Prillicy!~chatzilla@97.104.19.221> has joined #aichallenge
2010-11-21T19:32:47  <Error323> whois this guy: http://72.44.46.68/getplayer?player=Jamil_Najafov.0
2010-11-21T19:35:53  *** Utkarsh <Utkarsh!~admin@117.201.82.245> has joined #aichallenge
2010-11-21T19:37:53  <mega1> I've just played a couple of games against him
2010-11-21T19:38:11  <Error323> mmm
2010-11-21T19:38:21  <Error323> mega1: do you know who testingfrompiter is on official?
2010-11-21T19:38:29  <Error323> and do you wanna test a new version of mine? :)
2010-11-21T19:38:34  <mega1> no idea
2010-11-21T19:38:37  <Error323> its doing really well on tcp
2010-11-21T19:38:52  <mega1> this guy's running about 4 instantly disconnecting bots
2010-11-21T19:39:15  <Error323> indeed.
2010-11-21T19:39:53  <Error323> mega1: so have you got some time?:))
2010-11-21T19:39:56  <Error323> please
2010-11-21T19:40:07  <mega1> Error323: sure, pass me one
2010-11-21T19:41:39  <Error323> mega1: send.
2010-11-21T19:41:41  <medrimonia> gosh there's still such mistakes in my code...
2010-11-21T19:42:34  <medrimonia> Instead of sending source._num_ships ships, I was sending source._owner ships...
2010-11-21T19:44:02  <medrimonia> That's why I was often sending one ship... I understand better now...
2010-11-21T19:44:24  <Error323> O_O
2010-11-21T19:44:28  <Error323> noob :P
2010-11-21T19:45:07  <medrimonia> I don't know how I can write things like that
2010-11-21T19:45:42  <medrimonia> (in fact I know... It's because I'm working with very dirty matrix instead of using beautiful objects structures
2010-11-21T19:46:14  *** Utkarsh <Utkarsh!~admin@117.201.82.245> has quit IRC (Ping timeout: 276 seconds)
2010-11-21T19:46:30  <mega1> ew, this nafajov guy really is all over both tcp servers
2010-11-21T19:49:40  <medrimonia> I think this contest will have learned me a lot... about what I have avoid...
2010-11-21T19:50:21  <medrimonia> *to
2010-11-21T19:51:25  *** Utkarsh <Utkarsh!~admin@117.201.83.45> has joined #aichallenge
2010-11-21T19:54:59  *** Utkarsh <Utkarsh!~admin@117.201.83.45> has quit IRC (Read error: Connection reset by peer)
2010-11-21T20:00:02  *** Appleman1234 <Appleman1234!~Appleman1@CPE-60-226-176-19.qld.bigpond.net.au> has quit IRC (Ping timeout: 255 seconds)
2010-11-21T20:05:09  *** Utkarsh <Utkarsh!~admin@117.201.80.62> has joined #aichallenge
2010-11-21T20:06:00  <mega1> medrimonia: contests, above all things
2010-11-21T20:06:45  *** _Utkarsh <_Utkarsh!~admin@117.201.80.62> has joined #aichallenge
2010-11-21T20:06:46  <medrimonia> mega1: I didn't really understand what you said
2010-11-21T20:06:47  *** amstan <amstan!~amstan@206-248-160-127.dsl.teksavvy.com> has joined #aichallenge
2010-11-21T20:06:47  *** ChanServ sets mode: +o amstan
2010-11-21T20:08:07  <Error323> whoever is nafajov on tcp please kill your connection, because all you do is disconnect!
2010-11-21T20:08:23  <amstan> Error323: tell the admin
2010-11-21T20:08:40  <Error323> amstan: ain't around
2010-11-21T20:08:41  <Error323> :/
2010-11-21T20:09:21  <medrimonia> @seen dhartmei
2010-11-21T20:09:21  <contestbot> medrimonia: dhartmei was last seen in #aichallenge 7 weeks, 4 days, 8 hours, 24 minutes, and 42 seconds ago: <dhartmei> project selection problem?
2010-11-21T20:09:21  <amstan> Error323: there's low chance that the person doing it heard you though
2010-11-21T20:09:43  <Error323> i know
2010-11-21T20:09:50  <Error323> its really annoying >_<
2010-11-21T20:10:17  *** Utkarsh <Utkarsh!~admin@117.201.80.62> has quit IRC (Ping timeout: 276 seconds)
2010-11-21T20:10:24  <medrimonia> "if growth_player(pw,1,0,t_etat)>growth_player(pw,1,0,t_etat)+5:"
2010-11-21T20:10:41  <medrimonia> I'm really writing some things....
2010-11-21T20:11:07  * lavalamp_ freaks out
2010-11-21T20:12:14  <mega1> medrimonia: this contests has been a time sink. Teaches us to think twice before entering next time.
2010-11-21T20:12:22  <mega1> oh, wait. I did think twice.
2010-11-21T20:12:38  <lavalamp_> me too, didn't help
2010-11-21T20:12:51  <medrimonia> mega1: Yes, I'm already waiting for the next one =P
2010-11-21T20:14:31  <mega1> beaten by an off by 1 error at game end: http://72.44.46.68/canvas?game_id=1133458
2010-11-21T20:16:32  *** amstan <amstan!~amstan@206-248-160-127.dsl.teksavvy.com> has quit IRC (Ping timeout: 255 seconds)
2010-11-21T20:17:45  <medrimonia> You started it too late =P
2010-11-21T20:18:30  <mega1> yes, because it counts the growth at the end of the game
2010-11-21T20:18:36  <mega1> with which it would have been a draw
2010-11-21T20:18:37  <contestbot> Top 10 players: bocsimacko(3348), dmj111(3217), wagstaff(3213), GreenTea(3167), Raschi(3133), bix0r4ever(3131), oldman(3108), Manwe56(3085), _iouri_(3074), _Astek_(3071)
2010-11-21T20:18:44  *** yasith_ <yasith_!~yasith@112.134.121.68> has joined #aichallenge
2010-11-21T20:21:20  *** yasith__ <yasith__!~yasith@112.134.125.49> has quit IRC (Ping timeout: 276 seconds)
2010-11-21T20:23:02  <mega1> 4/0/48
2010-11-21T20:23:21  <medrimonia> Hmmm it's really hard to know which bot will be the best
2010-11-21T20:23:45  <lavalamp_> I'm not sure if I will ever unbreak my bot now :(
2010-11-21T20:24:04  <medrimonia> I've too much time out on tcp and on official it's not possible to really evaluate a bot
2010-11-21T20:25:06  <mega1> who's swann?
2010-11-21T20:25:53  <medrimonia> I don't know but he seems to be good
2010-11-21T20:26:23  *** _Utkarsh <_Utkarsh!~admin@117.201.80.62> has quit IRC (Ping timeout: 264 seconds)
2010-11-21T20:30:33  *** Utkarsh <Utkarsh!~admin@117.201.80.211> has joined #aichallenge
2010-11-21T20:31:21  <bhasker> lol http://72.44.46.68/canvas?game_id=1133868
2010-11-21T20:32:35  <mega1> dmj111: is that dmj.5 on the official?
2010-11-21T20:32:44  <medrimonia> Waow he attacks from very far
2010-11-21T20:33:03  *** mceier <mceier!~mceier@chello089077208118.chello.pl> has quit IRC (Quit: leaving)
2010-11-21T20:38:21  <Prillicy> that bot moved around that map skilfully
2010-11-21T20:38:32  <Prillicy> meanwhile I can't seem to clone a variable :(
2010-11-21T20:43:25  *** Utkarsh <Utkarsh!~admin@117.201.80.211> has quit IRC (Ping timeout: 245 seconds)
2010-11-21T20:43:30  <Prillicy> I'm trying to copy my planets variable and it's referencing it :(
2010-11-21T20:43:37  <Prillicy> :'(
2010-11-21T20:44:32  <Error323> You WIN against bocsimacko-b0.70
2010-11-21T20:44:32  <Error323> connected to 98.247.248.39:995, waiting for game
2010-11-21T20:44:39  <Error323> oops wrong window
2010-11-21T20:47:09  <Prillicy> yeah right
2010-11-21T20:47:11  <Prillicy> you're proud
2010-11-21T20:48:32  *** Utkarsh <Utkarsh!~admin@117.201.83.64> has joined #aichallenge
2010-11-21T20:48:42  <phreeza> heheheh
2010-11-21T20:51:54  <Error323> lol
2010-11-21T20:53:14  *** Utkarsh <Utkarsh!~admin@117.201.83.64> has quit IRC (Ping timeout: 245 seconds)
2010-11-21T21:03:19  *** Utkarsh <Utkarsh!~admin@117.201.84.153> has joined #aichallenge
2010-11-21T21:03:34  <burny> personally I think this contest would have been way more neat, if the bots had 100games to gather infoa bout the maps, and opponents play styles, and then 100games that count starts after that
2010-11-21T21:07:55  *** Appleman1234 <Appleman1234!~Appleman1@CPE-60-226-179-130.qld.bigpond.net.au> has joined #aichallenge
2010-11-21T21:12:13  *** Utkarsh <Utkarsh!~admin@117.201.84.153> has quit IRC (Ping timeout: 240 seconds)
2010-11-21T21:14:20  *** Naktibalda <Naktibalda!~quassel@cpc5-heme9-2-0-cust144.9-1.cable.virginmedia.com> has quit IRC (Remote host closed the connection)
2010-11-21T21:16:52  *** Utkarsh <Utkarsh!~admin@117.201.82.165> has joined #aichallenge
2010-11-21T21:18:39  <contestbot> Top 10 players: bocsimacko(3346), _Astek_(3266), wagstaff(3216), dmj111(3180), Raschi(3149), asavis(3101), Manwe56(3081), GreenTea(3077), bix0r4ever(3067), _iouri_(3066)
2010-11-21T21:21:16  *** Utkarsh <Utkarsh!~admin@117.201.82.165> has quit IRC (Ping timeout: 240 seconds)
2010-11-21T21:28:24  <mega1> flanagan, segar, swann, fufu, hermitian, weoi
2010-11-21T21:28:29  <mega1> who are you?
2010-11-21T21:41:25  *** Utkarsh <Utkarsh!~admin@117.201.85.75> has joined #aichallenge
2010-11-21T21:44:27  *** amstan <amstan!~amstan@69-196-188-214.dsl.teksavvy.com> has joined #aichallenge
2010-11-21T21:44:27  *** ChanServ sets mode: +o amstan
2010-11-21T21:47:04  *** phreeza <phreeza!~phreeza@68-076.eduroam.RWTH-Aachen.DE> has quit IRC (Quit: Computer has gone to sleep)
2010-11-21T21:47:22  *** Utkarsh <Utkarsh!~admin@117.201.85.75> has quit IRC (Ping timeout: 265 seconds)
2010-11-21T21:47:41  * lavalamp_ writes O(planets^2 * MaxDistance ^ 4) loop
2010-11-21T21:48:00  *** McLeopold <McLeopold!~scott@c-98-247-248-39.hsd1.wa.comcast.net> has joined #aichallenge
2010-11-21T21:48:30  <McLeopold> Error323: I want to apologize for that incredibly slow game...
2010-11-21T22:00:24  <antimatroid> boy is the top 2 onwards undecided
2010-11-21T22:01:58  *** Utkarsh <Utkarsh!~admin@117.201.84.29> has joined #aichallenge
2010-11-21T22:09:29  *** amriedle <amriedle!~amriedle@CPE0013f7f2be12-CM0013f7f2be0e.cpe.net.cable.rogers.com> has joined #aichallenge
2010-11-21T22:13:04  *** Appleman1234 <Appleman1234!~Appleman1@CPE-60-226-179-130.qld.bigpond.net.au> has quit IRC (Read error: Connection reset by peer)
2010-11-21T22:15:08  *** Appleman1234 <Appleman1234!~Appleman1@CPE-60-226-179-130.qld.bigpond.net.au> has joined #aichallenge
2010-11-21T22:17:33  *** McLeopold_ <McLeopold_!~scott@c-98-247-248-39.hsd1.wa.comcast.net> has joined #aichallenge
2010-11-21T22:17:41  *** McLeopold <McLeopold!~scott@c-98-247-248-39.hsd1.wa.comcast.net> has quit IRC (Ping timeout: 276 seconds)
2010-11-21T22:18:03  <McLeopold_> @ranking
2010-11-21T22:18:05  <contestbot> McLeopold_: Top 10 players: bocsimacko(3346), _Astek_(3226), dmj111(3177), Raschi(3153), asavis(3148), dimkadimon(3113), felixcoto(3109), wagstaff(3089), Manwe56(3071), _iouri_(3056)
2010-11-21T22:18:39  <contestbot> Top 10 players: bocsimacko(3346), _Astek_(3226), dmj111(3177), Raschi(3153), asavis(3148), dimkadimon(3113), felixcoto(3109), wagstaff(3089), Manwe56(3071), _iouri_(3056)
2010-11-21T22:18:45  <McLeopold_> I didn't realize Manwe56 was in the top ten.
2010-11-21T22:18:58  <McLeopold_> Now I don't feel so bad when I lose to him.
2010-11-21T22:20:00  *** amriedle <amriedle!~amriedle@CPE0013f7f2be12-CM0013f7f2be0e.cpe.net.cable.rogers.com> has quit IRC (Quit: Lost terminal)
2010-11-21T22:26:13  *** Utkarsh <Utkarsh!~admin@117.201.84.29> has quit IRC (Ping timeout: 240 seconds)
2010-11-21T22:32:32  *** Utkarsh <Utkarsh!~admin@117.201.83.182> has joined #aichallenge
2010-11-21T22:37:09  *** Utkarsh <Utkarsh!~admin@117.201.83.182> has quit IRC (Ping timeout: 265 seconds)
2010-11-21T22:37:21  <antimatroid> McLeopold_: how's the bot coming along
2010-11-21T22:37:31  <McLeopold_> I hate my life
2010-11-21T22:37:34  <antimatroid> i'm pretty keen to try some different varations once entries are in
2010-11-21T22:37:43  <antimatroid> mine is starting to pull together i think
2010-11-21T22:37:46  <antimatroid> i was getting worried
2010-11-21T22:38:00  <McLeopold_> I think I've been making it worse for a week straight
2010-11-21T22:38:14  <antimatroid> hehe, i just added in something to make it faster
2010-11-21T22:38:24  <antimatroid> i'm no longer regenerating every single future state at every node
2010-11-21T22:38:29  <antimatroid> only those that have changed
2010-11-21T22:38:29  <McLeopold_> It can do some pretty awesome stuff sometimes, but it's really up to the map type.
2010-11-21T22:38:49  <McLeopold_> Yeah, I did that too
2010-11-21T22:39:05  <McLeopold_> I think I've got a bunch of extra crap that I don't use anymore slowing it down.
2010-11-21T22:39:17  <antimatroid> yeah, i think i've got most of that out
2010-11-21T22:39:21  <lavalamp_> I need more time :(
2010-11-21T22:39:34  <antimatroid> lavalamp_: everyone does, so it doesn't matter
2010-11-21T22:39:50  * antimatroid is glad he has nothing to do all week but this
2010-11-21T22:40:04  <antimatroid> did you guys see the forum post from space invaders?
2010-11-21T22:40:14  <antimatroid> http://ai-contest.com/forum/viewtopic.php?f=17&t=1136
2010-11-21T22:40:20  <antimatroid> doesn't look very good, but we'll see
2010-11-21T22:40:46  <lavalamp_> no I mean I broke everything and am not sure if I can put it back together again
2010-11-21T22:41:02  <antimatroid> go back to and old version?
2010-11-21T22:41:15  <antimatroid> an*
2010-11-21T22:41:27  <antimatroid> my versioning is very poor, i just have a folder packed with zip files
2010-11-21T22:42:30  <lavalamp_> i could but I would just need to do the same things to it
2010-11-21T22:42:41  <antimatroid> ah crap, dimkadimon is in the top 10
2010-11-21T22:42:44  <antimatroid> he's not allowed to beat me
2010-11-21T22:42:54  <McLeopold_> http://tcp.zeroviz.us/canvas.cgi?game_id=19669
2010-11-21T22:43:35  <antimatroid> you were sort of handed that on a silver platter :P
2010-11-21T22:44:22  <McLeopold_> who, me/
2010-11-21T22:46:22  <antimatroid> yeah
2010-11-21T22:46:28  <McLeopold_> how so?
2010-11-21T22:46:34  <antimatroid> when he sent 50 ships across the map at the start
2010-11-21T22:47:12  <McLeopold_> hmm, maybe
2010-11-21T22:47:56  <McLeopold_> I was watching the game as it was played, thinking my bot shouldn't have gone for the center.
2010-11-21T22:48:05  <bhasker> interesting my bot is climbing of its own accord
2010-11-21T22:48:08  <bhasker> at #40 now
2010-11-21T22:48:16  <burny> any time(especially early one), ships are sent, and it doesn't capture.. it's a bad move
2010-11-21T22:49:14  <antimatroid> that depends, sometimes i will start going for a neutral, but it will no longer be optimal
2010-11-21T22:49:18  <antimatroid> there's nothing i can do about that
2010-11-21T22:49:20  *** Utkarsh <Utkarsh!~admin@117.201.81.115> has joined #aichallenge
2010-11-21T22:51:15  <burny> who wants to see an awesome game?
2010-11-21T22:51:17  <burny> http://ai-contest.com/visualizer.php?game_id=7698553
2010-11-21T22:51:39  <McLeopold_> Yeah, you got to play a real one!
2010-11-21T22:51:58  <McLeopold_> almost
2010-11-21T22:52:33  <McLeopold_> looks like you pegged you elo though, you where you thought you would be
2010-11-21T22:52:41  <janzert> burny: hehe
2010-11-21T22:52:57  <janzert> burny: btw the way if you didn't see it you were up to rank 502 at one time
2010-11-21T22:53:12  <burny> cool, was only 680 when I saw
2010-11-21T22:53:29  <burny> kinda sad, that such a gimpy strat could get to 500
2010-11-21T22:53:48  <antimatroid> burny: still think the games trivial? :P
2010-11-21T22:53:53  <burny> ya
2010-11-21T22:54:02  *** Utkarsh <Utkarsh!~admin@117.201.81.115> has quit IRC (Ping timeout: 260 seconds)
2010-11-21T22:54:03  <burny> er, not trivial.. simple
2010-11-21T22:54:10  <janzert> well, you *were* probably over rated then ;)
2010-11-21T22:54:25  <burny> janzert, I resubmited since then
2010-11-21T22:54:36  <burny> made it worse
2010-11-21T22:54:46  <janzert> hate it when that happens
2010-11-21T22:54:54  <antimatroid> lame, dimkadimon is up to third
2010-11-21T22:54:58  <McLeopold_> hahaha: --> Most likely to play next game within the next 3674 games.
2010-11-21T22:55:17  <burny> I change it so it sends exactly enough ships..
2010-11-21T22:55:17  <janzert> ?
2010-11-21T22:55:20  <McLeopold_> that number looks horrible
2010-11-21T22:55:23  <burny> it did better when it sent extra ships
2010-11-21T22:55:35  <janzert> means you just played not that long ago
2010-11-21T22:55:54  <antimatroid> burny: what do you mean by enough?
2010-11-21T22:55:55  <McLeopold_> Yeah, I know what it means, but it's still funny to me.
2010-11-21T22:56:03  <antimatroid> enough for me is more than they could possibly send there in that time
2010-11-21T22:56:05  <burny> if neutral has 50.. I send 51
2010-11-21T22:56:30  <antimatroid> except for defense, i will cancel out their moves without worrying about what else tehy might do
2010-11-21T22:56:36  <burny> used to send somwhere around 55-80, depending on stupid stuff
2010-11-21T22:57:08  <janzert> I debated about say something like "likely to not play for a long time" down to " likely to play very soon" but decided to just go the simple route and give the number
2010-11-21T22:57:29  <antimatroid> numbers are good :)
2010-11-21T22:57:39  <janzert> probably just giving a general idea would have reduced refresh load though :P
2010-11-21T22:57:44  <burny> anyone else see somethign screwy with that game though?
2010-11-21T22:58:09  <antimatroid> no?
2010-11-21T22:58:43  <janzert> besides deciding to take a zero growth planet as the first attack?
2010-11-21T22:58:51  <antimatroid> ehe
2010-11-21T22:59:00  <janzert> actually and only attack on a neutral
2010-11-21T22:59:37  <burny> guess it's just certain browsers/os's/etc.
2010-11-21T23:00:15  <McLeopold_> @stats
2010-11-21T23:00:15  <contestbot> McLeopold_: I have 1 registered users with 0 registered hostmasks; 1 owner and 0 admins.
2010-11-21T23:00:18  <burny> for me and someone else, the middle planet doesn't show up
2010-11-21T23:00:22  <McLeopold_> hmm
2010-11-21T23:00:27  <antimatroid> ahh, i feel better now, dimkadimon lost to 70th place
2010-11-21T23:00:30  <janzert> uh oh
2010-11-21T23:00:43  <janzert> contestbot got wiped?
2010-11-21T23:00:44  <contestbot> janzert: Run as fast as you can and don't look back.
2010-11-21T23:00:45  <antimatroid> burny: what browser?
2010-11-21T23:00:50  <McLeopold_> where's the stats page?
2010-11-21T23:00:50  <burny> ie
2010-11-21T23:01:03  <antimatroid> there's your problem :P
2010-11-21T23:01:31  <janzert> http://contestbot.hypertriangle.com/
2010-11-21T23:01:45  <McLeopold_> for games per minute
2010-11-21T23:01:50  <janzert> err, you mean game_stats
2010-11-21T23:01:52  <janzert> yeah
2010-11-21T23:01:55  <burny> with the middle planet being invisible..
2010-11-21T23:01:58  <janzert> @gamestats
2010-11-21T23:01:59  <contestbot> janzert: gamestats = http://ai-contest.com/game_stats.php
2010-11-21T23:02:02  <McLeopold_> thx
2010-11-21T23:02:10  <burny> that game looks like, my first attack is on their base.. and it just evaporates part way through..
2010-11-21T23:02:27  <burny> and then after they win, they poof from the game, and I have ships come out of no where
2010-11-21T23:02:27  <McLeopold_> janzert: you could do a quick eta
2010-11-21T23:02:34  <lavalamp_> aaak it compiled
2010-11-21T23:02:37  * lavalamp_ scared to run
2010-11-21T23:02:44  <virdo> burny: get chrome or firefox :)
2010-11-21T23:03:00  <janzert> McLeopold_: yeah, but I'd rather not ;)
2010-11-21T23:03:13  <janzert> McLeopold_: since making it accurate is actually a hard problem
2010-11-21T23:03:22  <burny> forget accurate..
2010-11-21T23:03:24  <janzert> as the gpm varies wildly
2010-11-21T23:03:30  <McLeopold_> well, only ballpark accurate
2010-11-21T23:03:39  <McLeopold_> give or take 15 mintues or so
2010-11-21T23:03:44  <janzert> people *will* complain when it isn't accurate
2010-11-21T23:03:52  <McLeopold_> yes they will
2010-11-21T23:04:13  <McLeopold_> just show a picture of the old ie download progress bar
2010-11-21T23:04:13  <burny> just say "4500games to go.. roughly 5800games per hour
2010-11-21T23:04:15  <burny> or whateer
2010-11-21T23:04:19  <janzert> McLeopold_: heh, yeah I could probably just say within the next hour then people will be happy when it averages 25 minutes :P
2010-11-21T23:04:22  *** bhasker_ <bhasker_!~bhasker@adsl-99-25-118-235.dsl.pltn13.sbcglobal.net> has joined #aichallenge
2010-11-21T23:05:07  <McLeopold_> does the last 24 period fluctuate much or just the last 5 minutes?
2010-11-21T23:05:22  <janzert> but really it's just something that I don't want to give any more though to
2010-11-21T23:05:38  <janzert> last 24 hours varies by about 10 or so
2010-11-21T23:05:46  <burny> janz, all I meant by it was..
2010-11-21T23:05:49  <janzert> it seems to be trending down as we head toward the finals
2010-11-21T23:06:05  <McLeopold_> there you go, queue length / ( last24 + 10)
2010-11-21T23:06:06  <burny> if you tell people what the average per hour is.. then they won't have to hit f5 20times, to measure it themselves
2010-11-21T23:06:30  <janzert> I think more people are submitting bots that actually play games now :)
2010-11-21T23:06:35  *** bhasker <bhasker!~bhasker@adsl-99-25-118-235.dsl.pltn13.sbcglobal.net> has quit IRC (Ping timeout: 276 seconds)
2010-11-21T23:06:42  <burny> while a "estimated 15:37 remaining" would be great..
2010-11-21T23:06:51  <burny> it's not worth the effort really
2010-11-21T23:07:09  <janzert> McLeopold_: except the actual current rate of games can be anywhere from half to double the 24 hour
2010-11-21T23:07:31  <McLeopold_> oh
2010-11-21T23:08:12  <janzert> yeah, we seem to go anywhere from 40 to 160 gpm averaged over 5 minutes
2010-11-21T23:08:19  <burny> or heck.. "50-140 games per minute", would be tons accurate
2010-11-21T23:09:12  <burny> janz.. the bot I had that got up to 500ish rank...
2010-11-21T23:09:30  <burny> the only thing it did, was pick a planet, and a place to attack, and send it off
2010-11-21T23:09:37  <janzert> heh, in the last 7 minutes the 5 minute average went from 38 to 143
2010-11-21T23:09:37  <Prillicy> I can't seem to connect to the thing, I think I'd need ip-passthrough on my router
2010-11-21T23:09:46  <burny> almsot identcal to teh default bot
2010-11-21T23:10:27  <janzert> burny: yeah, I really do believe it was over ranked. when coming from the bottom up like that the ranking will tend to overshoot before settling down
2010-11-21T23:11:22  <burny> maybe, should I resubmit it?
2010-11-21T23:11:32  <burny> but ya.. probably was over-ranked some, for sure
2010-11-21T23:12:07  <janzert> If you'd like to see where it settles in at but don't do it for me
2010-11-21T23:12:27  <burny> I don't care about that bot.. or the current one
2010-11-21T23:12:28  <janzert> If I were you I'd just work on trying to make a better one
2010-11-21T23:12:36  * janzert nods
2010-11-21T23:13:08  <janzert> wow, just noticed new users has really fallen off in the last couple of days
2010-11-21T23:13:32  <janzert> new submissions have stayed steady though
2010-11-21T23:14:06  <janzert> I expect they will probably climb some this week and really spike at the end :)
2010-11-21T23:15:02  <antimatroid> ahh, i fixed my time issues
2010-11-21T23:15:11  <antimatroid> i now delete my tree at the start of each turn rather than at the end
2010-11-21T23:16:08  <antimatroid> which also gives me a full second on the first turn, accounting for set up time, nice
2010-11-21T23:16:15  <burny> just re-subed.. think it's the same as original one
2010-11-21T23:16:31  <burny> might have a couple minor balance changes
2010-11-21T23:16:51  <antimatroid> i find i rush things when i try to test against people all the time
2010-11-21T23:16:56  <antimatroid> and i end up with a huge buggy mess
2010-11-21T23:18:03  <McLeopold_> janzert: if you took out crashed games from the average, it might stabalize
2010-11-21T23:18:23  <McLeopold_> somewhat
2010-11-21T23:18:37  <McLeopold_> But I agree, don't do anything about it right now.
2010-11-21T23:18:39  <contestbot> Top 10 players: bocsimacko(3352), _Astek_(3226), dmj111(3188), asavis(3174), felixcoto(3159), Raschi(3143), wagstaff(3091), bix0r4ever(3083), dimkadimon(3056), oldman(3055)
2010-11-21T23:18:51  <antimatroid> i think in the future it should be the case that if you have played more than x games and your propotion of crashes is greater than y then your current bot gets disabled
2010-11-21T23:19:21  <burny> agree
2010-11-21T23:19:28  <McLeopold_> antimatroid: I don't think that is good.  I like my do over scheme for games against crashed bots better.
2010-11-21T23:19:32  <burny> no point letting bots play after crashing 5-10times
2010-11-21T23:19:38  <McLeopold_> The crash bots don't really take much cpu
2010-11-21T23:19:53  <antimatroid> no, but they're annoying for people to play against
2010-11-21T23:19:55  <McLeopold_> And you should still get your fair share of "real" games.
2010-11-21T23:20:28  <McLeopold_> They could just not show them in the main listings, or flag them as crashed.
2010-11-21T23:20:34  <janzert> The people that are caused the most problem are the lowest ranked players who don't crash
2010-11-21T23:20:54  <McLeopold_> So, when it was your turn, you may see 2 games right away and think, "oh, one was a crash, at least I got a real game."
2010-11-21T23:21:12  <antimatroid> wouldn't it be cooler if they'd had two real games?
2010-11-21T23:21:24  <McLeopold_> no
2010-11-21T23:21:40  <McLeopold_> not for everyone else
2010-11-21T23:21:40  <janzert> so yeah in future contests I'd really like to see them removed after say 10-20 games if you crash on the first turn in over 10% of them
2010-11-21T23:22:01  <antimatroid> janzert: yeah, possibly even more like 20%
2010-11-21T23:22:21  <janzert> oops, I meant 90% was thinking the other way around
2010-11-21T23:22:24  <burny> if you crash atall in over 50% of games(and are ranked low)
2010-11-21T23:22:30  <antimatroid> i am glad one can no longer be dq'd for timing out
2010-11-21T23:22:34  <McLeopold_> janzert: if you say that up front and allow people to track their eligability for furture games (crash rate) then that would work.
2010-11-21T23:22:36  *** flag <flag!~root@69-165-173-172.dsl.teksavvy.com> has quit IRC (Quit: Lost terminal)
2010-11-21T23:22:39  <antimatroid> that would have been a huge problem for people this time i think
2010-11-21T23:22:58  <janzert> McLeopold_: yep, as long as it was in from the beginning or close to it, it should be fine
2010-11-21T23:23:18  <burny> my strat is going to be to write a bot that ALWAYS ties
2010-11-21T23:23:34  <burny> and I'll have another bot, with inline assembly implimentation of the STD::vector class
2010-11-21T23:23:40  <McLeopold_> and as long as they have good data proving there bot needs to be removed, then their should not be too many complaints.
2010-11-21T23:23:52  <janzert> as an aside I finally got code in to avoid repairing against recently played opponents
2010-11-21T23:24:00  <burny> the first bot, will just send 1ship back and forth between planets, hundreds of times per turn
2010-11-21T23:24:01  <McLeopold_> nice
2010-11-21T23:24:17  <burny> causeing both bots to time out, and thus a tie
2010-11-21T23:24:23  <janzert> it's about as simple as it can be right now but at least it's something
2010-11-21T23:24:33  <antimatroid> wow, i've more than doubled the size of my tree generation abilities in the last 12 hours :D
2010-11-21T23:24:46  <antimatroid> and fixed my time issues, awesome
2010-11-21T23:25:57  <McLeopold_> janzert: was it a problem with tron?  Or is this just a bigger contest?
2010-11-21T23:26:25  <janzert> McLeopold_: hmm?
2010-11-21T23:26:30  <burny> any indication of what teh strongest few bots are, that have zero 'looking into the future' code?
2010-11-21T23:26:36  <janzert> what problem?
2010-11-21T23:26:38  <McLeopold_> worrying about bad games
2010-11-21T23:26:45  <McLeopold_> getting enough games in
2010-11-21T23:26:55  <janzert> yes and yes
2010-11-21T23:27:27  <janzert> although the big problem in tron was just that the pairing was simply uniform randomly chosen opponents
2010-11-21T23:27:42  <janzert> what we have now is already certainly a huge step up
2010-11-21T23:27:56  <burny> what rng is used for the pairing?
2010-11-21T23:28:27  <burny> a horribe one?
2010-11-21T23:28:31  <janzert> burny: built in python which I'm pretty sure is Mersenne twister
2010-11-21T23:28:36  <janzert> no
2010-11-21T23:29:19  <antimatroid> janzert: agreed, what we have now is soooo much better
2010-11-21T23:29:33  <janzert> burny: the seeding is probably more of a concern than the actual generator
2010-11-21T23:29:54  <janzert> since we aren't running it for long periods of time
2010-11-21T23:29:59  <burny> is it bad code then, that results in the same pattern in opponents, or gettign 10games in 4seconds, etc.?
2010-11-21T23:30:46  <janzert> not sure what you mean by pattern of opponents, but the multiple games at once is problems with the way the matchups are generated
2010-11-21T23:31:11  <janzert> mostly multiple instances of the pairing code being run at once
2010-11-21T23:31:25  <lavalamp_> IT RUNS
2010-11-21T23:31:30  <burny> ah, so multiple isntances all think the same person is next?
2010-11-21T23:31:36  <janzert> right
2010-11-21T23:31:56  <lavalamp_> ...for 12 turns
2010-11-21T23:32:01  <burny> and by the pattern I mean..
2010-11-21T23:32:10  <janzert> lavalamp_: that 12 times better right? :)
2010-11-21T23:32:41  <burny> alternating between teh same 2 opponents 3 times ina row
2010-11-21T23:32:43  <lavalamp_> considering it was 0 turns a minute ago, I think it's more than that...
2010-11-21T23:33:32  <janzert> burny: yeah, that's just a result of using a pareto distribution without any guards against repeat pairings. shouldn't see it anymore
2010-11-21T23:34:06  <janzert> lavalamp_: way to go infinite improvement :)
2010-11-21T23:34:30  <janzert> just do it two more times and you'll be up there with mega1
2010-11-21T23:34:39  <McLeopold_> I just got a lead on mega1 in a tcp game...
2010-11-21T23:34:45  <McLeopold_> and it's gone
2010-11-21T23:40:22  <burny> could be nice to give a point once per day..
2010-11-21T23:40:47  <McLeopold_> ouch, divide by zero error. :/
2010-11-21T23:40:49  <burny> that can be used to CHOOSE your next opponent, and jump to near the front of que
2010-11-21T23:41:49  <lavalamp_> good news: 10x faster. bad news: doesn't work.
2010-11-21T23:42:18  <burny> 10steps forward, and infinity steps back!
2010-11-21T23:42:21  *** rebelxt_ <rebelxt_!181c0b18@gateway/web/freenode/ip.24.28.11.24> has joined #aichallenge
2010-11-21T23:42:51  <antimatroid> ahh, story of my life
2010-11-21T23:42:57  <antimatroid> now it makes too many moves sometimes
2010-11-21T23:44:50  *** davidjliu <davidjliu!473a6d0e@gateway/web/freenode/ip.71.58.109.14> has quit IRC (Quit: Page closed)
2010-11-21T23:45:22  <burny> actually: I just had an idea, of something that could actually be slightly neat, for the next contest..
2010-11-21T23:45:32  <burny> having multiple tiers of contestants..
2010-11-21T23:45:44  <Prillicy> ?
2010-11-21T23:45:53  <burny> eg. in addition to teh current contest ladder..
2010-11-21T23:46:11  <burny> there'd be one where bots are only allowed to send a single attack each turn
2010-11-21T23:46:28  <burny> another, where all attacks have to be 50% of the population.. etc.
2010-11-21T23:46:40  <Prillicy> you're thinking the next contest won't be a completely different decision option
2010-11-21T23:47:00  <Prillicy> galcon bot creation ends in feb
2010-11-21T23:47:19  <burny> I thought it was next week?
2010-11-21T23:47:36  <Prillicy> ?
2010-11-21T23:48:47  <antimatroid> Prillicy: entries close in under a week
2010-11-21T23:48:55  <Prillicy> bah
2010-11-21T23:48:58  <Prillicy> there goes my fun
2010-11-21T23:49:20  <antimatroid> you have till like the weekend
2010-11-21T23:49:30  <antimatroid> it's been going for 3 months now
2010-11-21T23:49:38  <antimatroid> well, almost
2010-11-21T23:50:06  <burny> I'm just thinking in the context of.. makign the contest be much simplified, for less experienced programmers, to still be able to compete a bit
2010-11-21T23:50:23  <antimatroid> burny: tron was a lot easier
2010-11-21T23:50:30  <burny> because the problem of "what is the best planet to attack this turn?".. is WAY easier, then the problem as a whole
2010-11-21T23:50:37  <antimatroid> but i think people are interested in doing fun games
2010-11-21T23:50:47  <antimatroid> if it's too simple everyone just does fairly similar minimax implementations
2010-11-21T23:51:19  <burny> I'm not saying make it easy INSTEAD OF hard..
2010-11-21T23:51:22  <burny> saying in ADDITION to hard
2010-11-21T23:51:44  <burny> 2 totally seperate ladders.. one for 'real' bots.. one for gimped nuby bots
2010-11-21T23:52:05  <antimatroid> meh, i prefer just one contest
2010-11-21T23:52:20  <burny> because you're capable of doing a 'real' one :P
2010-11-21T23:52:39  <antimatroid> people can still submit their best effort
2010-11-21T23:52:44  <antimatroid> and compete with people of similar quality
2010-11-21T23:52:46  <burny> the default AI given with the starter package, is beating like 40% of people
2010-11-21T23:53:09  <antimatroid> yeah, a lot of them aren't real contestants
2010-11-21T23:53:16  <antimatroid> we need a way to filter those early submissions out
2010-11-21T23:53:39  <antimatroid> i think people should have to resubmit every few weeks, and by that within the last couple of weeks before the deadline if they want their bot run at the end
2010-11-21T23:54:02  <antimatroid> or even just have them log in and say yes, keep running my bot, somehow
2010-11-21T23:55:27  <burny> I just think that lots of people who put 20-200 minutes of effort into this.. would like to put in a small ammount of effort, to have a half-assed bot, that can compete in a different ladder
2010-11-21T23:57:27  <virdo> it all boils down to the same thing -- think of top 1000 as one ladder and top 1000-2000 as another ;)
2010-11-21T23:58:06  <burny> like.. FPS's..
2010-11-21T23:58:13  <burny> lots of people suck at them, but like the concept..
2010-11-21T23:58:50  <burny> having a seperate FPS ladder, where you're not allowed to aima nd move at the same time.. makes it more approachable, for noobly's
2010-11-21T23:59:17  *** rebelxt_ <rebelxt_!181c0b18@gateway/web/freenode/ip.24.28.11.24> has quit IRC (Ping timeout: 265 seconds)