2010-11-19T00:02:19 *** hellman has joined #aichallenge 2010-11-19T00:06:24 * Mjothvitnir ahhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhh splat 2010-11-19T00:06:27 *** Mjothvitnir has quit IRC () 2010-11-19T00:17:54 *** davidjliu has joined #aichallenge 2010-11-19T00:18:02 Top 10 players: bocsimacko(3703), wagstaff(3696), bix0r4ever(3543), deccan(3526), protocolocon(3452), davidjliu(3442), Silver_S(3437), george(3435), shangas(3420), dmj111(3392) 2010-11-19T00:20:17 *** delt0r has joined #aichallenge 2010-11-19T00:20:47 *** delt0r___ has quit IRC (Ping timeout: 260 seconds) 2010-11-19T00:57:40 *** narnach has joined #aichallenge 2010-11-19T01:08:42 *** amstan_ has joined #aichallenge 2010-11-19T01:08:42 *** ChanServ sets mode: +o amstan_ 2010-11-19T01:12:10 *** amstan_ has quit IRC (Remote host closed the connection) 2010-11-19T01:18:04 Top 10 players: bocsimacko(3693), wagstaff(3543), bix0r4ever(3537), deccan(3474), protocolocon(3461), Silver_S(3460), george(3434), davidjliu(3433), shangas(3418), Raschi(3402) 2010-11-19T01:28:05 *** sigh has quit IRC (Ping timeout: 255 seconds) 2010-11-19T01:29:02 *** bhasker has quit IRC (Quit: bhasker) 2010-11-19T01:40:43 *** yasith has quit IRC (Ping timeout: 252 seconds) 2010-11-19T01:40:59 *** Appleman1234 has quit IRC (Ping timeout: 255 seconds) 2010-11-19T01:48:38 *** superflit has quit IRC (Ping timeout: 255 seconds) 2010-11-19T01:53:45 *** yasith has joined #aichallenge 2010-11-19T01:56:17 *** McLeopold has joined #aichallenge 2010-11-19T02:01:33 antimatroid: i fixed the visualizer thing you were annoyed with 2010-11-19T02:01:47 janzert: did you implement my visualizer changes yet? 2010-11-19T02:01:51 *** yasith has quit IRC (Ping timeout: 265 seconds) 2010-11-19T02:03:47 wow, dhartmei's got the change, I feel so proud 2010-11-19T02:06:54 you're referring to bringing the active planet to the top? no I haven't yet, I did start to look at it though :} 2010-11-19T02:09:00 ok, made 1 tweak just now 2010-11-19T02:13:42 *** yasith has joined #aichallenge 2010-11-19T02:14:30 *** narnach has quit IRC (Ping timeout: 245 seconds) 2010-11-19T02:18:04 Top 10 players: bocsimacko(3699), wagstaff(3574), bix0r4ever(3511), deccan(3511), Silver_S(3506), protocolocon(3461), george(3434), davidjliu(3430), shangas(3424), Raschi(3420) 2010-11-19T02:19:59 *** narnach has joined #aichallenge 2010-11-19T02:21:03 *** hellman has quit IRC (Remote host closed the connection) 2010-11-19T02:25:42 *** davidjliu has quit IRC (Ping timeout: 265 seconds) 2010-11-19T02:34:55 *** virdo has quit IRC (Ping timeout: 265 seconds) 2010-11-19T02:38:01 web programming feels like a giant vat of plagurism 2010-11-19T02:41:07 ok, changes are backported 2010-11-19T02:41:15 and live :) 2010-11-19T02:41:59 sweet! 2010-11-19T02:42:41 ooh, you missed the latest tweek 2010-11-19T02:43:00 what's that? 2010-11-19T02:43:14 a change to the switch active function 2010-11-19T02:43:31 it uses the z-order as well 2010-11-19T02:43:38 it's in there 2010-11-19T02:43:43 or should be 2010-11-19T02:43:56 the behavoir doesn't seem like it is 2010-11-19T02:44:32 huh, the code is there 2010-11-19T02:45:04 oh, needed to refresh 2010-11-19T02:45:10 ahh 2010-11-19T02:51:47 *** lavalamp_ has joined #aichallenge 2010-11-19T02:51:59 hi lavalamp_ 2010-11-19T02:52:04 yo 2010-11-19T02:52:16 where's that visualize pastebin? 2010-11-19T02:52:22 visualizer* 2010-11-19T02:52:27 @pastewars 2010-11-19T02:52:28 McLeopold: pastewars = http://visualizer.naktibalda.lt/ 2010-11-19T02:52:45 I've got crazy maps on port 996 2010-11-19T02:52:53 you should try it once or twice 2010-11-19T02:53:12 McLeopold: i will play later 2010-11-19T02:53:18 still refining :| 2010-11-19T02:53:20 cool 2010-11-19T02:53:32 aw man it hasn't been updated :( 2010-11-19T02:53:33 http://visualizer.naktibalda.lt/574 2010-11-19T02:53:34 it's got the no neutrals maps, or sort of 2010-11-19T02:53:49 *** Mistmanov has joined #aichallenge 2010-11-19T02:54:00 this game was awesome, my bots swapped bases three times 2010-11-19T02:54:03 haha, fleets from nowhere appear 2010-11-19T02:55:06 lavalamp_: i think you still have planet id bias 2010-11-19T02:55:32 hmm why? 2010-11-19T02:55:53 nah, there's a planet off teh screen 2010-11-19T02:55:59 it's using the old visualisetr 2010-11-19T02:56:26 nvm, I was trying to analyze why you attack the back line or front line first 2010-11-19T02:56:56 ah that's 'cause I'm retarded 2010-11-19T02:57:09 is it just me or are there new names in the top 10 2010-11-19T02:57:10 well, back line did better 2010-11-19T02:57:16 @rankings 2010-11-19T02:57:17 McLeopold: Top 10 players: bocsimacko(3682), wagstaff(3555), bix0r4ever(3497), deccan(3492), Silver_S(3471), Raschi(3453), protocolocon(3444), BaronTrozo(3427), george(3424), davidjliu(3421) 2010-11-19T02:57:26 a couple 2010-11-19T02:57:35 wagstaff has been up there before 2010-11-19T02:57:45 yeah, although not recently 2010-11-19T02:57:55 then deccan and barontrozo wtf? 2010-11-19T02:57:59 not sure who silver_s is 2010-11-19T02:59:04 http://tcp2.zeroviz.us/canvas.cgi?game_id=9 2010-11-19T02:59:49 that's funny, i just finished watching that game :) 2010-11-19T03:00:02 awesome 2010-11-19T03:02:10 http://tcp2.zeroviz.us/canvas.cgi?game_id=16 2010-11-19T03:02:20 Accoun: dhartmei put new maps on the old tcp yesterday :) 2010-11-19T03:02:45 damn i wish lots more maps were added early on 2010-11-19T03:03:23 maybe we can get them to run a crazy map competition in january ^^ 2010-11-19T03:03:24 map types or # of maps 2010-11-19T03:04:00 map types 2010-11-19T03:04:05 I think I need to lower the starting ship counts 2010-11-19T03:05:00 what's the ip/ port for your crazy map server? I wanna start a bot 2010-11-19T03:05:05 http://tcp2.zeroviz.us/canvas.cgi?game_id=17 2010-11-19T03:05:06 996 2010-11-19T03:05:12 i'd be willing to put a little bit of effort in to try and get it to work half decently wth n>2 players 2010-11-19T03:05:14 ^^ that map is a giant fail 2010-11-19T03:05:29 abandon all planets and go for the growth 5 2010-11-19T03:05:51 yeah, but you need a whole new game engine for that 2010-11-19T03:06:00 I was thinking about moving planets 2010-11-19T03:06:10 How would you calc turn remaining? 2010-11-19T03:06:26 distance changes every turn, would be messy but doable 2010-11-19T03:06:29 yes!!! http://72.44.46.68/ in new map set! 2010-11-19T03:06:33 and i don't know if it's be overly fair 2010-11-19T03:06:52 Accoun: and new visualizer 2010-11-19T03:07:08 I'm surprised dhartmei used my messy planet id and growth 2010-11-19T03:07:19 It really looks ugly :) 2010-11-19T03:07:25 my bots would fail with moving planets 2010-11-19T03:07:31 my bot would too 2010-11-19T03:07:39 i ignore coordinates after the first move 2010-11-19T03:07:40 yeah, I don't even know how I would calculate for that 2010-11-19T03:07:43 I precalc a bunch of tables 2010-11-19T03:07:44 but wouldn't take me long to fix 2010-11-19T03:07:56 you'd have to be given a schedule of the movements or something 2010-11-19T03:08:04 yeah exactly 2010-11-19T03:08:24 otherwise you couldn't predict 2010-11-19T03:08:37 yeah, you'd want to give them equations of ellipsoids on the map 2010-11-19T03:08:46 then planets on each ellipsoid and starting positions 2010-11-19T03:09:09 I think n players is better to work on first 2010-11-19T03:09:09 well, i think soid is 3d ? i'm terrible with names 2010-11-19T03:09:29 oh, and 3d maps 2010-11-19T03:09:34 with a cool visualizer :) 2010-11-19T03:09:48 yeah 2010-11-19T03:10:59 3d maps wouldn't change a thing for me 2010-11-19T03:11:32 my bot is soley dependant on planet distances 2010-11-19T03:11:35 well, it might 2010-11-19T03:11:43 i'd have to change to handle 3 coordinates, but neither 2010-11-19T03:11:46 and the game changes 2010-11-19T03:11:51 there would be a lot more planets at the same distance all at once 2010-11-19T03:11:56 ^^ 2010-11-19T03:12:12 So, there would be alot more r/p/s situations 2010-11-19T03:12:14 yeah but that doesn't change the math 2010-11-19T03:12:23 no, not the math, but I think the stragegy 2010-11-19T03:12:26 that's why i wanted to do 3d tron :( 2010-11-19T03:12:38 3d tron would be cool 2010-11-19T03:12:45 super cool, hard to watch 2010-11-19T03:12:51 oo yeah 2010-11-19T03:12:52 3d dots and boxes? 2010-11-19T03:12:58 yeah 2010-11-19T03:13:01 dots and cubes? 2010-11-19T03:13:15 dots and boxes AND cubes? 2010-11-19T03:13:20 oh god 2010-11-19T03:13:24 and hypercubes 2010-11-19T03:13:31 not quite sure how you'd score that 2010-11-19T03:13:34 dang, you beat me to that ;) 2010-11-19T03:13:36 heh make a visualizer for that! 2010-11-19T03:13:52 box = 1, cube = 5 2010-11-19T03:14:13 a cube is 6 boxes 2010-11-19T03:14:19 not always 2010-11-19T03:14:24 lots overlop 2010-11-19T03:14:26 overlap* 2010-11-19T03:14:34 ah 2010-11-19T03:14:37 well it is, but not all of the sides are unique to that cube 2010-11-19T03:15:05 although, I would prefer a game with simultaneous moves 2010-11-19T03:15:11 yeah, same 2010-11-19T03:15:18 like diplomacy, no press 2010-11-19T03:15:36 i like the idea of a 2d shooter game of sorts 2010-11-19T03:15:45 http://www.diplom.org/index.py 2010-11-19T03:16:35 how about zendo... 2010-11-19T03:17:04 http://en.wikipedia.org/wiki/Zendo_%28game%29 2010-11-19T03:17:28 that would be pretty tough 2010-11-19T03:18:04 Top 10 players: bocsimacko(3688), wagstaff(3531), bix0r4ever(3524), deccan(3495), Silver_S(3479), Raschi(3475), protocolocon(3449), BaronTrozo(3439), davidjliu(3430), george(3429) 2010-11-19T03:18:53 @tcp 2010-11-19T03:18:54 Accoun: tcp = (#1) http://www.benzedrine.cx/planetwars/, or (#2) http://72.44.46.68/, or (#3) 98.247.248.39:995 (zeroviz.us), or (#4) http://tcp.zeroviz.us/ 2010-11-19T03:19:11 http://www.benzedrine.cx/planetwars/ now with new map set! play please 2010-11-19T03:19:28 not sure that game would work very well 2010-11-19T03:21:34 *** aerique has joined #aichallenge 2010-11-19T03:23:37 that's similar to mastermind? 2010-11-19T03:24:52 I think it would 2010-11-19T03:25:10 *** McLeopold has quit IRC (Quit: Leaving) 2010-11-19T03:25:50 kinda similar 2010-11-19T03:26:31 but now that I think about it you might have to have already solved the game to enforce the rules :/ 2010-11-19T03:36:39 *** McLeopold has joined #aichallenge 2010-11-19T03:42:34 i love dhartmei very math! 2010-11-19T03:43:17 http://www.benzedrine.cx/ ligth in the dark! 2010-11-19T03:46:36 *** perror has joined #aichallenge 2010-11-19T03:50:06 *** narnach1 has joined #aichallenge 2010-11-19T03:50:06 *** narnach1 is now known as narnach_mb 2010-11-19T04:07:07 *** yasith has quit IRC (Ping timeout: 260 seconds) 2010-11-19T04:09:54 *** Appleman1234 has joined #aichallenge 2010-11-19T04:14:20 *** antimatroid1 has joined #aichallenge 2010-11-19T04:14:40 *** antimatroid has quit IRC (Ping timeout: 245 seconds) 2010-11-19T04:18:04 Top 10 players: bocsimacko(3308), wagstaff(3191), BaronTrozo(3176), bix0r4ever(3176), Raschi(3167), deccan(3143), Silver_S(3115), protocolocon(3095), davidjliu(3075), george(3069) 2010-11-19T04:21:54 *** yasith has joined #aichallenge 2010-11-19T04:33:23 *** Naktibalda has joined #aichallenge 2010-11-19T04:40:38 someone posted in the forums looking for another run with crazier maps and n>2 players 2010-11-19T04:41:50 my code could do it without too much work 2010-11-19T04:42:01 maybe not in one second though 2010-11-19T04:43:40 the biggest thing with adding more players would be taking out people that try to wait 2010-11-19T04:43:57 i expect you need to agree with the other player and send small amounts together each turn 2010-11-19T04:44:19 i don't think it would be too hard to signal that if lots of bots were doing it 2010-11-19T04:46:45 sigh.... my new code isn't magic 2010-11-19T04:46:49 ok sleep time 2010-11-19T04:46:57 *** lavalamp_ has quit IRC (Quit: Page closed) 2010-11-19T04:47:13 *** McLeopold has quit IRC (Quit: Leaving) 2010-11-19T04:51:00 Ah! Finding an actual bug instead of tweaking heuristics, that's better. 2010-11-19T05:04:31 *** yasith_ has joined #aichallenge 2010-11-19T05:05:20 *** Appleman1234 has quit IRC (Ping timeout: 245 seconds) 2010-11-19T05:06:40 *** yasith has quit IRC (Ping timeout: 240 seconds) 2010-11-19T05:12:24 *** hellman has joined #aichallenge 2010-11-19T05:18:05 Top 10 players: bocsimacko(3303), wagstaff(3253), BaronTrozo(3240), Silver_S(3134), deccan(3127), protocolocon(3107), dmj111(3080), george(3074), davidjliu(3067), shangas(3047) 2010-11-19T05:34:57 Accoun: dude.. wtf? 2010-11-19T05:41:25 I think he had some vodka 2010-11-19T05:41:28 =) 2010-11-19T05:41:38 it's not a one off 2010-11-19T05:42:20 i have the same old problem of not being able to generate a very big tree :| 2010-11-19T05:42:31 it seems to be making reasonably smart moves finally though 2010-11-19T05:43:26 *** Appleman1234 has joined #aichallenge 2010-11-19T06:14:04 *** perror has quit IRC (Ping timeout: 252 seconds) 2010-11-19T06:18:06 Top 10 players: bocsimacko(3291), BaronTrozo(3278), wagstaff(3275), Raschi(3181), bix0r4ever(3145), Silver_S(3115), protocolocon(3102), deccan(3085), george(3075), davidjliu(3069) 2010-11-19T06:18:32 *** Mistmanov has quit IRC (Quit: Page closed) 2010-11-19T06:37:40 @rankings 2010-11-19T06:37:42 antimatroid1: Top 10 players: BaronTrozo(3293), bocsimacko(3292), wagstaff(3240), Silver_S(3105), protocolocon(3101), george(3075), deccan(3073), dmj111(3069), bix0r4ever(3067), davidjliu(3063) 2010-11-19T06:37:55 look out 2010-11-19T06:41:10 fixed 2010-11-19T06:42:03 anyone know what "rufes band" actually is? 2010-11-19T07:05:03 *** mceier has joined #aichallenge 2010-11-19T07:10:48 *** Rubicon-|-Cross has joined #aichallenge 2010-11-19T07:18:05 Top 10 players: bocsimacko(3290), BaronTrozo(3245), wagstaff(3203), bix0r4ever(3173), Raschi(3112), protocolocon(3101), Silver_S(3091), george(3077), davidjliu(3057), dmj111(3049) 2010-11-19T07:31:29 *** hellman has quit IRC (Quit: buffer overflow) 2010-11-19T07:44:27 i hate debugging game trees 2010-11-19T07:44:31 it was horrible enough for tron 2010-11-19T07:44:33 this is just bleh 2010-11-19T08:18:06 Top 10 players: bocsimacko(3290), wagstaff(3227), BaronTrozo(3214), bix0r4ever(3120), Raschi(3099), protocolocon(3089), Silver_S(3087), george(3071), davidjliu(3051), shangas(3045) 2010-11-19T08:19:41 *** delt0r___ has joined #aichallenge 2010-11-19T08:20:57 *** delt0r has quit IRC (Ping timeout: 240 seconds) 2010-11-19T08:25:34 what's this huge reduction in elo rating? 2010-11-19T08:32:46 the top 10 has been a lot more competitive today 2010-11-19T08:32:49 that could just be it 2010-11-19T08:33:17 oh man, another new person in the top 10 2010-11-19T08:33:23 although they haven't had many games 2010-11-19T08:33:28 it means nothing 2010-11-19T08:33:29 a PHP bot too 2010-11-19T08:33:33 what is the world coming to 2010-11-19T08:33:55 only if you're in the top 10 after 20 games you can be greatfull :P 2010-11-19T08:33:57 :D 2010-11-19T08:34:04 'what is the world coming to' :D 2010-11-19T08:34:33 rabidus: i was offended enough at java doing so well 2010-11-19T08:34:36 let along php 2010-11-19T08:34:56 php isn't that bad.. 2010-11-19T08:35:06 never used it :P 2010-11-19T08:35:11 tried few functions.. c++ vectors vs php arrays 2010-11-19T08:35:33 php was just 60% slower, which is kinda much. but i was thinking something like 60000% 2010-11-19T08:35:46 c++ <3 2010-11-19T08:37:54 gah wtf 2010-11-19T08:38:11 i'm getting layers in my tree where it's not the player i would expect with repeating players turns 2010-11-19T08:38:46 php is ment for very other things then writing an ai bot :P 2010-11-19T08:39:02 the range of languages amazes me 2010-11-19T08:39:12 devs should be prazed for that 2010-11-19T08:39:17 yeah 2010-11-19T08:39:23 i plan to learn haskell once this is done 2010-11-19T08:40:36 @rankings 2010-11-19T08:40:37 antimatroid1: Top 10 players: bix0r4ever(3335), bocsimacko(3278), Raschi(3211), BaronTrozo(3207), wagstaff(3194), trinhanhminh1405(3106), protocolocon(3081), Silver_S(3080), george(3067), davidjliu(3053) 2010-11-19T08:40:41 he's only had 2 games though 2010-11-19T08:41:25 both raschi and bix0r resubmitted 2010-11-19T08:44:24 Error323_: yeah, one of the nice things about this challenge is no restrictions on the language used (except for them using an ancient version of ubuntu and the consequences of that) 2010-11-19T08:45:21 aerique: hopefully that will be fixed in the future 2010-11-19T08:45:32 i'm undecided how often i would like to see these contests 2010-11-19T08:45:39 this is the second one this year 2010-11-19T08:45:47 it's a huge drain on time 2010-11-19T08:46:20 i think i fixed that stupid bug :) 2010-11-19T08:46:37 minimax bot is slowly coming together properly now 2010-11-19T08:47:34 antimatroid1: definitely shouldn't be too often otherwise I'd get nothing else done 2010-11-19T08:47:56 tron was only for 3 weeks 2010-11-19T08:48:02 but this definitely needed 3 months 2010-11-19T08:48:05 antimatroid1: either once a year for three months or twice but for a shorter period 2010-11-19T08:48:20 3 weeks was a little short 2010-11-19T08:48:27 not for tron 2010-11-19T08:48:36 it was for me :) 2010-11-19T08:48:42 :) 2010-11-19T08:49:10 i have to juggle a fulltime job and a family though, i don't know what your situation is 2010-11-19T08:49:22 *** perror has joined #aichallenge 2010-11-19T08:49:26 student, easy peasy for me, i just don't do any uni work :P 2010-11-19T08:49:35 8-D 2010-11-19T08:49:41 i haven't done any actual work in months either 2010-11-19T08:49:49 i have a pretty good job where i can just work when i can be bothered 2010-11-19T08:49:59 <3 research assistant work 2010-11-19T08:50:42 it helps that the person i work for is away a lot for other research positions 2010-11-19T08:57:08 *** medrimonia1 is now known as medrimonia 2010-11-19T09:03:32 *** bhasker has joined #aichallenge 2010-11-19T09:09:40 awww, raschi's on a submission spree 2010-11-19T09:10:12 http://ai-contest.com/profile_submissions.php?user_id=10464 2010-11-19T09:17:07 *** d0ugal has joined #aichallenge 2010-11-19T09:17:45 "Wait for it... Wait for it... Wait for it... FINISH HIM" http://ai-contest.com/visualizer.php?game_id=7938884 2010-11-19T09:18:08 Top 10 players: bix0r4ever(3343), bocsimacko(3282), BaronTrozo(3200), wagstaff(3179), trinhanhminh1405(3096), protocolocon(3082), Silver_S(3081), george(3067), davidjliu(3061), luksian(3048) 2010-11-19T09:23:55 nice 2010-11-19T09:23:58 lucky, but nice :P 2010-11-19T09:26:38 <_W_> cute 2010-11-19T09:31:49 antimatroid1: What was lucky about it? :) 2010-11-19T09:33:04 he almost came back on you 2010-11-19T09:33:15 Yeah, but not really 2010-11-19T09:33:39 he probably only lost cause he tried to expand stupidly 2010-11-19T09:33:52 actually, he was pretty far behind then anyway 2010-11-19T09:33:53 hmmmm 2010-11-19T09:33:57 Yeah 2010-11-19T09:34:35 Just highlights a bug in a heuristic I have, it doesn't tkae into account the groth rate 2010-11-19T09:34:47 i remember first seeing bocsi's bot play and thinking to myself: "whew. this certainly is different from what everyone else seems to be doing." 2010-11-19T09:34:48 but when his % number of ships started to catch up, I went back into full on attack 2010-11-19T09:38:14 *** Bobng has joined #aichallenge 2010-11-19T09:43:17 *** d0ugal has left #aichallenge ("Leaving...") 2010-11-19T09:45:09 brx: i was like: "haha, he doesn't coordinate attacks" and then i lost on turn 40 2010-11-19T09:50:28 *** bhasker_ has joined #aichallenge 2010-11-19T09:50:28 *** bhasker has quit IRC (Read error: Connection reset by peer) 2010-11-19T09:55:35 lol, raschi resubmitted again 2010-11-19T09:59:37 *** Palmik has joined #aichallenge 2010-11-19T10:14:59 ... and again 2010-11-19T10:15:12 that's just getting ridiculous 2010-11-19T10:15:30 he has 111 pages of submissions 2010-11-19T10:17:32 *** bhasker_ has quit IRC (Quit: bhasker_) 2010-11-19T10:17:38 *** JamesMG has joined #aichallenge 2010-11-19T10:18:08 Top 10 players: bocsimacko(3276), wagstaff(3169), BaronTrozo(3166), trinhanhminh1405(3159), bix0r4ever(3141), protocolocon(3079), Silver_S(3073), davidjliu(3073), george(3067), luksian(3053) 2010-11-19T10:43:04 *** tim_ has joined #aichallenge 2010-11-19T10:48:00 why did the ratings drop so much again? 2010-11-19T10:48:11 and who are all these new names? ;-) 2010-11-19T10:48:36 mega1: i suspect it's from all the new names 2010-11-19T10:48:46 you guys in the top 10 seem to be beating each other about 2010-11-19T10:49:02 although you still seem to be holding them at bay 2010-11-19T10:49:23 i've decided raschi is way worse than bix0r at resubmitting, he has 111 pages of submissions 2010-11-19T10:49:38 heh 2010-11-19T10:50:03 he started with c# then went to java :\ 2010-11-19T10:50:35 I don't think I'm holding them at bay, I just have a longer history of games with a lot of wins until yesterday 2010-11-19T10:51:26 they all get lots of losses too 2010-11-19T10:51:41 but that's all they have 2010-11-19T10:51:41 i think you're just used to almost always winning :P 2010-11-19T10:51:52 I'm tempted to upload a new version 2010-11-19T10:52:07 possibly worth seeing how it does 2010-11-19T10:52:20 while it'd be nice for you to be sitting up top now, it's the final result that counts most 2010-11-19T10:53:13 (not that i need to tell you that :P) 2010-11-19T10:55:03 mega1: What's your nickname on official server? 2010-11-19T10:55:21 bocsimacko :-) 2010-11-19T10:55:50 hmmm I'm understanding some things better know 2010-11-19T10:55:54 hehe 2010-11-19T10:56:57 it's incredible how a "better" bot can have worse results on tcp 2010-11-19T10:57:17 depends whose playing when you run it i guess 2010-11-19T10:57:22 who's** 2010-11-19T10:57:52 there often does only seem to be a few people running at a time 2010-11-19T10:58:35 the top 10 is going to be very competitive this time 2010-11-19T10:58:48 *** tim_ has quit IRC (Remote host closed the connection) 2010-11-19T10:59:02 there are a number of people who haven't uploaded recently, some in ages (McLeopold) 2010-11-19T10:59:45 For me, the main issue is that every two weeks, I rewrite all my bot 2010-11-19T11:00:04 i've been rewriting mine for almost 2 weeks now :| 2010-11-19T11:00:21 I started rewriting mine last Saturday 2010-11-19T11:01:10 And I was beginning to think "that's so stupid, I can't be ranked in the top 100 with the new" 2010-11-19T11:01:26 And now... it seems to be better than the old one 2010-11-19T11:01:36 But there's still a lot of work to do 2010-11-19T11:01:57 yeah, i don't hold much hopes for top 10 anymore 2010-11-19T11:02:02 but somewhere in the top 50 would be nice 2010-11-19T11:02:12 preferably at the better end 2010-11-19T11:02:43 but since i haven't had a running bot in ages, it's hard to say how it will do 2010-11-19T11:03:24 *** aerique has quit IRC (Quit: ...) 2010-11-19T11:04:22 New upload now for me, I'll see what it can do. 2010-11-19T11:05:18 it seems to me that when I do a full rewrite it gets considerably better 2010-11-19T11:05:26 but as I sit and tweak my existing code it gets worse 2010-11-19T11:10:19 yeah, that's the problems i had 2010-11-19T11:10:26 i think it was introducing too many bugs 2010-11-19T11:10:35 i've taken my time with this final revision 2010-11-19T11:11:21 I've still got an old version of my attack part, my bot just looks like some patchwork 2010-11-19T11:12:39 Wasn't Hazard #1 before? 2010-11-19T11:13:10 maybe at some point 2010-11-19T11:13:18 mega1 has been on top for a good month or two 2010-11-19T11:13:31 one 2010-11-19T11:13:51 just one? 2010-11-19T11:13:56 seems like longer 2010-11-19T11:14:08 No I thinks it's right 2010-11-19T11:14:11 that's still a really long time to hold first though 2010-11-19T11:14:44 it wasn't before the twentieth october 2010-11-19T11:14:52 there's 1 point difference in elo up there atm 2010-11-19T11:14:52 Oct 20th 15:25:53 successfully entered into contest Lisp 2010-11-19T11:15:15 first submission was the 10th 2010-11-19T11:15:22 no, sep 19th 2010-11-19T11:15:46 Yes but at the beginning you weren't even in the top 100 right mega1? 2010-11-19T11:15:54 *** Utkarsh has joined #aichallenge 2010-11-19T11:16:36 I remember that the first time I really saw bocsimacko in the top ten, he was first... and it hasn't really changed since 2010-11-19T11:16:37 medrimonia: I still resisted in the first weeks. 2010-11-19T11:16:52 then it timed out a lot 2010-11-19T11:16:58 *** Mjothvitnir has joined #aichallenge 2010-11-19T11:17:03 Ah ok 2010-11-19T11:17:14 it was doing well on tcp before finally I managed to sort out the issues with the official server 2010-11-19T11:18:09 Top 10 players: bocsimacko(3280), bix0r4ever(3279), BaronTrozo(3164), wagstaff(3157), protocolocon(3074), Raschi(3072), davidjliu(3069), george(3059), Silver_S(3059), dmj111(3055) 2010-11-19T11:18:12 I wasn't playing on tcp before end of october that's why I hadn't seen your name before^^ 2010-11-19T11:20:41 *** RainCT has joined #aichallenge 2010-11-19T11:27:09 Do you know who is CCCF on the tcp server? 2010-11-19T11:27:56 is that felixcoto? 2010-11-19T11:28:16 that's a wild guess on my part 2010-11-19T11:29:28 I don't know 2010-11-19T11:30:59 I just hope that there won't be hundreds of good players submitting during the last few days 2010-11-19T11:31:48 people seem to think it is felixcoto 2010-11-19T11:33:54 Ok, I'll have to know the relation between the official/tcp/irc Nicknames ^^ 2010-11-19T11:36:12 *** rebelxt has joined #aichallenge 2010-11-19T11:36:21 I find it really curious how losses started to pile up suddenly. 2010-11-19T11:36:44 Can it be that the servers are running more at the same time? 2010-11-19T11:36:57 janzert: any changes on the servers? 2010-11-19T11:38:25 mega1: did it start happening after the latest map change or when the servers slowed down last night? i had a very weird losing streak too 2010-11-19T11:39:01 it's hard to tell: http://ai-contest.com/profile_games.php?user_id=8565 2010-11-19T11:39:14 first loss is: Nov 18th 21:16:48 2010-11-19T11:39:34 from my observations, you've had a few streaks of loses since the new maps came on 2010-11-19T11:39:46 but i still think it's performed better than other bots 2010-11-19T11:40:59 Nov 18th 9pm - that's when I noticed servers slowed down a lot (GPM dropped by 50%). but we are now back to normal 170 GPM: http://ai-contest.com/game_stats.php 2010-11-19T11:41:34 170 is above normal isn't it? 2010-11-19T11:41:59 GPM has been 150-160 recently with 4 servers running 2010-11-19T11:42:37 less than 3900 bots online atm 2010-11-19T11:42:54 wait no, it just doesn't show the last page of numbers anymore :| 2010-11-19T11:43:04 what, it does now :\ 2010-11-19T11:43:05 weird 2010-11-19T11:43:15 almost 4400 bots 2010-11-19T11:44:22 mega1, i can't tell if it's related to your losses but i saw both myself and bhasker resubmit last night at the same time and we both dropped from ~30-40 to 100+... slowly climbing back up now.. however my submission is heuristic so time-outs are most certainly not an issue 2010-11-19T11:44:31 mega1, smth is going on - you normally don 2010-11-19T11:44:42 don't lose games to 150+ ranked bots 2010-11-19T11:45:49 in light of the sudden drop in ratings there may indeed be a problem here 2010-11-19T11:46:29 *** narnach_mb has quit IRC (Quit: Leaving.) 2010-11-19T11:47:50 i haven't noticed any problems with games watched 2010-11-19T11:47:58 people don't seem to be timing out and games seem to finish 2010-11-19T11:48:20 i think there's quite a few new faces near the top though and it's less sure about who the leaders are 2010-11-19T11:48:28 *** Appleman1234 has quit IRC (Ping timeout: 252 seconds) 2010-11-19T11:49:00 the rankings are anchored from the bottom too yeah? 2010-11-19T11:49:28 *** McLeopold has joined #aichallenge 2010-11-19T11:50:17 for me a sudden losing streak might have been related to map change. my funneling logic seems to make wrong decisions for many new maps. more testing needed 2010-11-19T11:51:34 oh yeah, ratings might be messed up during the last week with many people submitting strong bots they've been working on on tcp for weeks... 2010-11-19T11:52:06 and off tcp 2010-11-19T11:52:27 medrimonia: up into top 20 2010-11-19T11:52:39 top 10* 2010-11-19T11:52:46 only off two games though :P 2010-11-19T11:52:53 off tcp - maybe.. but one gotta be very smart to make good improvements using local testing only.. 2010-11-19T11:53:56 Yes I've seen, I was ten with one game 2010-11-19T11:53:58 quite a few of the top people i haven't seen on tcp at all 2010-11-19T11:54:46 they could be going under other names of course 2010-11-19T11:54:52 i still have no idea who testingFromPiter is 2010-11-19T11:55:04 I've got to wait to see what it look's like after more that 24 hours 2010-11-19T11:55:05 it seems all names in current top 12-14 have been testing on tcp 2010-11-19T11:55:29 XD 5 with two games 2010-11-19T11:55:33 medrimonia: yeah it should stabilize in 24-48h 2010-11-19T11:56:29 don't start celebrating just yet - this has happened to me before as well ;-) feels good though hehe 2010-11-19T11:56:49 McLeopold: it might be good to paginate the username games page on your TCP server 2010-11-19T11:57:04 it's starting to take some time for mine to load and I can't imagine it's being nice to your bandwidth :) 2010-11-19T11:57:36 Hmm.. okay, I might just expire a bunch of game too 2010-11-19T11:58:17 ? 2010-11-19T11:58:18 Pagination would require I write a really website instead of static html. That's a big switch. 2010-11-19T11:58:55 ah. I can just rename myself instead 2010-11-19T11:59:18 If you don't mind. That's a good hack. 2010-11-19T11:59:18 probably overkill to have 2000+ games of history across tweaks anyways 2010-11-19T11:59:42 *** hellman has joined #aichallenge 2010-11-19T12:01:23 rebelxt: I know, I don't expect to be in the top 20 with this bot 2010-11-19T12:04:33 *** Utkarsh has quit IRC (Ping timeout: 240 seconds) 2010-11-19T12:05:26 *** tapwater has joined #aichallenge 2010-11-19T12:06:10 mega1: I love playing against you in tcp server, it always show me a weakness of my bot^^ 2010-11-19T12:06:31 you're welcome :-) 2010-11-19T12:11:20 *** Utkarsh has joined #aichallenge 2010-11-19T12:15:27 mega1: http://tcp.zeroviz.us/canvas.cgi?game_id=11618 2010-11-19T12:15:38 I don't actually think this is a draw map 2010-11-19T12:15:46 *** lavalamp_ has joined #aichallenge 2010-11-19T12:15:47 Which means you missed a move. :) 2010-11-19T12:16:18 @tcp 2010-11-19T12:16:19 lavalamp_: tcp = (#1) http://www.benzedrine.cx/planetwars/, or (#2) http://72.44.46.68/, or (#3) 98.247.248.39:995 (zeroviz.us), or (#4) http://tcp.zeroviz.us/ 2010-11-19T12:16:24 McLeopold: I'm not sure either. Tell me what you think is the winner. 2010-11-19T12:16:32 http://tcp.zeroviz.us/canvas.cgi?game_id=11394 2010-11-19T12:16:40 If I had continued well, maybe? 2010-11-19T12:18:08 going for that extra 19 was stupid 2010-11-19T12:18:09 Top 10 players: bix0r4ever(3288), bocsimacko(3266), Raschi(3204), BaronTrozo(3117), wagstaff(3116), murrayr(3080), protocolocon(3072), davidjliu(3057), george(3046), dmj111(3038) 2010-11-19T12:18:12 hmm, interesting 2010-11-19T12:19:19 this gives me some small hope: http://tcp.zeroviz.us/canvas.cgi?game_id=12414 2010-11-19T12:19:19 That 7 followed by 18 is cool. 2010-11-19T12:19:52 both launched same turn 2010-11-19T12:20:22 *** Utkarsh has quit IRC (Ping timeout: 252 seconds) 2010-11-19T12:20:44 that was my old bot even. 2010-11-19T12:20:45 lavalamp_: that's a weird game 2010-11-19T12:21:09 It never had a sense of being behind or ahead. 2010-11-19T12:22:01 I'm disappointed in how my bot has been doing on tcp considering how well it did in local testing :( 2010-11-19T12:23:07 hm obviously I have a bug in my antisnipe/defense code :( 2010-11-19T12:25:03 I guess that's what happens when you code all night because you can't sleep 2010-11-19T12:25:39 *** Utkarsh has joined #aichallenge 2010-11-19T12:26:40 actually, I don't think this one is either, I might add it back in. http://tcp.zeroviz.us/zeroviz_538_36.txt.html 2010-11-19T12:28:07 looks a lot like a draw map to me 2010-11-19T12:28:13 McLeopold: do you have a game that shows why it may not be a draw? 2010-11-19T12:28:32 not yet, I'm just calculating in my head, but the game should be similar to the other one 2010-11-19T12:28:34 i can't see anything worth doing other than waiting 2010-11-19T12:28:41 where I was able to hold the 17 planet. 2010-11-19T12:28:46 I think the 25 is the key 2010-11-19T12:29:15 I need a distance table... 2010-11-19T12:29:16 hmm, maybe 2010-11-19T12:29:46 *** Utkarsh has quit IRC (Ping timeout: 240 seconds) 2010-11-19T12:29:53 i don't like your chances 2010-11-19T12:30:08 http://tcp.zeroviz.us/canvas.cgi?game_id=11030 2010-11-19T12:30:34 not that smiley did the best counter... 2010-11-19T12:31:15 mega1: I read your comment backwards 2010-11-19T12:31:33 nah way, there's 1 distance difference 2010-11-19T12:31:39 If it's not a draw, then you should have done a first move 2010-11-19T12:32:14 antimatroid1: so draw or not? 2010-11-19T12:32:21 hmm, i'm not sure 2010-11-19T12:32:34 i think it's actually quite tricky now 2010-11-19T12:32:53 say you are red in the above linked game, on the second turn, blue has to decide what to sent at your expansion 2010-11-19T12:33:03 but if they send everything, you can respond the following turn by taking them out 2010-11-19T12:33:09 you can also wait to defend anything the do send 2010-11-19T12:33:36 I need the distance to the 10 neutral 2010-11-19T12:33:37 that doesn't look like a draw map to me 2010-11-19T12:34:00 mega1: thoughts on what i just said? 2010-11-19T12:34:02 If going for the 25 means you give up the 10, then it's a draw, because the best move is to wait 2010-11-19T12:34:29 i need human mode with exact ship numbers :( 2010-11-19T12:34:40 antimatroid1: that might be so 2010-11-19T12:35:01 I need more time to analyze it 2010-11-19T12:35:28 i think it would be hard to get the 10 if they expanded, but i could be wrong 2010-11-19T12:35:30 *** Utkarsh has joined #aichallenge 2010-11-19T12:36:01 theres a distance of 14 to the 10 neutral from home 2010-11-19T12:36:26 http://pastebin.com/ZYCbkGQA 2010-11-19T12:36:32 wait, i think it's 15 sorry 2010-11-19T12:36:49 no, i can't count 2010-11-19T12:36:57 it was 14, 13 to your other neutral 2010-11-19T12:37:39 that must be really confusing 2010-11-19T12:37:40 from home, it's 13 distance to your neutral with 25 ships, 14 distance to the neutral with 10, and 14 to the other persons neutral with 25 2010-11-19T12:39:18 *** rebelxt has left #aichallenge 2010-11-19T12:40:05 I'm gonna tweak AntBot to play that map 2010-11-19T12:40:19 I think it is not a draw. :) 2010-11-19T12:41:31 It's very close to r/p/s :) 2010-11-19T12:42:29 *** Utkarsh has quit IRC (Ping timeout: 255 seconds) 2010-11-19T12:46:09 i don't know what to do after going for the 25 if the other one does the same 2010-11-19T12:46:14 but i think that's the best first move 2010-11-19T12:47:09 I'll try my nash bot on that map 2010-11-19T12:52:06 Hmm, if you go for the 25, maybe the opponent get's the 30? 2010-11-19T12:53:28 *** wh1teside_ has joined #aichallenge 2010-11-19T12:54:47 *** wh1teside has quit IRC (Ping timeout: 276 seconds) 2010-11-19T12:58:17 yeah, that might be right 2010-11-19T12:58:58 i don't think it's a trivial draw anyway, but quite a bots would probably think it is 2010-11-19T13:02:15 I have a stupidly simple idea to solve that map. 2010-11-19T13:02:29 If it works, it invalidates 1000's of lines of code. 2010-11-19T13:02:35 those are sometimes the best 2010-11-19T13:02:36 :) 2010-11-19T13:02:37 And I'm going to celebrate 2010-11-19T13:03:13 Sigh...new bot comming up... 2010-11-19T13:08:09 McLeopold: do you have the map id? 2010-11-19T13:08:30 zeroviz_538_36.txt 2010-11-19T13:09:51 that may be a draw against really good bots, but it should be not be tossed out 2010-11-19T13:10:08 it's still a good test of strength 2010-11-19T13:11:03 all maps are draw maps 2010-11-19T13:11:09 the question is what's the strategy to give it 2010-11-19T13:11:19 (well, including mixed strategies) 2010-11-19T13:11:51 @rankings 2010-11-19T13:11:52 antimatroid1: Top 10 players: bix0r4ever(3313), bocsimacko(3264), murrayr(3187), Raschi(3186), wagstaff(3115), BaronTrozo(3111), protocolocon(3070), dmj111(3066), george(3050), davidjliu(3048) 2010-11-19T13:12:50 i'm not sure i could bear it if java won 2010-11-19T13:13:39 java: the superior AI language 2010-11-19T13:13:58 my nash bot also just sits there on the that map 2010-11-19T13:13:59 ahaha 2010-11-19T13:14:19 i'm going to try human mode on it 2010-11-19T13:14:25 not that it's a particularly good bot 2010-11-19T13:14:37 lisp, javascript or python: anyone else and I'll start handing out secrets :) 2010-11-19T13:14:59 * mega1 holds his new submission back 2010-11-19T13:15:07 you'd take jscript over C variants or Go? 2010-11-19T13:15:19 having my doubts about you now McLeopold 2010-11-19T13:15:28 c? mah 2010-11-19T13:16:04 McLeopold seems to have a good taste 2010-11-19T13:16:25 didn't the go founder say he wouldn't have done it if he had known about something else? 2010-11-19T13:16:59 or do I have that backwards? 2010-11-19T13:17:17 *** smellyhippy has quit IRC (Ping timeout: 255 seconds) 2010-11-19T13:17:25 McLeopold: I know the Groovy founder said that about Clojure 2010-11-19T13:17:37 oh, yeah, that's what I'm thinking of 2010-11-19T13:18:10 Top 10 players: bix0r4ever(3313), bocsimacko(3264), murrayr(3187), Raschi(3186), wagstaff(3115), BaronTrozo(3111), protocolocon(3070), dmj111(3066), george(3050), davidjliu(3048) 2010-11-19T13:18:36 ugh, my bot is so broken right now I can't stand to run it 2010-11-19T13:18:38 Clojure and Jython are in too. :) 2010-11-19T13:19:08 Maybe haskell, but I still don't get it. 2010-11-19T13:19:16 I'm gonna write a bash bot 2010-11-19T13:19:22 just to make the ugliest thing possible 2010-11-19T13:19:57 btw, I've started a final tournament transition checklist. Information on it at http://ai-contest.com/forum/viewtopic.php?f=20&t=1131 2010-11-19T13:20:03 for i in `seq $planet_count` 2010-11-19T13:20:34 JamesMG: you want ugly, write perl (BUUURN) 2010-11-19T13:20:43 haha 2010-11-19T13:21:36 time to rewrite all my order generation code 2010-11-19T13:22:06 *** smellyhippy has joined #aichallenge 2010-11-19T13:22:06 *** smellyhippy has joined #aichallenge 2010-11-19T13:22:19 janzert: with the way it's keeping old peoples ranking with new submissions 2010-11-19T13:22:23 will it clear that properly? 2010-11-19T13:22:56 yes, that is handled by the step to run the ranking code twice after archiving all current games 2010-11-19T13:23:05 :) just making sure 2010-11-19T13:23:12 yep :) 2010-11-19T13:23:33 i do think it's important that the leaderboard stays live 2010-11-19T13:24:12 you'll give accoun one less thing to call cheating on too 2010-11-19T13:24:16 yeah, that's my current plan although I haven't discussed any of it with Jeff yet 2010-11-19T13:24:44 hehe, not really too worried about that since I'm sure he'll find something 2010-11-19T13:25:44 i think i'd probably be in favour of keeping any draw maps too, as i don't think any will be as bad as some found with tron 2010-11-19T13:26:33 and players can't then just assume they wont be given a map on which they need to do nothing 2010-11-19T13:26:42 yeah, given that the current generator seems to have a low probability of generating a trivial draw now it will probably be fine 2010-11-19T13:27:13 removing maps (after start) changes # of games played, (cheaters) replaying those opponents on new map looks bad also (cheaters kept replaying games till their favorite won) 2010-11-19T13:27:47 :) 2010-11-19T13:28:03 oh, one mor thing, it would be good to finalise how the tournament is to be run 2010-11-19T13:28:13 like what top x bots will get a round robin at the end 2010-11-19T13:28:17 nah, at least now removing means just no new games will be played on it 2010-11-19T13:28:25 the old games will still count 2010-11-19T13:28:41 antimatroid1: I'm hoping for none 2010-11-19T13:29:03 but it depends on how good we can make the rankings while running the full tournament 2010-11-19T13:29:04 see, that is what it takes, remove the "complaint" before it happens 2010-11-19T13:30:08 janzert: the quickest way to find draw maps seems to be looking for 200 turns with no moves 2010-11-19T13:30:11 i.e. my hope is that we can get good rankings top to bottom without resorting to eliminating bottom ranks 2010-11-19T13:30:41 if the number look obviously bad, then it is probably a draw. 2010-11-19T13:30:53 i'm still not sure about removing them for this 2010-11-19T13:30:54 If you have to think about it or calculate, it should probably stay in 2010-11-19T13:30:58 will non-starters be removed or will they kep playing? 2010-11-19T13:31:55 McLeopold: the way I'm going to monitor it is just by watching percentage of draws among top players, but if there are only one or two I'll probably just leave it alone anyway 2010-11-19T13:32:09 yes, that will work 2010-11-19T13:32:10 you could fire a _couple_ of simple maps at all bots and those that don't answer "go" don't play 2010-11-19T13:32:27 but by "watching" you should target 200 turns with no moves first 2010-11-19T13:33:27 Mjothvitnir: I'd rather not eliminate anyone that we allowed in during this phase 2010-11-19T13:33:52 and bots that crash on the first turn have really quick games anyway 2010-11-19T13:33:56 if they don't respond with a go, the won't take up too much cpu 2010-11-19T13:34:01 janzert: has anything been changed on the servers in the last 24 hours? 2010-11-19T13:34:17 you mean the change in elo range? 2010-11-19T13:34:18 yes 2010-11-19T13:34:19 but, you may want to consider a bot having played recently only if there was more than 1 turn 2010-11-19T13:34:51 janzert: I'm suddenly losing way more and wondering if it's the opponent or something else too 2010-11-19T13:34:59 true, for the final, but I think that that should have been a test along with "compile failure" all along. 2010-11-19T13:35:10 mega1: nothing that should effect that no 2010-11-19T13:35:34 no more games played at the same time for instance? 2010-11-19T13:35:38 nope 2010-11-19T13:35:44 hmm 2010-11-19T13:35:50 the server games in last 24 hours would be much more meaning full :) 2010-11-19T13:36:21 *** Mistmanov has joined #aichallenge 2010-11-19T13:36:48 Mjothvitnir: yeah, eliminating immediate crash bots before entering them into the competition is something that should definitely be there for the next one 2010-11-19T13:37:38 mega1: are the loses on line symmetry? 2010-11-19T13:37:41 i actually saw some code to handle that in one of the php submit files 2010-11-19T13:37:47 left over from an old contest 2010-11-19T13:38:12 McLeopold: the line symmetric maps have been there for quite some time 2010-11-19T13:38:25 but yeah, we don't have tests and we should have had them 2010-11-19T13:38:27 so they should not be a bigger problem suddenly 2010-11-19T13:38:33 hmm 2010-11-19T13:38:43 i think he competition just got a lot better 2010-11-19T13:38:51 in the last day I've added some additional restrictions on how a map is chosen, tweaked the visualizer, eliminated a database join when handing out matches to workers and started the final tournament checklist 2010-11-19T13:39:04 so nothing that should be able to effect game outcomes 2010-11-19T13:39:17 besides the new maps 2010-11-19T13:39:19 janzert: what are the restrictions on the map selection? 2010-11-19T13:39:32 Zannick: new maps in the last 24 hours? 2010-11-19T13:40:00 mega1: you beat bix0r's new entry at least :) 2010-11-19T13:40:25 mega1: hm, no 2010-11-19T13:41:40 mega1: the map must be enabled in the database and highest priority that the current pairing hasn't played and from that set one that minimizes the combined count of plays by each player 2010-11-19T13:42:22 the part added in the past day was minimizing the plays by each player 2010-11-19T13:43:08 do enabled maps have different priorities? 2010-11-19T13:43:15 they can yes 2010-11-19T13:43:28 why? when? how? 2010-11-19T13:43:49 currently the last batch of maps is at priority 1 and the ones before that at the 2nd priority 2010-11-19T13:44:17 of course there almost certainly won't be enough games played that the 2nd priority maps will get used at all 2010-11-19T13:44:39 I just didn't completely disable them after adding the most recent maps is all 2010-11-19T13:46:30 *** bhasker has joined #aichallenge 2010-11-19T13:46:43 is he like trying to game the elo rating system? 2010-11-19T13:46:50 just uploading every time he drops below a # 2010-11-19T13:47:48 priority 2 is the radius 12 then? 2010-11-19T13:48:02 yeah, the r528 maps 2010-11-19T13:49:00 janzert: that's a good map choosing method, I like it 2010-11-19T13:49:10 but if I'm thinking about it right it would take over a month of games at the current rate before they would be used 2010-11-19T13:49:42 maybe a little less 2010-11-19T13:49:58 nah, maybe more :) 2010-11-19T13:50:14 basically, a pairing must play all priority 1 maps before they can do 2 2010-11-19T13:50:20 yep 2010-11-19T13:50:39 what happens when the have played all maps? 2010-11-19T13:51:05 it will repeat and start spitting out warnings in the log 2010-11-19T13:51:08 the tournament ends and we declare a winner? 2010-11-19T13:52:47 so, instead of hasn't play, it's played the least? 2010-11-19T13:54:11 mega1: could you humor me and toss your bot on http://tcp2.zeroviz.us/ 2010-11-19T13:54:15 for a little? 2010-11-19T13:56:59 i'll chuck mine on, but the only thing it's doing is attacking and defending 2010-11-19T13:57:06 and it's still not doing that for me entirely right 2010-11-19T13:57:35 McLeopold: in a sec 2010-11-19T13:58:00 what else is there? 2010-11-19T13:58:23 expand 2010-11-19T13:58:31 resitribute 2010-11-19T13:58:35 redistribute* 2010-11-19T13:58:39 redirect 2010-11-19T13:58:43 oh, funneling? 2010-11-19T13:59:17 whatever people call it 2010-11-19T14:00:20 ship propogation 2010-11-19T14:00:28 leveraging ship populations 2010-11-19T14:00:44 RageBot fails on my crazy maps. :( 2010-11-19T14:00:46 doesn't seem to be working 2010-11-19T14:00:52 port 996? 2010-11-19T14:00:56 yeah 2010-11-19T14:01:05 i'm probably just being retarded 2010-11-19T14:01:14 http://tcp2.zeroviz.us/ am i just using that as the host? 2010-11-19T14:01:25 oh, no 2010-11-19T14:01:34 well... 2010-11-19T14:01:40 it's all the same ip address 2010-11-19T14:01:47 let me try 2010-11-19T14:02:52 can't connect 2010-11-19T14:03:12 maybe it's not me being retarded for once 2010-11-19T14:04:19 using the same ip as for the normal with port 996 2010-11-19T14:05:18 okay, try again 2010-11-19T14:05:43 (no wonder there were no connections...) 2010-11-19T14:07:02 still doesn't seem to like me 2010-11-19T14:07:21 hmm.. I can get in 2010-11-19T14:07:28 zeroviz.us:996? 2010-11-19T14:07:28 McLeopold: yes, it's really played the least and if least is larger than 0 log a warning 2010-11-19T14:07:40 yeah 2010-11-19T14:07:49 *** tim_ has joined #aichallenge 2010-11-19T14:08:03 nope,i'm on 2010-11-19T14:08:15 Go Ant! 2010-11-19T14:08:30 caned me 2010-11-19T14:09:11 \o/ 2010-11-19T14:09:14 http://72.44.46.68/canvas?game_id=1095852 2010-11-19T14:09:40 so now that dhartmei's server is also running the official maps what is different about your server mcleopold? 2010-11-19T14:09:57 it's got cooler maps 2010-11-19T14:09:57 Finding and analyzing map types. 2010-11-19T14:10:14 ah k 2010-11-19T14:12:53 so what caused the sudden drop in elo across the board on official servers? 2010-11-19T14:13:01 now the highest is 3200ish 2010-11-19T14:13:05 from like 3800? 2010-11-19T14:13:32 it's a very boring game between me and rage bot on a map with one home planet atm 2010-11-19T14:13:56 mcleopold: you really shouldn't have lost to me :P 2010-11-19T14:14:14 oh, it was your old submit, that's alright 2010-11-19T14:15:53 *** RainCT has quit IRC (Ping timeout: 276 seconds) 2010-11-19T14:16:07 *** janzert has quit IRC (Ping timeout: 260 seconds) 2010-11-19T14:17:00 lol, it really does a bad job when there's no checks about what to send 2010-11-19T14:17:11 not surprising really 2010-11-19T14:18:11 Top 10 players: bocsimacko(3292), bix0r4ever(3269), murrayr(3232), wagstaff(3155), BaronTrozo(3128), medrimonia(3126), dmj111(3106), protocolocon(3105), george(3079), davidjliu(3075) 2010-11-19T14:18:42 *** janzert has joined #aichallenge 2010-11-19T14:18:57 @tcp 2010-11-19T14:18:58 bhasker: tcp = (#1) http://www.benzedrine.cx/planetwars/, or (#2) http://72.44.46.68/, or (#3) 98.247.248.39:995 (zeroviz.us), or (#4) http://tcp.zeroviz.us/ 2010-11-19T14:19:10 what is your bot name antimatroid? 2010-11-19T14:19:33 contestbot learn tcp as http://tcp2.zeroviz.us/ (crazy!) 2010-11-19T14:19:33 McLeopold: As you wish. 2010-11-19T14:20:11 bhasker: the rating changes are caused by having bayeselo calculate and use the correct weight for draws instead of it's default which is calculated for chess 2010-11-19T14:20:40 bhasker: antimatroid on tcp2, but it's literally only picking out offense/defense moves there, no expansion,redistribution,redirecting etc. 2010-11-19T14:22:34 i'm still trying to sort the first things out before i'm willing to let it do the rest 2010-11-19T14:22:51 *** Blkt has joined #aichallenge 2010-11-19T14:23:26 good evening everyone 2010-11-19T14:24:38 ah k 2010-11-19T14:24:55 *** RainCT has joined #aichallenge 2010-11-19T14:25:07 so for tcp2 whats the server address/port? 2010-11-19T14:25:18 zerovis.us:996 2010-11-19T14:25:18 tcp2.zeroviz.us 995? 2010-11-19T14:25:20 ah k 2010-11-19T14:25:24 996 2010-11-19T14:26:22 it's doing a pretty poor job of just attacking atm anyway 2010-11-19T14:26:25 *** RainCT has quit IRC (Remote host closed the connection) 2010-11-19T14:26:34 dang, I can get a lead on mega1, but I can't hold it 2010-11-19T14:27:09 nem is szervezo 2010-11-19T14:27:12 hanem versenyzo 2010-11-19T14:27:17 hmm i get permission denied 2010-11-19T14:27:18 ah, sorry 2010-11-19T14:27:22 wrong window 2010-11-19T14:27:26 haha 2010-11-19T14:27:31 i was like wtf 2010-11-19T14:28:16 nm works now 2010-11-19T14:28:51 first game wiped by mega1 in 40 turns 2010-11-19T14:28:56 36 :-\ 2010-11-19T14:29:19 these games are short 2010-11-19T14:29:44 yeah, they start more developed 2010-11-19T14:29:59 lol freaking hell 2010-11-19T14:30:04 my bot screwed up bigtime 2010-11-19T14:30:05 antimatroid1: are current one is awesome! 2010-11-19T14:30:17 for me... 2010-11-19T14:30:32 *** hellman has quit IRC (Remote host closed the connection) 2010-11-19T14:30:54 I want to produce some fast growth maps 2010-11-19T14:31:05 are current on is awesome what? 2010-11-19T14:31:08 200 ships to start, 30 planets, ship count < 30 2010-11-19T14:31:18 yeah sure 2010-11-19T14:31:38 http://tcp2.zeroviz.us/canvas.cgi?game_id=630 2010-11-19T14:31:44 lol 2010-11-19T14:31:51 my bot isn't even considering what you will sent at it, so the first move it's like sending everything off to just a couple of planets and being wiped 2010-11-19T14:32:02 *** p4p4 has joined #aichallenge 2010-11-19T14:32:06 llike that 2010-11-19T14:32:25 everything to the corner :D 2010-11-19T14:33:16 ah well, it's 6 30am 2010-11-19T14:33:17 time for bed 2010-11-19T14:33:25 mega1 stop wiping my bot 2010-11-19T14:33:27 !@#!@3 2010-11-19T14:35:55 sigh getting wiped in 35 turns is no fun 2010-11-19T14:36:49 *** wh1teside__ has joined #aichallenge 2010-11-19T14:37:33 21 turns 2010-11-19T14:37:36 this is depressing 2010-11-19T14:38:01 :) 2010-11-19T14:39:17 well i can still beat antimatroid:p 2010-11-19T14:39:41 thats not much to brag about atm :P 2010-11-19T14:39:52 it's like beating rage bot aha 2010-11-19T14:40:02 yeah, bhasker, we have a match 2010-11-19T14:40:03 slightly harder maybe 2010-11-19T14:40:24 :p 2010-11-19T14:40:37 *** wh1teside_ has quit IRC (Ping timeout: 260 seconds) 2010-11-19T14:40:41 hmm, we are both conservative 2010-11-19T14:40:42 lost game 2010-11-19T14:40:44 bhasker: if you want to feel better, look at http://zeroviz.us:8080/JamesMG.dev.2.html 2010-11-19T14:41:05 k james that does make me feel better:) 2010-11-19T14:41:06 JamesMG: you can lose the 8080 now and go with tcp.zeroviz.us 2010-11-19T14:41:07 I'm on the border of reverting back to Nov 8 because of this :) 2010-11-19T14:41:13 ok McLeopold 2010-11-19T14:41:32 interesting close match i think 2010-11-19T14:41:37 the draw? 2010-11-19T14:41:40 almost even 2010-11-19T14:41:48 no no the match my bot is having against antbot 2010-11-19T14:41:51 oh ok 2010-11-19T14:42:49 hmmm my new bot seems to be quite better than the old one 2010-11-19T14:43:42 But I haven't enough games to know really how much it's gonna be ranked. 2010-11-19T14:43:45 it still loses even when it doesn't need to expand aha, bah i'll look at it when i get up 2010-11-19T14:43:47 aww lost it 2010-11-19T14:43:47 :( 2010-11-19T14:44:01 oh vicotyr 2010-11-19T14:45:15 my bot gets locked down still :-\ 2010-11-19T14:45:17 doesn't even defend 2010-11-19T14:45:48 okay, crazy_fast maps are in 2010-11-19T14:46:51 bhasker got the first crazy_fast 2010-11-19T14:46:51 who is "mb" now? 2010-11-19T14:47:00 and lost 2010-11-19T14:47:27 what's crazy fast mean? 2010-11-19T14:47:36 http://tcp2.zeroviz.us/canvas.cgi?game_id=668 2010-11-19T14:48:18 growth doesn't seem to be above 5? 2010-11-19T14:48:32 no, the visualizer doesn't support than...yet 2010-11-19T14:49:00 will there be planets with growth >5? 2010-11-19T14:49:33 If I update the visualizer, yes 2010-11-19T14:49:54 would that break your code? 2010-11-19T14:49:59 *** rebelxt has joined #aichallenge 2010-11-19T14:50:13 hmm might break my evaluation function 2010-11-19T14:50:20 i should change that piece anyway 2010-11-19T14:50:50 http://tcp2.zeroviz.us/canvas.cgi?game_id=670 2010-11-19T14:50:53 I changed that in my code so > +5 planets don't break me when they did the map change 2010-11-19T14:50:59 bhasker, you're getting sniped badly :( 2010-11-19T14:51:08 yea:-\ 2010-11-19T14:51:17 *** Accoun has quit IRC () 2010-11-19T14:51:39 bhasker: haha, our bots have been climbing back up pretty close to each other since last night's resubmission ;-) 2010-11-19T14:51:46 yea 2010-11-19T14:52:00 to test it I made maps that started players off with 1000 ships and the planets had growth rates up to 100 I think 2010-11-19T14:52:07 was it 0.23 you submitted? 2010-11-19T14:52:17 pretty much with one bug fixed 2010-11-19T14:52:25 my bot is too conservative 2010-11-19T14:52:40 lavalamp_: check out my map generator, you can do that with command line options 2010-11-19T14:52:41 cool, it should do well 2010-11-19T14:52:44 i am considering removing the concept of available ships altogether 2010-11-19T14:52:54 hmm, it tries to take enemies the same time it takes a neutral too 2010-11-19T14:52:56 and just launch as long as it improves my position 2010-11-19T14:53:17 McLeopold: cool 2010-11-19T14:53:20 bhasker: don't think you can remove it completely.. but interesting idea.. 2010-11-19T14:53:38 except i need an evaluation function thats very accurate 2010-11-19T14:53:49 my current one sucks:-\ 2010-11-19T14:54:17 python McMaps.py -n crazy_fast -c 50 -p 31 31 -d 2 34 -r 10 -g 2 5 -s 1 30 -b 1 1 -x 24.0 -t radial --home 200 5 2010-11-19T14:54:19 bhasker: my most reckless bot does best too.. if i take scheduled fleets into account when calculating available ships, i don't do well heh 2010-11-19T14:54:28 yea 2010-11-19T14:54:30 :-\ 2010-11-19T14:55:22 let me try a modification 2010-11-19T14:55:34 bhasker: i haven't got the code running for that atm, take a look at some of the games 2010-11-19T14:55:43 it repeatedly sends out ships when it shouldn't 2010-11-19T14:55:51 it's also sending fleets all the way across the map 2010-11-19T14:56:01 my bot? 2010-11-19T14:56:02 or yours? 2010-11-19T14:56:04 no mine 2010-11-19T14:56:06 oh k 2010-11-19T14:56:12 ouch mega1: http://tcp2.zeroviz.us/canvas.cgi?game_id=683 2010-11-19T14:57:29 *** dvladim has joined #aichallenge 2010-11-19T14:57:53 i think mega1's evaluation function could use a code read from us. he should put it on pastebin so that we can help debug ;-)) 2010-11-19T14:59:04 it'd be interesting to see how well an open source bot could play after the contest 2010-11-19T14:59:25 *** dvladim has left #aichallenge 2010-11-19T15:03:14 *** Accoun has joined #aichallenge 2010-11-19T15:04:18 i'd be interested to get my Monte Carlo bot to completion.. it might do reasonably well given longer turn time and semi-heavy playouts 2010-11-19T15:05:16 my move generator uses monte carlo 2010-11-19T15:06:14 antimatroid1: you're really falling down on these crazy_fast maps 2010-11-19T15:06:18 http://tcp2.zeroviz.us/canvas.cgi?game_id=715 2010-11-19T15:06:38 lavalamp_: but then you are using heuristic eval function to calculate value of each move, correct? 2010-11-19T15:07:00 correct 2010-11-19T15:07:14 well, value of each "playout" 2010-11-19T15:07:28 lavalamp_: i was doing "light" playouts (i.e. rage bot style, for instance) simulating moves from each player until the end of the game or until winner is clear 2010-11-19T15:07:46 I have another heuristic that weights the order I search in 2010-11-19T15:07:58 rebelxt, yeah I do super heavy playouts 2010-11-19T15:08:15 *** yasith_ has quit IRC (Ping timeout: 245 seconds) 2010-11-19T15:08:52 lavalamp_: interesting, are you actually simulating multiple turn playouts (a number of turns for both players)? 2010-11-19T15:09:50 no, in theory I then switch to a payoff table to see which of the generated moves is best 2010-11-19T15:10:16 the monte carlo stuff is used just to find the best order to execute goals in 2010-11-19T15:10:45 considering only a single player 2010-11-19T15:11:28 I have high hopes for this stupid bot if I can get it working and fast enough for a depth 2 search 2010-11-19T15:11:28 ok, i see. mine is different - i tried simulating games till the end using different "light" move generators. kinda like what MC Go programs do 2010-11-19T15:11:55 yeah, I'm sorta treating MC my own way :) 2010-11-19T15:12:10 but even having one minute per turn is not enough to produce good results with light playouts :( 2010-11-19T15:12:24 the thing is, in this game random "light" moves are NOT even close to being good 2010-11-19T15:12:37 whereas in go the mistakes kinda cancel out 2010-11-19T15:14:00 I didn't lose against mega1!: http://tcp2.zeroviz.us/canvas.cgi?game_id=678 2010-11-19T15:14:03 correct, random light moves lead to completely random playouts.. so high number of playouts does not help 2010-11-19T15:14:23 oh crap, i might have won if it went longer 2010-11-19T15:14:42 McLeopold: lol at the map 2010-11-19T15:14:43 yeah that's why I have such a heavy goal execution... :/ 2010-11-19T15:14:58 sometimes it works: http://72.44.46.68/canvas?game_id=1095846 2010-11-19T15:16:31 McLeopold: nah he must have a horizon effect near the end 2010-11-19T15:16:37 maybe 2010-11-19T15:16:40 I know my bot does 2010-11-19T15:16:56 suddenly it sees that your ships will never land ^^ 2010-11-19T15:17:24 http://tcp2.zeroviz.us/canvas.cgi?game_id=758 2010-11-19T15:17:26 VICTORY 2010-11-19T15:18:02 nice 2010-11-19T15:18:11 I almost got him on that map too 2010-11-19T15:18:12 Top 10 players: bix0r4ever(3312), bocsimacko(3293), Raschi(3226), wagstaff(3155), murrayr(3126), medrimonia(3103), BaronTrozo(3103), dmj111(3101), protocolocon(3095), george(3074) 2010-11-19T15:18:21 McLeopold: are some of these crazy maps your local test maps? 2010-11-19T15:18:25 hehe:D 2010-11-19T15:18:35 no, those were all hand made 2010-11-19T15:19:43 McLeopold: that's what i was wondering.. i have some super simple test maps with 3-6 planets for testing sniping, close starts, etc.. could be useful on the tcp2 2010-11-19T15:19:59 true 2010-11-19T15:20:02 its depressing that at start of tournament one could hope to beat the top bots and there were lot of good ones 2010-11-19T15:20:06 now it looks like there is one good one 2010-11-19T15:20:09 and many average ones:( 2010-11-19T15:20:42 but, I think it's only useful locally 2010-11-19T15:20:43 hey, first page is way above average. well, 50% of submissions are above avg! 2010-11-19T15:20:58 lol 2010-11-19T15:21:02 I created a schedule bot that would do the perfect starts for my hand made maps. 2010-11-19T15:21:05 McLeopold: yeah 2010-11-19T15:21:10 *** yasith_ has joined #aichallenge 2010-11-19T15:22:16 McLeopold: that's a good idea.. i only make sure that my latest bot never loses on hand-made test maps to old versions 2010-11-19T15:22:23 *** wh1teside_ has joined #aichallenge 2010-11-19T15:23:15 bhasker: I think that the good thing is that the old #1 bot are now not even in the first page! (for those who haven't resubmit 2010-11-19T15:23:33 true 2010-11-19T15:25:13 bhasker: you doing invalid orders on tcp2 2010-11-19T15:25:19 yea i know :-\ 2010-11-19T15:25:23 this version has bugs 2010-11-19T15:25:27 need to fix that 2010-11-19T15:25:41 its somehow assuming more ships to be available on some planets 2010-11-19T15:25:47 than total left on it 2010-11-19T15:25:50 have to track that down 2010-11-19T15:25:56 the very best and quickest way to get good elo on tcp: http://72.44.46.68/getplayer?player=semyglass ;-) 2010-11-19T15:26:19 *** wh1teside__ has quit IRC (Ping timeout: 265 seconds) 2010-11-19T15:26:34 lol 2010-11-19T15:27:04 rebelxt: My log looks like the same when I'm fixing my bot^^ 2010-11-19T15:27:46 lol, are you editing the same file you are playing on tcp? 2010-11-19T15:30:39 bhasker: interesting game now 2010-11-19T15:31:00 i do that all the time rebel:D 2010-11-19T15:31:15 i watching my eval function result 2010-11-19T15:31:19 looks like i am winning! 2010-11-19T15:31:30 2010-11-19 12:31:24,043-v0.1-INFO- Turn:105 updateStateObj() - My Value of current state:186.674878 2010-11-19T15:31:30 2010-11-19 12:31:24,043-v0.1-INFO- Turn:105 updateStateObj() - Enemy Value of current state:50.405556 2010-11-19T15:31:39 hurray 2010-11-19T15:31:47 I think I see a bug 2010-11-19T15:31:53 in mine? 2010-11-19T15:32:02 no, I'm the one that lost 2010-11-19T15:32:13 ah this version is aggressive 2010-11-19T15:32:21 it doesn't do available ship calculations 2010-11-19T15:32:24 properly 2010-11-19T15:32:35 so it will attack more often 2010-11-19T15:32:41 though might end up losing because of that too 2010-11-19T15:32:45 by sending ships the wrong way 2010-11-19T15:34:26 its launching all ships from 5 growth planets to 3 growth planets 2010-11-19T15:34:30 and losing the 5 growht 2010-11-19T15:34:36 i am amazed it actually managed to win 2010-11-19T15:35:21 http://tcp2.zeroviz.us/canvas.cgi?game_id=799 2010-11-19T15:35:23 interesting game 2010-11-19T15:35:44 *** Naktibalda has quit IRC (Quit: ChatZilla 0.9.86 [Firefox 3.6.8/20100722155716]) 2010-11-19T15:36:39 what makes that one interesting? 2010-11-19T15:37:03 you can find lots of mistakes to analyze 2010-11-19T15:37:15 there is so much going on I'm not sure where to look 2010-11-19T15:37:25 yeah 2010-11-19T15:38:02 its just funny to watch, i just launch all over 2010-11-19T15:38:06 and no idea why 2010-11-19T15:38:08 :-\ 2010-11-19T15:38:25 :) ok 2010-11-19T15:42:01 reckless bot does better 2010-11-19T15:42:12 rule to calculate available ships 2010-11-19T15:42:21 if not underAttack: all ships are available 2010-11-19T15:42:26 else do a proper calculations 2010-11-19T15:42:40 it does stupid stuff sometimes 2010-11-19T15:42:40 *** smloh1 has joined #aichallenge 2010-11-19T15:42:45 but overall seems to do better 2010-11-19T15:42:49 than a conservative version 2010-11-19T15:44:16 *** p4p4 has quit IRC (Remote host closed the connection) 2010-11-19T15:44:48 rebelxt http://72.44.46.68/canvas?game_id=1097026 2010-11-19T15:44:52 reckless version in action 2010-11-19T15:46:34 my bot gets creamed so bad by you guys that i'm not sure it is worth running on the tcp servers 2010-11-19T15:46:49 which one is your bot? 2010-11-19T15:48:13 none (tcp) at the moment. (official 550-700ish) 2010-11-19T15:48:39 I need to rewrite it, but I find that depressing 2010-11-19T15:48:46 ah k 2010-11-19T15:53:29 *** smloh1 has quit IRC (Quit: Page closed) 2010-11-19T15:54:16 pound-pound-pound 2010-11-19T15:56:01 wow 2010-11-19T15:59:44 well, one win for a rotten-stump brain isn't bad :) 2010-11-19T16:02:30 *** aerique has joined #aichallenge 2010-11-19T16:03:28 2/14/1 2010-11-19T16:07:18 *** Naktibalda has joined #aichallenge 2010-11-19T16:08:46 *** McLeopold has quit IRC (Ping timeout: 240 seconds) 2010-11-19T16:09:54 hey guys, I'm using JBotManager. My bot (rather simple, started yesterday) does work well but for tcp games I always loose after a few turns, I guess due to timeouts. However, my bot seems to be computing pretty fast in "on computer combat" mode. any ideas? 2010-11-19T16:10:45 which tcp server are you using? 2010-11-19T16:11:23 or alternatively... what name does your bot have? 2010-11-19T16:11:38 http://72.44.46.68/getratings are you in that list? 2010-11-19T16:12:12 let me check 2010-11-19T16:13:10 72.44.46.68:995, but I'm not in the list 2010-11-19T16:13:32 name of bot is "python schummelbot.py" if I understand it correctly ;-) 2010-11-19T16:14:23 tcp server port _username_ -p pass command 2010-11-19T16:14:42 ok, i didnt put anything there :) 2010-11-19T16:14:57 do I have to sign up somewhere for that? 2010-11-19T16:15:03 nope 2010-11-19T16:15:19 so first connect sets the acc + pw? 2010-11-19T16:15:27 in the manager, in the tcp tab, at the top, you can see the account name and pw 2010-11-19T16:15:35 basically 2010-11-19T16:15:43 you should fill some random stuff in there 2010-11-19T16:17:34 jbot manager rembers so it really can be random 2010-11-19T16:17:54 ok. i did that and in the last 2 round I didnt have a single timeout (using the same code) 2010-11-19T16:17:59 weird, but ok 2010-11-19T16:18:02 :) 2010-11-19T16:18:12 Top 10 players: bix0r4ever(3312), bocsimacko(3291), wagstaff(3141), murrayr(3130), BaronTrozo(3101), medrimonia(3094), protocolocon(3093), dmj111(3087), davidjliu(3072), george(3070) 2010-11-19T16:18:17 there get cleared out every so often so you can make a new name with a new version of your code 2010-11-19T16:18:31 howdy 2010-11-19T16:18:37 yo 2010-11-19T16:18:40 hi =) 2010-11-19T16:18:46 hey guys 2010-11-19T16:18:51 Mistmanov: stop developing 2010-11-19T16:18:54 ^)^ 2010-11-19T16:19:10 eh, no matter how much I develop, my bot refuses to get any better :p 2010-11-19T16:19:23 hahah, yea i kind of got the same 2010-11-19T16:19:31 although changing to the new maps imrproved my bot's winrate from 30-70 to 50-50.. 2010-11-19T16:19:40 so that's pretty andy 2010-11-19T16:19:49 2/8/0 on regular zeroviz, even a rotten stump can win sometimes 2010-11-19T16:20:27 *** delt0r has joined #aichallenge 2010-11-19T16:20:44 yeah that's really annoying 2010-11-19T16:20:47 i've got one more idea 2010-11-19T16:20:58 but if that fails aswell i'll go back to my old version :/ 2010-11-19T16:21:10 *** delt0r___ has quit IRC (Ping timeout: 250 seconds) 2010-11-19T16:23:23 *** perror has quit IRC (Quit: Bye all !) 2010-11-19T16:24:29 bhasker: ur bot is doing better ain't it 2010-11-19T16:29:57 http://72.44.46.68/canvas?game_id=1097545 I lost this game due to "invalid input". whats a good way to find out what I did wrong (because this does not happen everytime) 2010-11-19T16:31:22 right before issuing an order 2010-11-19T16:31:31 you should check for the basic errors 2010-11-19T16:31:48 (sending too many ships, sending from a planet that isn't yours, sending ships to the planet that they came from) 2010-11-19T16:32:08 that last one isn't an error 2010-11-19T16:32:11 just debug everything that's you'll finally sent 2010-11-19T16:32:15 although it's not useful 2010-11-19T16:32:23 I mean when destination is the same as source 2010-11-19T16:32:30 that should give an invalid error right? 2010-11-19T16:32:34 does your issueorders method/function/procedure check src!=dest, src.owner=you, src.ships >= sentships 2010-11-19T16:32:38 yeah that doesn't trigger invalid order 2010-11-19T16:33:02 src==dest is supposed to be an error 2010-11-19T16:33:02 ow ok :o 2010-11-19T16:33:33 really? my bot did that for the longest time without issue 2010-11-19T16:33:50 bhasker: watched that game - that was reckless and you got away with it ;-) 2010-11-19T16:33:58 well "supposed to" and "reality" aren't always the same 2010-11-19T16:35:44 Mistmanov, glad new maps are working out for you. i had opposite effect - had to fix a number of things 2010-11-19T16:35:58 yeah it's a bit random 2010-11-19T16:36:19 oh well, i'm not going to win anyway, so the exact results don't worry me too much 2010-11-19T16:37:02 good games with your new bot on tcp 2010-11-19T16:37:27 =) 2010-11-19T16:37:43 yeah, winning is out of reach for most of us. still fun to waste some time on it ;-) 2010-11-19T16:38:09 fun/frustrating 2010-11-19T16:38:10 whatever :p 2010-11-19T16:38:15 Endgame Conditions: A bot issues a command with the same source and destination planet and forfeits the game. 2010-11-19T16:38:46 bug hunting is so... ugh right now my bot suddenly starts crashing on a part of the code that has been working just fine for days 2010-11-19T16:38:53 but, you seem to be right, I don't see that check in the code 2010-11-19T16:39:33 lol 2010-11-19T16:39:34 haha, yeah debugging is a !@#%$^... i so regret not writing proper test cases from the very beginning 2010-11-19T16:39:43 so do i 2010-11-19T16:39:48 a proper visualizer et 2010-11-19T16:39:50 etc 2010-11-19T16:39:52 but nooooo 2010-11-19T16:39:58 starting out with one file 2010-11-19T16:40:03 "how hard can it be?" 2010-11-19T16:40:05 xD 2010-11-19T16:41:19 yeah, interface looked deceptively simple, lol 2010-11-19T16:43:20 *** aerique has quit IRC (Quit: ...) 2010-11-19T16:44:19 is there any other way for "invalid input"!? i check what you said, but unfortunately it still crashes: if p.owner==player.ME and p.id!=target.id and p.ship_count>=nships: p.send_fleet(target, nships) 2010-11-19T16:44:40 bhasker and mistmanov, our latest bots seem to be about the same strength looking at tcp scores.. we kept trading places 2010-11-19T16:45:04 can be =) 2010-11-19T16:45:06 tim_: n_ships<=0 is not accepted 2010-11-19T16:45:14 tim: is nships >0 ? 2010-11-19T16:45:55 you can't send a negative number of ships and on tcp, you can't send 0 ships 2010-11-19T16:46:10 also: are you issuing multiple orders per turn? if so, are you updating the planet's shipscounts? 2010-11-19T16:47:00 Mistmanov is right too 2010-11-19T16:47:38 you should really get your program to print all the orders it's sending. problems then often become instantly clear 2010-11-19T16:49:33 *** rebelxt has left #aichallenge 2010-11-19T16:49:54 men 2010-11-19T16:50:10 have you guys ever looked at mega1's games 2010-11-19T16:50:12 again, there might be women here 2010-11-19T16:50:14 :p 2010-11-19T16:50:22 lol 2010-11-19T16:50:29 in theory at least.. 2010-11-19T16:50:32 i seriously doubt that :P 2010-11-19T16:50:50 me? I'm careful to issue a single order every turn. 2010-11-19T16:50:54 Oh I think there's at least one woman doing the ai_contest 2010-11-19T16:51:16 maybe 2010-11-19T16:51:23 a girl from my university started working on the starting thing 2010-11-19T16:51:28 and then gave up after 2 days or something 2010-11-19T16:51:29 but if she's also on irc? =) 2010-11-19T16:52:13 She maybe, you'll never know^^ 2010-11-19T16:52:42 And I think that if a girl is in irc now, she wouldn't say that she's a girl^^ 2010-11-19T16:52:52 :) 2010-11-19T16:53:03 I'd be scared of creepy stalking nerds... 2010-11-19T16:53:10 * Mistmanov is a creepy stalking nerd 2010-11-19T16:53:16 :( 2010-11-19T16:53:25 lol 2010-11-19T16:53:26 right, the topic was AI contests 2010-11-19T16:53:41 yeah so where was I? 2010-11-19T16:53:43 ah ya 2010-11-19T16:53:44 men 2010-11-19T16:53:50 have you guys ever looked at mega1's games 2010-11-19T16:53:53 ^_^ 2010-11-19T16:53:57 how do you get the contestbot to cough when it last saw someone? 2010-11-19T16:54:01 that's bosmicko right? 2010-11-19T16:54:04 yes 2010-11-19T16:54:09 @seen Mjothvitnir 2010-11-19T16:54:09 JamesMG: Mjothvitnir was last seen in #aichallenge 11 seconds ago: how do you get the contestbot to cough when it last saw someone? 2010-11-19T16:54:15 *** Bobng has quit IRC (Quit: Leaving) 2010-11-19T16:54:31 I think everyone has seen his games 2010-11-19T16:54:37 @seen janzert 2010-11-19T16:54:37 Mjothvitnir: janzert was last seen in #aichallenge 2 hours, 34 minutes, and 25 seconds ago: bhasker: the rating changes are caused by having bayeselo calculate and use the correct weight for draws instead of it's default which is calculated for chess 2010-11-19T16:54:37 somehow he coordinates his attacks so that the closest planet gives the final blow 2010-11-19T16:54:41 very hard to defend 2010-11-19T16:54:45 nice stuff 2010-11-19T16:55:00 interesting, if you msg it it doesn't work 2010-11-19T16:55:36 @rankings 2010-11-19T16:55:37 Error323_: Top 10 players: bix0r4ever(3366), bocsimacko(3248), smloh1(3175), wagstaff(3121), BaronTrozo(3119), murrayr(3116), protocolocon(3098), davidjliu(3091), smiley1983(3087), dmj111(3082) 2010-11-19T16:55:40 I assumed that was fairly common at this point Error323_? 2010-11-19T16:55:45 ^_^ 2010-11-19T16:55:56 well yes 2010-11-19T16:56:08 I do like to watch it though :) 2010-11-19T16:56:10 my bot still doesn't coordinate attacks -_- 2010-11-19T16:56:14 but not howto build that properly :P 2010-11-19T16:57:09 I loved the effect that mcleopold showed yesterday (?), where he went to capture some neutral in a far of corner, but also some other neutrals on the way 2010-11-19T16:57:21 and those extra neutrals also added some ships to the attack 2010-11-19T16:57:28 Yes that onw was really beautiful 2010-11-19T16:57:32 one* 2010-11-19T16:57:32 yes that what mega1 does all the time 2010-11-19T16:57:36 also in attacks 2010-11-19T16:58:04 it's awesome 2010-11-19T16:58:04 my way of coding something like that would involve 20 nested for loops 2010-11-19T16:58:11 :/ 2010-11-19T16:58:11 lol 2010-11-19T16:58:26 well if that would make it work 2010-11-19T16:58:27 Raschi submitted more than 25 times today!!! 2010-11-19T16:58:30 go for it i'd say 2010-11-19T16:58:32 I tried to get my bot to do that after seeing McLeopold's 2010-11-19T16:58:37 now mine is broken 2010-11-19T16:58:44 lol 2010-11-19T16:58:45 I think that was McLeopold's goal all along :) 2010-11-19T16:58:50 :D 2010-11-19T16:58:52 well, the 25 loops would probably make me timeout 2010-11-19T16:59:12 Some parts of my script are already getting executed something like 10.000 times per turn 2010-11-19T16:59:25 I was going to just say "wow" 2010-11-19T16:59:32 The strangest seen I've seen from mega1 was to attack a neutral planet without taking it 2010-11-19T16:59:47 but I bet some of my estimations and things get called that many times also 2010-11-19T16:59:50 My bot has gone after the neutral planet and he took it back after 2010-11-19T16:59:58 I fixed my bot after... 2010-11-19T17:00:09 well that's probably the danger of it 2010-11-19T17:00:23 since it can't look ahead that much sometimes that will be the sideeffect 2010-11-19T17:00:24 medrimonia: yes, that may happen in the above mentioned coordinated attack if it later realizes that it's not going to work 2010-11-19T17:00:52 mega1: will your bot make an intelligent decision to abandon it's coordinated attack at some point then? 2010-11-19T17:01:16 mine just lets the attack fail due to a lack of ships or whatever, it's kinda sad 2010-11-19T17:01:16 no, it doesn't even know that it's in a coordinated attack 2010-11-19T17:02:15 I colonize only if I'm sure that I'll have the advantage for some turns, or at least I'm supposed to... 2010-11-19T17:03:00 Oh nice game! 2010-11-19T17:03:24 http://ai-contest.com/visualizer.php?game_id=7994668 2010-11-19T17:04:22 something like turn 52 2010-11-19T17:05:04 That's absolutely not as this that my bot is supposed to work... but it worked 2010-11-19T17:05:15 at least that's something 2010-11-19T17:05:17 in my current version on tcp and official i accidently disable sniping >_< 2010-11-19T17:05:36 you have a specific algorithm for sniping? 2010-11-19T17:05:42 yes 2010-11-19T17:05:47 Mistmanov: yep 2010-11-19T17:05:54 mine just does it by default (it sees a weakly defended friendly -->> attack it) 2010-11-19T17:05:56 sniping, attacking, defending 2010-11-19T17:06:00 ye so does mine 2010-11-19T17:06:06 but it shouldn't 2010-11-19T17:06:27 or well sniping is so important imo that it should get high priority 2010-11-19T17:06:47 I don't know if it makes much difference 2010-11-19T17:06:53 well it does here 2010-11-19T17:07:04 since i sort my actions using some heuristic 2010-11-19T17:07:05 mine is like : defending, snipping, attacking adversary, expand, redistribute with other extra stuff 2010-11-19T17:07:19 but sniping should get more priority 2010-11-19T17:07:31 hmm 2010-11-19T17:07:54 a heuristic i need to change really 2010-11-19T17:07:56 :w 2010-11-19T17:08:00 oops wrong window 2010-11-19T17:08:30 hmmm in fact, what I wanted to do (but I haven't enough time) was to score each move (snip, attack,defend etc...) and sort it after. 2010-11-19T17:08:47 But now my bot is very conservative 2010-11-19T17:08:51 yes that's my idea aswell 2010-11-19T17:08:52 ^_^ 2010-11-19T17:09:03 i do it sort of already 2010-11-19T17:09:14 but it can be done much better 2010-11-19T17:09:14 *** jaspervdj has quit IRC (Quit: NEVER GONNA GIVE YOU UP NEVER GONNA LET YOU DOWN) 2010-11-19T17:09:28 I started out with that, but scoring them properly was beyond my skill 2010-11-19T17:09:44 *** tim_ has quit IRC (Remote host closed the connection) 2010-11-19T17:09:44 I sort each move inside of snip and all now 2010-11-19T17:09:46 so now I just go "generally, defense is more profitable than offense", so do defense first 2010-11-19T17:10:24 Mistmanov: I've said that I score my move, not that I score them rightly. =P 2010-11-19T17:10:25 raschi just reuploads if he loses the first game 2010-11-19T17:10:31 or any game for that matter :-) 2010-11-19T17:10:32 indeed 2010-11-19T17:10:40 raschi just reuploads. 2010-11-19T17:10:46 its kinda desparate :P 2010-11-19T17:10:50 desperate* 2010-11-19T17:10:53 You don't need to add any thing after =P 2010-11-19T17:10:54 or whatever :P 2010-11-19T17:10:56 to what end? Not like that's a valid long term strategy 2010-11-19T17:11:02 eh, if he enjoys seeing his name on the leader board, let him I'd say 2010-11-19T17:11:17 the only leaderboard that counts is made after the uploading is turned of =) 2010-11-19T17:11:21 exactly 2010-11-19T17:11:24 indee 2010-11-19T17:11:25 d 2010-11-19T17:11:45 Mistmanov: given that, why haven't you uploaded anything new? 2010-11-19T17:12:00 because I don't care about the leaderboard :p 2010-11-19T17:12:21 packing up a .zip file is more trouble than I'm willing to take 2010-11-19T17:12:26 -_- 2010-11-19T17:12:28 lol 2010-11-19T17:12:30 "make zip" 2010-11-19T17:12:34 ^_^ 2010-11-19T17:12:41 ok, I'll upload tonights bot 2010-11-19T17:12:50 lol well not for me :P 2010-11-19T17:12:54 i was just wondering =) 2010-11-19T17:13:24 ok, so maybe it was a bit because seeing a new submission rank lower feels bad :p 2010-11-19T17:13:51 ah so instead of not caring, you care a lot :P 2010-11-19T17:14:55 somewhat.. 2010-11-19T17:14:59 uploading now ;) 2010-11-19T17:17:05 :) 2010-11-19T17:18:12 Top 10 players: bix0r4ever(3369), bocsimacko(3258), BaronTrozo(3129), wagstaff(3124), murrayr(3117), protocolocon(3102), smloh1(3091), davidjliu(3088), george(3084), dmj111(3083) 2010-11-19T17:19:11 two people from "rufes band" in the top 10? can't remember seeing them before.. 2010-11-19T17:19:15 grrrr tomorrow I'll have a real connection at least during some hours 2010-11-19T17:19:41 protocolocon is here since a few days at least 2010-11-19T17:19:49 I'll would even say weeks 2010-11-19T17:20:06 but BaronTrozo... I've never seen 2010-11-19T17:21:16 *** amriedle has joined #aichallenge 2010-11-19T17:22:55 wow, my dev bot is performing worse than my minimax attempt 2010-11-19T17:23:14 been a successful day 2010-11-19T17:25:18 heh my bot's lucky run is over 2010-11-19T17:26:14 janzert: jeff has 2 accounts, jeff_cameron and j3camero 2010-11-19T17:27:41 medrimonia: barontrozo has been on tcp for quite a while, but interestingly his ratig there quite bad 2010-11-19T17:27:48 +is 2010-11-19T17:31:23 i'll resubmit aswell in a sec 2010-11-19T17:31:31 been coding on 5 beers 2010-11-19T17:31:36 lets see how that works out 2010-11-19T17:31:37 :D 2010-11-19T17:32:13 so anything new E323? 2010-11-19T17:32:45 *** smloh1 has joined #aichallenge 2010-11-19T17:32:48 some minor tweaks 2010-11-19T17:32:57 tomorrow i intend to try something really newish 2010-11-19T17:33:02 and that'll be the last of it 2010-11-19T17:33:07 *** McLeopold has joined #aichallenge 2010-11-19T17:33:27 invalid orders 4tl need to track that bug down 2010-11-19T17:33:28 http://72.44.46.68/canvas?game_id=1098231 2010-11-19T17:33:33 an interesting game to watch 2010-11-19T17:35:16 indee 2010-11-19T17:35:17 d 2010-11-19T17:35:20 complete change 2010-11-19T17:35:21 hehe 2010-11-19T17:36:07 yup 2010-11-19T17:36:28 http://tcp2.zeroviz.us/canvas.cgi?game_id=875 2010-11-19T17:36:37 that game just looks neat at the end 2010-11-19T17:36:41 medrimonia: how do your bots on tcp relate to the submitted one? 2010-11-19T17:37:17 at turn 53 ish 2010-11-19T17:37:58 rainfall 2010-11-19T17:38:03 yeah 2010-11-19T17:38:29 this game makes a great screensaver 2010-11-19T17:40:07 *** lavalamp_ has quit IRC (Quit: Page closed) 2010-11-19T17:41:28 hehe 2010-11-19T17:41:37 interesting i made some change to my eval function 2010-11-19T17:41:41 bot seems to be doing a little better 2010-11-19T17:41:49 http://72.44.46.68/canvas?game_id=1098283 2010-11-19T17:41:54 it makes interesting attacks 2010-11-19T17:41:57 i truely loved voidptr's fireworksbot 2010-11-19T17:42:11 http://tcp2.zeroviz.us/canvas.cgi?game_id=780 2010-11-19T17:42:18 it should get some kind of originality price :P 2010-11-19T17:42:20 mega1: sorry I was afk 2010-11-19T17:42:46 the submitted version is the r64 2010-11-19T17:42:48 ok lets upload this one aswell 2010-11-19T17:42:53 E323-18.0-2-g4a95d58 2010-11-19T17:43:02 I don't even think I've tested it on tcp 2010-11-19T17:43:20 so it seems 2010-11-19T17:43:26 no. so it's related to the r63 2010-11-19T17:43:26 that was a nice game to watch mcleopold 2010-11-19T17:43:28 so how did you test it? 2010-11-19T17:43:36 mega1: turn 14, lower right 2010-11-19T17:43:38 wtf? 2010-11-19T17:43:56 from the r63 to the r64, there was just a minor change 2010-11-19T17:44:16 because I had sometimes an index error in the 63 2010-11-19T17:45:36 mega1: I've a lot of problem on the tcp, because I've a bad connection 2010-11-19T17:46:02 http://72.44.46.68/canvas?game_id=1098350 lol creaping out of death 2010-11-19T17:46:03 wow, all those Ts 2010-11-19T17:46:53 Ts? 2010-11-19T17:46:54 you should be about 65% without those timeouts 2010-11-19T17:47:08 Timeouts are displayed as Loss(T) 2010-11-19T17:47:23 Error323_: nice draw 2010-11-19T17:47:24 Thanks 2010-11-19T17:47:43 Yes I'll try some bots tomorrow at my university 2010-11-19T17:47:53 j's and i's look to much alike at this time of day :( 2010-11-19T17:48:05 McLeopold: its not a draw 2010-11-19T17:48:06 =) 2010-11-19T17:48:08 i won 2010-11-19T17:48:17 but the visualizer doesn't see that 2010-11-19T17:48:21 show that* 2010-11-19T17:48:39 yea, it shows 698 to 698 2010-11-19T17:48:59 yea, though my growthrate is slightly higher 2010-11-19T17:49:02 1 planet more 2010-11-19T17:49:02 whatever happened to all the shampoo bots 2010-11-19T17:49:05 flag you still around? 2010-11-19T17:49:11 so i won by 1 more ship 2010-11-19T17:49:16 at turn 200 2010-11-19T17:49:16 ? 2010-11-19T17:49:19 awesomeness =) 2010-11-19T17:49:34 do you get growth at the very end? 2010-11-19T17:50:10 *** davidjliu has joined #aichallenge 2010-11-19T17:50:19 i think the engine simulates one turn beyond 200 2010-11-19T17:50:21 or 2 2010-11-19T17:50:23 not sure 2010-11-19T17:50:34 that would be a bug? 2010-11-19T17:50:46 no 2010-11-19T17:50:54 i think it takes the orders at turn 200 2010-11-19T17:50:58 and calculates state at 201 2010-11-19T17:51:00 what the visualizer shows is the state at the beginning of the turn not the end 2010-11-19T17:51:13 i.e. 1 of 200 is the starting state 2010-11-19T17:51:20 before the first turn 2010-11-19T17:51:29 and 200 of 200 is the state just before the last turn 2010-11-19T17:51:44 ah 2010-11-19T17:52:39 I'll update that then, so the visualizer shows correctly. 2010-11-19T17:52:55 I don't think the engine records the final state though 2010-11-19T17:53:11 basically it is the state sent to the bots for that turn 2010-11-19T17:53:22 but it uses it for determining the winner 2010-11-19T17:53:37 and since the bots do not ever receive the final state the visualizer doesn't get it either 2010-11-19T17:53:54 it ought to at least record the ending state, no? 2010-11-19T17:53:59 bhasker: http://72.44.46.68/canvas?game_id=1098450 :D 2010-11-19T17:54:01 it would be nice 2010-11-19T17:54:05 :P 2010-11-19T17:54:45 yea it blew it midway 2010-11-19T17:54:56 I've been kind of surprised that it hasn't been a larger issue in the forums and had someone pursue it in the code 2010-11-19T17:54:58 ye 2010-11-19T17:55:08 oh now i lost 2010-11-19T17:55:10 curses 2010-11-19T17:55:33 @repo 2010-11-19T17:55:34 McLeopold: repo = http://code.google.com/p/ai-contest/ 2010-11-19T17:55:38 but since it's mostly just a cosmetic thing from my perspective I haven't bothered 2010-11-19T17:55:44 bhasker: the amount of times i've lost at turn 200 is just annoying 2010-11-19T17:55:46 :/ 2010-11-19T17:56:32 it would be nice for the entire data of the game to be correct, so you could look at the last state and determine the outcome. 2010-11-19T17:56:50 otherwise, you'd have to simulate one turn everytime you wanted to know who won 2010-11-19T17:57:03 (assuming there wasn't a flag) 2010-11-19T17:59:33 wow i lost against bogusbrainpower o_O 2010-11-19T17:59:57 *** Rubicon-|-Cross has quit IRC (Quit: Leaving) 2010-11-19T17:59:59 McLeopold: yep, I agree. I just don't have to do that at all ;) 2010-11-19T18:01:18 *** Mistmanov has quit IRC (Quit: bedtime -_-) 2010-11-19T18:04:17 janzert: it should be putting the final state in, according to engine.py 2010-11-19T18:04:28 maybe it's dhartmei's engine that doesn't? 2010-11-19T18:06:00 hah welcome to club e323 2010-11-19T18:06:21 heh i lose to flavor! 2010-11-19T18:06:43 McLeopold: hmm, wierd. It's been quite a while since I looked at it either I must either have misunderstood something then or am misremembering 2010-11-19T18:07:03 I'm pretty sure the visualizer on the main site never shows the end state though either 2010-11-19T18:07:09 lol interesting game http://72.44.46.68/canvas?game_id=1098465 2010-11-19T18:07:15 so I doubt it's something specific to dhartmei's setup 2010-11-19T18:07:19 *** JamesMG has quit IRC (Quit: Leaving.) 2010-11-19T18:08:48 I don't like these rock paper scissor maps too much. 2010-11-19T18:08:57 :) 2010-11-19T18:09:03 a lot of the new maps i feel have to do with who is aggresive 2010-11-19T18:09:06 and lucky enough to get away with 2010-11-19T18:09:09 good thing that bix0r likes them even less: http://ai-contest.com/profile.php?user_id=8565 2010-11-19T18:09:24 if someone happens to throw to center and other doesn't 2010-11-19T18:09:25 you win 2010-11-19T18:09:29 *** nigg_ has joined #aichallenge 2010-11-19T18:10:01 mega1: hopfully it isn't just 1 round of r/p/s in a game, but several 2010-11-19T18:10:02 take-center > expand > wait > take-center 2010-11-19T18:10:24 yea but sometimes center is expensive and you get sniped 2010-11-19T18:14:09 *** Migi32 has joined #aichallenge 2010-11-19T18:14:38 bocsimacko resubmitted and he's first again after 4 games? Is rating no longer reset? 2010-11-19T18:14:46 Nope 2010-11-19T18:15:00 your first matches are against bot near your last rank 2010-11-19T18:15:12 ok 2010-11-19T18:15:33 that's probably better 2010-11-19T18:18:14 Top 10 players: bocsimacko(3458), bix0r4ever(3360), wagstaff(3137), murrayr(3137), BaronTrozo(3115), protocolocon(3104), smloh1(3090), Raschi(3085), davidjliu(3083), george(3080) 2010-11-19T18:18:21 woooot! i will enjoy my transient time in the top 10 while i can :) 2010-11-19T18:19:16 smloh1: you've been as high as 3rd I think 2010-11-19T18:20:03 that was because i had just resubmitted and managed to beat shangas 2010-11-19T18:20:16 I freakin hate it when i lose to noobs >_< 2010-11-19T18:20:17 bah 2010-11-19T18:20:33 that is, much lower ranked ppl 2010-11-19T18:21:56 i'm sure as more games are played i'll drop out again, but it's nice to be reported by contestbot 2010-11-19T18:21:57 any news on the final tournament? 2010-11-19T18:22:46 mega1: what sort of news? 2010-11-19T18:23:04 swiss sytem / round robin / etc 2010-11-19T18:23:15 map selection as we have today? 2010-11-19T18:23:19 not that I've heard of 2010-11-19T18:23:38 right now I'm going to push for keeping the same as we have 2010-11-19T18:23:38 it would be nice to know 2010-11-19T18:23:58 and slowly eliminating bottom ranked players if we aren't getting enough games 2010-11-19T18:24:06 because if it's a top 10 round robin I may take a different approach 2010-11-19T18:24:20 to give good ranking order 2010-11-19T18:24:37 I get the feeling that what works against the top bots may not work against the lower ranked ones and vice versa. 2010-11-19T18:25:03 but I may be able to do some more work on pairing before the contest ends so we'll see 2010-11-19T18:25:20 I don't want to go to a completely new untested system just for the final though 2010-11-19T18:25:29 if we can help it at least :) 2010-11-19T18:25:40 sure, and I'm not here to push you into that direction 2010-11-19T18:26:05 of course you've seen as much communication from Jeff as I have so... 2010-11-19T18:26:25 it's always possible he will step in and do something completely different 2010-11-19T18:26:25 did we have a top 100 round robin for tron? 2010-11-19T18:26:31 no 2010-11-19T18:26:38 how was it? 2010-11-19T18:26:38 mcleopold tcp2 is very lonely 2010-11-19T18:26:41 going to switch to tcp1 2010-11-19T18:26:50 :( 2010-11-19T18:26:59 he did cut it down to just the top 100 competitors competing for the last day of the final tournament though 2010-11-19T18:27:00 its just you and me duking it out 2010-11-19T18:27:10 but the pairing didn't change for it at all 2010-11-19T18:27:14 not really, that's my old bot 2010-11-19T18:27:22 just there so something is 2010-11-19T18:27:35 but tcp2 really shows you your flaws 2010-11-19T18:28:26 by old bot does alot better on the crazy_fast 2010-11-19T18:28:29 ah 2010-11-19T18:28:57 bhasker: like your invalids 2010-11-19T18:29:13 oh i know whats causing that, just don't have the time right now to track it down 2010-11-19T18:29:25 i am not adjusting my remaining ships after lauching a fleet 2010-11-19T18:29:30 so i end up double launching 2010-11-19T18:29:34 and errors out 2010-11-19T18:29:35 :-\ 2010-11-19T18:31:55 bhasker: were you also in the tron contest? 2010-11-19T18:32:07 nope 2010-11-19T18:32:10 janzert: let us know when you have the pairing decided or you even at the idea phase 2010-11-19T18:32:13 i got introduced to this by a1k0n 2010-11-19T18:32:15 janzert: why are the servers slow? :P 2010-11-19T18:32:20 who wisely dropped out after i got hooked 2010-11-19T18:32:21 :-\ 2010-11-19T18:32:31 hehe 2010-11-19T18:33:03 mega1: look like the best idea is keep an idea on the "final tournament checklist" thread. Jeff posted in there a little earlier, I'm replying now 2010-11-19T18:35:09 s/keep an idea on/keep an eye on 2010-11-19T18:35:28 right 2010-11-19T18:35:43 *** Palmik has quit IRC (Remote host closed the connection) 2010-11-19T18:39:08 hmmmm just thinking it know, it's 23h59 server time the closing of uploads right? 2010-11-19T18:40:00 that's my assumption but I need Jeff to confirm that 2010-11-19T18:40:51 Thanks 2010-11-19T18:52:15 amstan: are you sure there was a top 25 round robin? my memory is that Jeff just limited the games played to the top contestants he didn't change the pairing at all 2010-11-19T18:52:22 for the tron contest that is 2010-11-19T18:52:32 janzert: that's what's what i meant 2010-11-19T18:53:17 ok, I also thought it was top 100 but I could quite likely be misremembering that part 2010-11-19T18:55:09 @rankings 2010-11-19T18:55:10 Error323_: Top 10 players: bocsimacko(3532), bix0r4ever(3350), wagstaff(3129), Error323(3120), protocolocon(3105), murrayr(3095), BaronTrozo(3093), smloh1(3091), davidjliu(3081), george(3078) 2010-11-19T18:55:17 ahhhhhhhhhhhhhh yes #4 2010-11-19T18:55:23 * Error323_ enjoys it while he can 2010-11-19T18:55:27 *** Error323_ is now known as Error323 2010-11-19T18:55:59 congrats Error323 2010-11-19T18:56:06 well 2010-11-19T18:56:10 only played 2 games :P 2010-11-19T18:56:55 let see what will happen next 2010-11-19T18:57:19 well i'll probably drop like a brick and stabilize around 50 2010-11-19T18:57:22 =] 2010-11-19T18:58:08 oh man i remember testingFromPiter was really good on old mapset 2010-11-19T18:58:13 he must not have like this 2010-11-19T19:00:44 *** mceier has quit IRC (Quit: leaving) 2010-11-19T19:09:41 lol 2010-11-19T19:09:45 won again' 2010-11-19T19:09:52 i'm gonna kick to nr 1 maybe 2010-11-19T19:09:59 mega1: watch out :P 2010-11-19T19:10:21 please beat that bix0r guy down 2010-11-19T19:12:14 Anyone know if or how specification.php is generated from specification.md? 2010-11-19T19:13:37 janzert: how many maps will the eventual server hold? 2010-11-19T19:13:39 also 50? 2010-11-19T19:13:40 woah, every tag in specification.php is malformed, i.e, to close an anchor say it has 2010-11-19T19:13:50 for the finals? 2010-11-19T19:13:56 i'm probably already heavily overfitting on the current set :/ 2010-11-19T19:13:57 yes 2010-11-19T19:14:05 at this point most likely 50 2010-11-19T19:14:53 use another set for testing :) 2010-11-19T19:14:54 *** smloh1 has quit IRC (Ping timeout: 265 seconds) 2010-11-19T19:15:01 yea 2010-11-19T19:15:03 i do 2010-11-19T19:15:05 i will 2010-11-19T19:15:16 but still 2010-11-19T19:15:19 it'll happen xD 2010-11-19T19:15:28 *** rebelxt has joined #aichallenge 2010-11-19T19:15:29 on the contest server itself most players have only played each map twice 2010-11-19T19:17:25 just saw pairings go by for the top 3 players so expect to get a game result soon :) 2010-11-19T19:17:41 i'm 4 2010-11-19T19:17:45 but i'll upgrade soon 2010-11-19T19:17:53 prolly to #2 or #1 :D 2010-11-19T19:17:58 i gotta see that 2010-11-19T19:17:59 hehe 2010-11-19T19:18:01 before i goto bed :P 2010-11-19T19:18:13 http://ai-contest.com/profile.php?user_id=10154 2010-11-19T19:18:14 Top 10 players: bocsimacko(3534), bix0r4ever(3358), wagstaff(3129), Error323(3119), murrayr(3108), protocolocon(3104), BaronTrozo(3092), smloh1(3091), davidjliu(3081), george(3077) 2010-11-19T19:18:15 =] 2010-11-19T19:18:22 \o/ 2010-11-19T19:18:24 *** McLeopold has quit IRC (Quit: Leaving) 2010-11-19T19:26:14 *** Appleman1234 has joined #aichallenge 2010-11-19T19:27:18 ahw 2010-11-19T19:27:20 only rank 3 2010-11-19T19:27:23 oh well 2010-11-19T19:28:06 looks like mega1's new submission handled wagstaff just fine 2010-11-19T19:28:30 it was over pretty quick 2010-11-19T19:28:48 yeah, no problem at all 2010-11-19T19:28:56 one more time 2010-11-19T19:28:58 @rankings 2010-11-19T19:29:00 Error323: Top 10 players: bocsimacko(3532), bix0r4ever(3357), Error323(3222), wagstaff(3129), murrayr(3108), protocolocon(3102), davidjliu(3082), BaronTrozo(3081), george(3073), dmj111(3071) 2010-11-19T19:29:02 <33333333 2010-11-19T19:29:18 Error323:3222 almost right 2010-11-19T19:29:31 haha ye 2010-11-19T19:29:32 xD 2010-11-19T19:29:58 let the mentalist go to bed 2010-11-19T19:30:12 i have to look at the next :P 2010-11-19T19:30:21 * Error323 slightly obsessive 2010-11-19T19:40:33 mega1: seriously http://ai-contest.com/visualizer.php?game_id=7998319 2010-11-19T19:40:39 that is just awesome 2010-11-19T19:41:17 holy crap 2010-11-19T19:41:19 i won again 2010-11-19T19:41:20 good god 2010-11-19T19:41:24 wtf is going on 2010-11-19T19:41:30 how can i sleep like this 2010-11-19T19:42:33 this isn't shabby either: http://ai-contest.com/visualizer.php?game_id=7999505 2010-11-19T19:42:40 * mega1 returns the compliment 2010-11-19T19:43:25 haha ye thx 2010-11-19T19:43:28 ^_^\ 2010-11-19T19:45:49 I won mine too! 2010-11-19T19:46:11 If it keeps going like that, I'll stay in the top20 2010-11-19T19:46:32 yeah same here 2010-11-19T19:46:36 well ur more stable 2010-11-19T19:46:38 good job 2010-11-19T19:46:51 I wasn't expecting my bot to be ranked so good 2010-11-19T19:46:56 i just love the sight of all green: http://ai-contest.com/profile.php?user_id=10154 2010-11-19T19:47:45 Let's go win against mega1 now =P 2010-11-19T19:47:51 =] 2010-11-19T19:47:55 He's used to all green 2010-11-19T19:48:14 i hacked the server and never team up against him 2010-11-19T19:48:22 this way i'll bypass and become #1 2010-11-19T19:48:26 hahaha 2010-11-19T19:48:53 xD 2010-11-19T19:49:31 dammit everytime i upload i always start up losing:-\ 2010-11-19T19:49:39 so takes a long ass time to even climb to 40 2010-11-19T19:49:56 congrats E323 2010-11-19T19:50:00 btw take a screenshot 2010-11-19T19:50:02 before you drop 2010-11-19T19:50:04 thx 2010-11-19T19:50:07 haha yeah 2010-11-19T19:50:11 and post it on facebook 2010-11-19T19:50:13 for glory 2010-11-19T19:50:20 i don't have facebook :P 2010-11-19T19:50:27 ah well somewhere 2010-11-19T19:50:28 then 2010-11-19T19:50:34 twitteR? 2010-11-19T19:50:49 http://fhuizing.no-ip.org/cows.jpg however this is me 2010-11-19T19:50:52 on the right 2010-11-19T19:51:48 ah 2010-11-19T19:51:55 my bot is slowly making its way to the top 50 2010-11-19T19:51:59 @50 now 2010-11-19T19:52:00 bhasker: No! 2010-11-19T19:52:05 @rainbow 2010-11-19T19:52:05 bhasker: (rainbow ) -- Returns colorized like a rainbow. 2010-11-19T19:52:11 @rainbow does a happy dance 2010-11-19T19:52:11 bhasker: does a happy dance 2010-11-19T19:52:47 http://72.44.46.68/canvas?game_id=1099625 double snipe 2010-11-19T19:52:50 FATALITY 2010-11-19T19:53:14 oh triple snipe actually 2010-11-19T19:53:33 heh thats rock paper scissors for you 2010-11-19T19:53:38 he over extended 2010-11-19T19:53:41 ye 2010-11-19T19:53:46 might have been fine if he had not gone for the 57 2010-11-19T19:53:50 might have survived 2010-11-19T19:53:51 *** si has joined #aichallenge 2010-11-19T19:54:17 *** si is now known as Guest37126 2010-11-19T19:54:37 *** medrimonia1 has joined #aichallenge 2010-11-19T19:56:26 bhasker: what do you do at yahoo btw? 2010-11-19T19:56:42 work for y! mail 2010-11-19T19:57:20 *** medrimonia has quit IRC (Ping timeout: 255 seconds) 2010-11-19T19:57:25 Error323: You don't play on zerovis.us:8080? 2010-11-19T19:57:29 *** medrimonia1 is now known as medrimonia 2010-11-19T19:57:43 medrimonia: whats the use of zerovis now that tcp is up 2 date? 2010-11-19T19:57:51 the original tcp that is 2010-11-19T19:57:55 or isn't it up to date 2010-11-19T19:58:25 *** nigg_ has quit IRC (Read error: Connection reset by peer) 2010-11-19T19:58:30 I would say that... you've got higher chances to play against someone of the top 100 on zerovis 2010-11-19T19:59:23 mkay 2010-11-19T19:59:28 i'm on it now 2010-11-19T19:59:37 against you :P 2010-11-19T19:59:43 *** lavalamp_ has joined #aichallenge 2010-11-19T19:59:46 *** nigg_ has joined #aichallenge 2010-11-19T20:00:08 Against me? I bet i'll time out^^ 2010-11-19T20:00:27 (bad connection) 2010-11-19T20:00:53 i think its a 200 turn game 2010-11-19T20:01:01 which i usually lose 2010-11-19T20:01:14 I can't see any of the games^^ 2010-11-19T20:01:20 me neither 2010-11-19T20:01:23 *** nigg_ has quit IRC (Client Quit) 2010-11-19T20:01:25 oh i won :P 2010-11-19T20:01:28 I just look at the error 2010-11-19T20:01:36 @rainbow \o/ 2010-11-19T20:01:36 Error323: \o/ 2010-11-19T20:01:47 I'll see the game^^ 2010-11-19T20:01:56 *** nigg_ has joined #aichallenge 2010-11-19T20:03:53 and again 2010-11-19T20:04:16 wel its been fun but i'm falling of my cahir 2010-11-19T20:04:19 too much booze 2010-11-19T20:04:22 :D 2010-11-19T20:04:26 gnite 2010-11-19T20:04:41 I've still to add some line for expansion 2010-11-19T20:05:44 and my defense has still to be fixed 2010-11-19T20:05:45 *** bobng_ has joined #aichallenge 2010-11-19T20:05:56 *** nigg_ has quit IRC (Client Quit) 2010-11-19T20:06:20 *** Appleman1234 has quit IRC (Ping timeout: 245 seconds) 2010-11-19T20:06:22 *** bobng_ is now known as bobng 2010-11-19T20:11:57 *** Guest37126 has quit IRC (Quit: Page closed) 2010-11-19T20:12:40 *** yasith_ has quit IRC (Ping timeout: 240 seconds) 2010-11-19T20:13:46 whoaaaaaaaaaoooooooooot 2010-11-19T20:13:51 won against wagstaff :P 2010-11-19T20:13:58 damn i'm on a roll for some reason 2010-11-19T20:14:02 Waow congrats! 2010-11-19T20:14:30 lol 1 ship more 2010-11-19T20:14:33 http://ai-contest.com/visualizer.php?game_id=8006597 2010-11-19T20:14:53 hahaha I was looking^^ 2010-11-19T20:14:55 but this is getting really interesting 2010-11-19T20:14:57 beautiful 2010-11-19T20:16:10 i'll maybe kick to #1 now 2010-11-19T20:16:19 and then i'll defintly have to screenshot it 2010-11-19T20:16:44 it seems i've reached the ballmer peak somehow: http://xkcd.com/323/ 2010-11-19T20:17:03 XD 2010-11-19T20:17:09 Didn't know about that 2010-11-19T20:18:09 hilarious 2010-11-19T20:18:15 Top 10 players: bocsimacko(3540), bix0r4ever(3287), Error323(3245), wagstaff(3118), protocolocon(3105), dmj111(3097), murrayr(3092), BaronTrozo(3089), davidjliu(3086), george(3071) 2010-11-19T20:18:58 *** nann has joined #aichallenge 2010-11-19T20:21:22 mega1: still there? 2010-11-19T20:21:26 yes 2010-11-19T20:21:31 :D 2010-11-19T20:21:42 I'm in a different time zone thought at the moment 2010-11-19T20:21:52 *** Apophis has quit IRC (Ping timeout: 240 seconds) 2010-11-19T20:21:57 02:18 am here 2010-11-19T20:22:08 I know 2010-11-19T20:22:32 and I lost 2010-11-19T20:22:58 my improvement is rather hackish and it didn't work in that loss 2010-11-19T20:23:03 :/ 2010-11-19T20:23:15 gotta find out why the "correct" version loses so much 2010-11-19T20:23:43 no you gotta stop and give us a chance ;-) 2010-11-19T20:23:58 *** Blkt has quit IRC (Quit: Error: do not makunbound t please!) 2010-11-19T20:26:15 craaaaaaaap 2010-11-19T20:26:19 @rankings 2010-11-19T20:26:20 Error323: Top 10 players: bocsimacko(3384), Error323(3320), bix0r4ever(3254), wagstaff(3114), protocolocon(3110), dmj111(3098), davidjliu(3092), BaronTrozo(3078), george(3078), murrayr(3055) 2010-11-19T20:26:22 nr 2 2010-11-19T20:26:36 well highest i've ever been anyway 2010-11-19T20:27:42 Gosh I was starting a "little improve" and in fact, I think it'll make something like a tenth part of my bot^^ 2010-11-19T20:28:53 *** bhasker has quit IRC (Ping timeout: 265 seconds) 2010-11-19T20:32:01 If you win the next one you could be #1 Error... =P 2010-11-19T20:32:10 indeed :P 2010-11-19T20:38:15 *** Migi32 has quit IRC (Quit: www.ai-contest.com) 2010-11-19T20:47:30 medrimonia: 2010-11-19T20:47:31 oops 2010-11-19T20:47:38 mega1: want my binary again? 2010-11-19T20:47:45 i wanna know how good it does against urs 2010-11-19T20:47:59 I have spare processing time if you want it 2010-11-19T20:48:20 arf... lose against #111... 2010-11-19T20:48:25 jail's set up so it's gonna take only a short time 2010-11-19T20:48:44 Error323: uh-oh 2010-11-19T20:48:52 http://ai-contest.com/visualizer.php?game_id=8008959 2010-11-19T20:49:04 don't really know how to hande games like this 2010-11-19T20:49:05 fuuuuuuuuuuuk 2010-11-19T20:49:07 =) 2010-11-19T20:50:03 mega1: i mailed it 2010-11-19T20:50:04 I've more ships, more prod, but ... 2010-11-19T20:50:07 let me know tomorrow :) 2010-11-19T20:50:09 <--- zzzz 2010-11-19T20:50:20 sorry for that loss, good night 2010-11-19T20:50:45 hahha 2010-11-19T20:50:49 don't appologize ;) 2010-11-19T20:50:49 *** Cold-Phoenix has joined #aichallenge 2010-11-19T20:50:53 nice work 2010-11-19T20:50:56 cya 2010-11-19T20:55:31 *** yasith has joined #aichallenge 2010-11-19T20:55:55 *** Cold-Phoenix has quit IRC () 2010-11-19T21:04:21 medrimonia: your bot redistributed ships the wrong way 2010-11-19T21:04:47 which game? 2010-11-19T21:06:12 against barnik? 2010-11-19T21:07:46 yes 2010-11-19T21:08:05 sent them to that planet on the right 2010-11-19T21:08:17 while that planet should have sent ships to the middle 2010-11-19T21:09:59 *** Foobarrel has joined #aichallenge 2010-11-19T21:13:27 in fact 2010-11-19T21:13:43 my right planet was at the same distance from me than from my enemy 2010-11-19T21:13:55 *** Appleman1234 has joined #aichallenge 2010-11-19T21:14:02 that's why it was considered as an important planet 2010-11-19T21:14:12 And I tried to keep it 2010-11-19T21:14:27 But I think I should have attacked in fact 2010-11-19T21:15:01 to keep the number of ships low. 2010-11-19T21:15:38 But with the way I've done my code, it'll be hard 2010-11-19T21:15:56 it looks to me as if barnik had sufficent force in the middle to take over you front line planet 2010-11-19T21:16:38 He has less ships in overall 2010-11-19T21:16:52 so my right planet should have attacked in the center 2010-11-19T21:17:24 by the time it gets there you're lost, no? 2010-11-19T21:17:43 it should have attacked since the beginning^^ 2010-11-19T21:17:48 *** Foobarrel has left #aichallenge ("Leaving") 2010-11-19T21:18:16 Top 10 players: bocsimacko(3390), Error323(3278), bix0r4ever(3207), protocolocon(3096), wagstaff(3094), dmj111(3085), davidjliu(3068), xpctxpeer(3067), BaronTrozo(3055), george(3054) 2010-11-19T21:19:05 or I should have redistributed the ships in the other way 2010-11-19T21:19:24 But I've still to work on this kind of situation 2010-11-19T21:20:13 On mirror map, when I have a planet on the symetrical line, I can't defend well 2010-11-19T21:28:40 It's time to go to sleep I think 2010-11-19T21:28:49 good night 2010-11-19T21:28:59 Good night 2010-11-19T21:29:53 *** yasith has quit IRC (Ping timeout: 255 seconds) 2010-11-19T21:41:49 * lavalamp_ feels optimistic after seeing rating on McLeopold's server 2010-11-19T21:58:05 * lavalamp_ optimistic feeling dissapates 2010-11-19T22:03:25 that was short lived 2010-11-19T22:10:41 *** davidjliu has quit IRC (Quit: Page closed) 2010-11-19T22:11:21 *** cfaftw has joined #aichallenge 2010-11-19T22:11:35 did demerzel ever get his C bot working>? 2010-11-19T22:18:15 Top 10 players: bocsimacko(3379), bix0r4ever(3231), Error323(3189), wagstaff(3104), protocolocon(3098), dmj111(3084), murrayr(3078), davidjliu(3069), george(3055), BaronTrozo(3049) 2010-11-19T22:22:33 what is "server time"? 2010-11-19T22:22:41 (as in time zone) 2010-11-19T22:41:08 *** Naktibalda has quit IRC (Remote host closed the connection) 2010-11-19T22:49:33 *** cfaftw has quit IRC (Quit: Leaving) 2010-11-19T23:18:16 Top 10 players: bocsimacko(3467), bix0r4ever(3282), felixcoto(3221), Error323(3143), wagstaff(3101), protocolocon(3100), davidjliu(3084), dmj111(3072), medrimonia(3061), george(3053) 2010-11-19T23:32:52 Mjothvitnir: @time 2010-11-19T23:32:55 @time 2010-11-19T23:32:55 antimatroid1: 11:32 PM, November 19, 2010 2010-11-19T23:33:45 um that doesn't seem right. 2010-11-19T23:34:55 janzert in the forum says that the contest is "i.e. central standard time zone", is that zone correct?? 2010-11-19T23:35:41 Mjothvitnir: that's the zone the server shows yes 2010-11-19T23:36:18 but contestbot showed eastern standard just now 2010-11-19T23:36:32 so contestbot isn't in that time zone or didn't he get switched to standard time? 2010-11-19T23:36:54 no idea, server time right now is Fri Nov 19 22:36:50 CST 2010 2010-11-19T23:37:02 *** amriedle has quit IRC (Quit: Lost terminal) 2010-11-19T23:37:14 now that sounds like central time :) 2010-11-19T23:37:21 :) 2010-11-19T23:38:27 ok, until I hear otherwise I'll assume I have one hour less, so to speak. :( 2010-11-19T23:38:51 one hour less? it's one hour more than eastern time 2010-11-19T23:39:00 I'm MST. 2010-11-19T23:39:06 ahh, yes 2010-11-19T23:40:22 Mjorthvitnir: i would suggest uploading hours if not a day before the deadline 2010-11-19T23:40:43 just in case 2010-11-19T23:46:40 Also to be clear, and you may already know this, Jeff has not confirmed CST as the correct timezone to use 2010-11-19T23:47:15 it's just the one to that makes sense to me since all the times listed in profiles and such are CST now 2010-11-19T23:48:44 *** cfaftw has joined #aichallenge