2010-10-16T00:01:31 Top 10 players: dmj111(3538), MariusCG(3418), deepblue(3391), Hazard(3341), ofrias(3301), fglider(3297), felixcoto(3292), EBraun(3290), ruilov(3285), FlameN(3283) 2010-10-16T00:03:28 *** Epona has quit IRC (Read error: Connection reset by peer) 2010-10-16T00:03:56 *** McLeopold has joined #aichallenge 2010-10-16T00:04:35 i wonder if honolulu-18 is in honolulu right now.. 2010-10-16T00:05:18 *** Epona has joined #aichallenge 2010-10-16T00:05:27 hey, is anyone here? 2010-10-16T00:05:28 *** yasith has quit IRC (Ping timeout: 264 seconds) 2010-10-16T00:05:34 hi 2010-10-16T00:05:43 here. 2010-10-16T00:06:18 maybe honolulu-18 is the 18th patriotic representative honolulu? 2010-10-16T00:06:47 i don't know. my bot has dropped rank from 100 to 900 since i submitted it. 2010-10-16T00:06:59 doh 2010-10-16T00:07:38 i was wondering about honolulu-18 since I am here in waikiki right now 2010-10-16T00:08:16 ooh i just beat oldman5, i hate that old guy always stomping on my dreams 2010-10-16T00:11:11 yeah, teach him a lesson! 2010-10-16T00:11:41 i just peeled out on his lawn! hahahaah!! 2010-10-16T00:11:47 it's just after midnight here, and i'm finding wine does not help me code. 2010-10-16T00:12:22 try a mai tai, I find they work wonders 2010-10-16T00:14:51 *** davidjliu has quit IRC (Quit: Page closed) 2010-10-16T00:15:49 whats up with the contest server, one game in the last 10 hours 2010-10-16T00:17:27 *** andrew-101 has quit IRC (Read error: Connection reset by peer) 2010-10-16T00:21:09 *** yasith has joined #aichallenge 2010-10-16T00:24:03 so whose bot is oldman 2010-10-16T00:24:36 well, its oldman 2010-10-16T00:25:53 oldman is a tough bot 2010-10-16T00:26:05 mean mean oldman 2010-10-16T00:26:05 yea 2010-10-16T00:26:24 though my new version is doing quite well i must say 2010-10-16T00:26:34 if he jumps out to an early lead on me - which is most of the time, I never recover 2010-10-16T00:26:57 what is your bots name rwa? 2010-10-16T00:26:58 I have big plans. Big plans I tell you!! 2010-10-16T00:27:01 vortex 2010-10-16T00:27:09 aah 2010-10-16T00:27:18 interesting bot, i learnt quite a bit from watching your bot 2010-10-16T00:27:23 so we meet again.. bwahahah 2010-10-16T00:27:43 my bot is renamed now 2010-10-16T00:27:48 runs under the codename of malazan! 2010-10-16T00:28:00 ah, yeah, that's a strong bot 2010-10-16T00:28:22 yea i rewrote the core of my bot 2010-10-16T00:28:28 I might have AntBot 2 up in a few minutes. 2010-10-16T00:28:40 nice want to see how my bot does against antbot now 2010-10-16T00:28:49 crowbot i think it holds its own quite well 2010-10-16T00:28:59 Yeah, but it's stupid. 2010-10-16T00:29:24 how is antbot smarter? 2010-10-16T00:29:29 Once it starts attacking a planet, it just keeps going. 2010-10-16T00:29:33 *** Titankiller has joined #aichallenge 2010-10-16T00:29:40 Even if it can't get it. 2010-10-16T00:29:58 punch through eh? 2010-10-16T00:30:12 Yeah, it usually works, but not always. 2010-10-16T00:30:46 yeah 2010-10-16T00:30:54 who's bot is honolulu? 2010-10-16T00:30:58 mostly if the other bot gives up for a second and the planet flips to you 2010-10-16T00:31:04 it usually is enough to turn the tide 2010-10-16T00:32:08 well that was cool 2010-10-16T00:32:16 one line of code just doubled the speed of my bot 2010-10-16T00:33:29 I might not timeout 75% of the games on main server now :p 2010-10-16T00:33:31 beat up by an oldman again :( 2010-10-16T00:34:12 my bot spends almost no time thinking. its retarded 2010-10-16T00:35:24 sounds like it's thinking-cycles challenged 2010-10-16T00:35:27 if you are not searching game tree there really isn't much thinking to do 2010-10-16T00:35:45 its just like can i get this planet-> yes/no, attack 2010-10-16T00:35:56 can i hold this planet -> yes/no attack 2010-10-16T00:36:09 and if i have ships left over where should i send them 2010-10-16T00:36:14 thats pretty much the core of my bot 2010-10-16T00:36:57 throw in some strategy! 2010-10-16T00:37:01 my bot isnt even that clever, its more about action than determine who kills who 2010-10-16T00:37:16 strategy = constraining your opponents actions so their choices are predictable! 2010-10-16T00:38:50 aka code in more situations 2010-10-16T00:39:37 my submission, mid september, did only one thing correctly: if at all possible, protect a planet that is under attack from being taken. everything else it does is stupid. 2010-10-16T00:40:25 now i keep trying small tweaks to make it beat itself, but it really needs some proper future-tree (bhasker: "game tree"?) 2010-10-16T00:40:44 i'm not sure i can call it "state space", but maybe that's what it is. 2010-10-16T00:41:33 every time i try to make my bot look into the future, it becomes too defensive and loses more :( 2010-10-16T00:42:21 the only way to consistently beat my bot is to be smart about which enemy planets you attack - and approximately 900 bots are doing that. 2010-10-16T00:44:00 krokkrok: i haven't used future-looking for anything other than defending the planets i already have, so i may have fallen into the same strategic flaw as yourself. 2010-10-16T00:44:52 we both need more sophisticated ways of deciding how to respond to future states than just "don't let the enemy take what's mine." 2010-10-16T00:46:44 i took out my code that made a planet defend *itself* and my win ratio goes up. i dont know why 2010-10-16T00:47:38 overcounting needed ships? 2010-10-16T00:47:45 do you make neutral or enemy planets conquer themselves? 2010-10-16T00:49:47 i think the math is correct, but, logic says it cant be because abandoning a planet when there is a very strong enemy plant close by seems like a bad idea 2010-10-16T00:49:50 *** yasith has quit IRC (Ping timeout: 265 seconds) 2010-10-16T00:50:36 *** mimidoux has quit IRC (Ping timeout: 265 seconds) 2010-10-16T00:52:22 lol, i said enemy plant. my bot really is retarded. cant even beat a plant :( 2010-10-16T00:52:49 *** rebelxt_ has joined #aichallenge 2010-10-16T00:53:43 hasta la vista 2010-10-16T00:53:46 *** krokkrok has quit IRC (Quit: Page closed) 2010-10-16T01:01:31 Top 10 players: dmj111(3543), MariusCG(3423), deepblue(3396), Hazard(3346), ofrias(3305), felixcoto(3297), EBraun(3294), fglider(3290), ruilov(3289), savchenko(3288) 2010-10-16T01:05:08 *** yasith has joined #aichallenge 2010-10-16T01:06:18 yellowseed, do you want some advice? 2010-10-16T01:06:38 from you McLeopold? yes :-) 2010-10-16T01:07:00 You said earlier "don't let the enemy take what's mine." 2010-10-16T01:07:18 It's all yours. Every planet. 2010-10-16T01:07:24 yes, that's what my last submission tries to do. 2010-10-16T01:07:27 ah- 2010-10-16T01:07:29 -nice. 2010-10-16T01:07:41 Defense code is attack code. 2010-10-16T01:08:00 so... suddenly it's tricky again: to decide which planets get priority. 2010-10-16T01:08:46 it's not as simple as growth_rate or proximity, 2010-10-16T01:08:56 My official submission just tries to growth. It doesn't know who's who, just what it hasn't taken yet. 2010-10-16T01:09:10 some sort of ... i don't know. it's fun to think about different ways to approach- 2010-10-16T01:09:19 *** krokokrusa has joined #aichallenge 2010-10-16T01:09:52 -huh. it doesn't know who's who, but it's not blind to how growth_rate affects the strength of planets that happen to be owned by the enemy- 2010-10-16T01:10:54 No, it's not. It knows how a planet switching hands effect the growth difference. 2010-10-16T01:10:54 thanks! i hope to meet you in the top 100 some day soon. 2010-10-16T01:11:23 are you on tcp? 2010-10-16T01:11:28 oh.. ok.. 2010-10-16T01:11:57 i played once on tcp, just to see that the new planetwars kit (thanks ulope and apinkin) tcp script works. 2010-10-16T01:12:02 (it does) 2010-10-16T01:12:28 you play a few different strategies under different names on tcp, right? 2010-10-16T01:12:33 Keep playing on it. You should run into one of my bots. 2010-10-16T01:12:35 Yes 2010-10-16T01:12:56 *** delt0r_ has joined #aichallenge 2010-10-16T01:13:55 :-) i'm done coding today, but i've got no plans but coding tomorrow and a hefty espresso is sure to be part of my morning rituals. 2010-10-16T01:14:08 goodnight 2010-10-16T01:14:34 *** delt0r___ has quit IRC (Ping timeout: 260 seconds) 2010-10-16T01:14:49 yellowseed: glad to see you are using python kit too 2010-10-16T01:15:06 <- apinkin on github 2010-10-16T01:15:10 Yes. If I can't win this, I'll make sure python does. :) 2010-10-16T01:15:16 rebelxt: hey, i recognize that username from the git logs ;-) 2010-10-16T01:16:15 yes, the most legible language will trump all others! 2010-10-16T01:16:43 I can develop quickly in python. I'm not slowed down by type checks. 2010-10-16T01:16:56 I just toss lists and dictionaries around like... 2010-10-16T01:17:05 something... 2010-10-16T01:18:05 hehe. i sure hope we don't need to port to c++ at the end ;) 2010-10-16T01:20:25 ulope: are you here? 2010-10-16T01:23:30 good news: official server is still playing some games. bad news: match results are getting mixed up again (i.e. winner is recorded as a loser) 2010-10-16T01:24:35 *** Titankiller has quit IRC (Quit: Nettalk6 - www.ntalk.de) 2010-10-16T01:26:04 McLeopold: are you trying a different approach in your new bot? i need some inspiration to re-do my bot again 2010-10-16T01:26:30 layers 2010-10-16T01:27:09 what does your current bot do, at a high level 2010-10-16T01:27:11 I'm reworking my bot significantly... 2010-10-16T01:27:54 current production bot is pretty simple - simulation of the future 2010-10-16T01:28:24 huge design space around that idea though... 2010-10-16T01:28:36 calculating ships needed, surplus, etc and making decisions on what to do next 2010-10-16T01:28:51 you could rewrite your forwarding logic 2010-10-16T01:29:30 i have 4 versions of forwarding/funneling logic ;-) but yeah, there is another idea in the works 2010-10-16T01:29:42 Have you solved close starts? 2010-10-16T01:30:46 You could find better openings to some of your games? 2010-10-16T01:31:13 McLeopold: mostly. i normally win close starts but there is still an issue 2010-10-16T01:31:48 as far as simulating the future, the real question is: what does the future hold? 2010-10-16T01:32:16 McLeopold: yeah openings need attention. using knapsack plus greedy (like most people probably do) 2010-10-16T01:32:52 what's greedy? 2010-10-16T01:34:08 greedy is no dynamic programming. just pick best move, execute, then see if there are many enough ships left for next best move, etc 2010-10-16T01:34:36 oh 2010-10-16T01:34:50 and dp is rearranging until you got the most 2010-10-16T01:34:57 right 2010-10-16T01:35:11 those are the bots I snipe :) 2010-10-16T01:35:56 my bot is pretty simple.. there are a bunch of heuristics though to handle sniping, anti-sniping, close starts, early funneling, etc, etc 2010-10-16T01:36:01 whats napsack 2010-10-16T01:36:03 You could try and play positionally 2010-10-16T01:36:27 Epona: given 100 ships, which planets should you choose to give you the most growth 2010-10-16T01:36:34 what i'm Epona, http://en.wikipedia.org/wiki/Knapsack_problem 2010-10-16T01:36:55 err merged two sentences into one ;) 2010-10-16T01:37:59 ah 2010-10-16T01:38:17 so its a way of looking at your tactics 2010-10-16T01:39:15 Epona, knapsack just helps you solve ship allocation problem a bit better than greedy algorithms 2010-10-16T01:42:24 http://en.wikipedia.org/wiki/Dijkstra%27s_algorithm 2010-10-16T01:42:28 another good one^ 2010-10-16T01:43:07 McLeopold: good ideas. i'm experimenting in most of these directions. too bad there is not enough time with work/etc. maybe i will take a couple of days vacation in November and drill into this 2010-10-16T01:45:51 Epona: yeah shortest path is useful when funneling through intermediate hops (which i could not make work as good as direct funneling) 2010-10-16T01:46:30 summation is the key 2010-10-16T01:46:34 imo 2010-10-16T01:48:10 *** Appleman1234 has quit IRC (Ping timeout: 245 seconds) 2010-10-16T01:52:45 later, all 2010-10-16T01:54:54 *** rebelxt_ has quit IRC (Quit: Page closed) 2010-10-16T01:55:51 AntBot is up. And just did the dumbest thing ever. 2010-10-16T01:55:57 http://72.44.46.68/canvas?game_id=511352 2010-10-16T01:56:40 nice. it pulled off a ragebot move ;) 2010-10-16T01:56:48 yeah, stf 2010-10-16T01:56:50 wtf 2010-10-16T01:57:05 that was a bad typo 2010-10-16T01:57:08 haha 2010-10-16T01:57:22 i've just spent the last few hours rewriting from scratch again 2010-10-16T01:57:24 ... 2010-10-16T01:58:14 my old code just got too scattered and broken in many pieces, figured it's time for a rebuild 2010-10-16T01:58:27 You might consider rolling your own framework 2010-10-16T01:58:43 That's how I can get new bots built really fast. 2010-10-16T01:58:51 keep a template? 2010-10-16T01:59:17 i do have a generic one i just copy... so I guess I sort of do? 2010-10-16T01:59:58 Well, push stuff to functions that you can mix a match. 2010-10-16T02:00:17 ahhh 2010-10-16T02:00:18 *** Appleman1234 has joined #aichallenge 2010-10-16T02:00:19 makes sense 2010-10-16T02:00:40 I've got a few funneling routines. I can switch them in and out. 2010-10-16T02:00:46 then just switch between different functions (enable/disable) for different behaviours 2010-10-16T02:00:47 right 2010-10-16T02:01:14 Plus, then you can start turning pieces on or off depending on the situation of the game. 2010-10-16T02:01:20 how are you guys doing funnelling? atm I'm just doing some dot-products to keep within a set deviation from course 2010-10-16T02:01:32 Top 10 players: dmj111(3561), MariusCG(3472), deepblue(3409), Hazard(3357), felixcoto(3315), ofrias(3311), Alvraro(3307), ruilov(3298), savchenko(3298), fglider(3278) 2010-10-16T02:01:43 A* or dijkstra's? 2010-10-16T02:01:53 dot product? 2010-10-16T02:01:59 yeah 2010-10-16T02:02:10 just to make sure that my angles are below a limit 2010-10-16T02:02:12 isn't that for adding vectors? 2010-10-16T02:02:17 no 2010-10-16T02:02:22 but 2010-10-16T02:02:39 you can calculate the cosine between two vectors easily 2010-10-16T02:03:19 then i just make sure that my potential path nodes are in the limit 2010-10-16T02:03:21 I struggled with how not to send fleets at a 90 degree angle, instead of forward. 2010-10-16T02:03:54 hmm 2010-10-16T02:04:16 *** yasith has quit IRC (Ping timeout: 252 seconds) 2010-10-16T02:04:30 Well, if you know the direction you want them to go, then just go there. 2010-10-16T02:04:51 Use A* 2010-10-16T02:05:01 That should find the shortest path. 2010-10-16T02:05:04 can ships change direction 2010-10-16T02:05:07 true 2010-10-16T02:05:11 Epona: no 2010-10-16T02:05:29 Epona: it's different from Galcon 2010-10-16T02:05:48 that would be so epic 2010-10-16T02:06:03 I don't think games would ever finish 2010-10-16T02:17:14 argh i hate debugging 2010-10-16T02:17:29 *** yellowseed has quit IRC (Quit: yellowseed) 2010-10-16T02:18:03 *** yellowseed has joined #aichallenge 2010-10-16T02:18:29 *** McLeopold has quit IRC (Quit: Leaving) 2010-10-16T02:22:28 so glad ships can't change direction 2010-10-16T02:22:45 *** yasith has joined #aichallenge 2010-10-16T02:22:51 yes, that would be horrible 2010-10-16T02:22:56 strategy wise 2010-10-16T02:23:06 yeah 2010-10-16T02:23:08 the space is so wide open already 2010-10-16T02:23:28 it would kind of be cool if growth proportional rather than constant 2010-10-16T02:23:48 proportional to what? 2010-10-16T02:23:53 current pop I guess 2010-10-16T02:23:56 number of ships on the planet 2010-10-16T02:24:06 but i think you might end up with close starts always being draws 2010-10-16T02:24:07 that would give a huge disadvantage to the trailer though.. 2010-10-16T02:24:12 well that would certainly change things... 2010-10-16T02:25:29 this tree bot is giving me worlds of pain to even get it up and running 2010-10-16T02:25:47 it wont even generate the tree without crashing at this point :| 2010-10-16T02:26:13 damn 2010-10-16T02:26:14 that sucks 2010-10-16T02:26:32 i only started it yesterday, i'll get it there, i just hate fixing things 2010-10-16T02:26:45 a tree bot will take a few days... 2010-10-16T02:27:29 I had an epic idea for how to improve my tree bot, but I doubt I can finish the calcs in 1 sec... 2010-10-16T02:28:06 i think the code is to the point of testing once i get bugs ironed out 2010-10-16T02:28:22 it's a complete mess from me changing things to calculate it for either player though 2010-10-16T02:28:47 ah, are you modeling the opponent on a tree too? 2010-10-16T02:29:02 i'm doing minimax, i need them to make moves :P 2010-10-16T02:29:21 well, approximate minimax, can't check every move at every turn 2010-10-16T02:29:25 right 2010-10-16T02:29:43 so the the question is what level of approx to choose for the opponent so things don't get out of hand... 2010-10-16T02:29:58 to start with each player will be doing under 10 turns each 2010-10-16T02:30:01 i will see how that goes 2010-10-16T02:31:11 i may even just run 2 layers (moves for each player in each layer) then run those terminal nodes with one move per player to try and evaluate those nodes 2010-10-16T02:35:37 *** otherAntimatroid has joined #aichallenge 2010-10-16T02:40:09 *** Snowplan has joined #aichallenge 2010-10-16T02:53:23 *** sigh has joined #aichallenge 2010-10-16T02:58:47 *** Epona has quit IRC (Quit: Leaving) 2010-10-16T03:01:32 Top 10 players: dmj111(3612), MariusCG(3481), deepblue(3414), Hazard(3363), felixcoto(3314), ofrias(3312), Alvraro(3310), savchenko(3304), ruilov(3300), rsergio(3286) 2010-10-16T03:02:22 *** E323|ZzzZz is now known as Error323 2010-10-16T03:06:48 http://www.ai-contest.com/visualizer.php?game_id=5908617 muggle stole that one right out from under me 2010-10-16T03:08:22 *** skandy_ has quit IRC (Ping timeout: 265 seconds) 2010-10-16T03:09:21 cool match 2010-10-16T03:27:04 except the part where i lost :P 2010-10-16T03:32:15 *** JCS^ has quit IRC (Quit: \x00) 2010-10-16T03:33:08 *** yasith has quit IRC (Ping timeout: 255 seconds) 2010-10-16T03:44:53 *** yasith has joined #aichallenge 2010-10-16T03:46:07 hm, got 3 games in a row in 5 minutes at the official server 2010-10-16T03:46:23 that's some fluctiation 2010-10-16T03:49:07 *** Snowplan has left #aichallenge 2010-10-16T03:53:21 *** jahn has joined #aichallenge 2010-10-16T03:58:16 woah 2010-10-16T04:01:33 Top 10 players: dmj111(3543), MariusCG(3456), deepblue(3412), Hazard(3344), felixcoto(3295), ofrias(3294), Alvraro(3286), ruilov(3285), savchenko(3284), Manwe56(3272) 2010-10-16T04:06:35 *** jahn has quit IRC (Quit: Page closed) 2010-10-16T04:08:48 *** nikre has joined #aichallenge 2010-10-16T04:12:15 *** Mathnerd314 has quit IRC (Quit: ChatZilla 0.9.86-rdmsoft [XULRunner 1.9.2.8/20100722155716]) 2010-10-16T04:25:59 *** tapwater has quit IRC (Ping timeout: 276 seconds) 2010-10-16T04:28:31 *** tapwater has joined #aichallenge 2010-10-16T04:31:07 *** tapwater has quit IRC (Client Quit) 2010-10-16T04:39:18 whoa, vortex.ca7235 might not be half bad 2010-10-16T04:41:02 should go to bed before it starts losing...lol 2010-10-16T04:41:12 savor the false victory 2010-10-16T04:46:34 *** jukkerknott has quit IRC (Quit: Leaving) 2010-10-16T04:47:15 i just woke up 2010-10-16T04:47:22 good morning 2010-10-16T04:47:47 good night 2010-10-16T04:48:00 i think i am having hangover of the year 2010-10-16T04:48:08 ugh 2010-10-16T04:48:11 I hate hangovers 2010-10-16T04:48:16 um... 2010-10-16T04:48:17 i love 2010-10-16T04:48:21 (NOT) 2010-10-16T04:48:32 which you would think would mean I don't get drunk... but no... 2010-10-16T04:48:38 just opened new beer can 2010-10-16T04:48:50 damn, took a loss 2010-10-16T04:48:53 this has to fix something 2010-10-16T04:48:57 to the blasted oldman 2010-10-16T04:49:26 how old are you? 2010-10-16T04:49:36 lol.. old 2010-10-16T04:49:39 just... curious :D 2010-10-16T04:49:43 38 2010-10-16T04:49:44 i'm 25 2010-10-16T04:49:46 ok. :) 2010-10-16T04:50:24 did you think I was old or young? 2010-10-16T04:50:48 i was just thinking that line 'to the blasted oldman' 2010-10-16T04:50:50 22, youngest :) 2010-10-16T04:51:09 frig, here comes my losing streak... 2010-10-16T04:51:33 i'm curious how old oldman is 2010-10-16T04:51:53 hmm... 2010-10-16T04:52:04 my winning bot is still on my desk as pseudo code 2010-10-16T04:52:11 half there, half on my head 2010-10-16T04:52:13 has he ever been on chat? 2010-10-16T04:52:14 my winning bot is crashing 2010-10-16T04:52:20 don't think s 2010-10-16T04:52:21 my winning bot is losing 2010-10-16T04:52:29 i am still writing it 2010-10-16T04:52:39 i think i will never write mine 2010-10-16T04:52:46 ooh, RageBot, I can beat him... 2010-10-16T04:52:52 my bot has a serious flaw that I know how to fix 2010-10-16T04:52:53 i never lose to rage bot 2010-10-16T04:53:02 me either 2010-10-16T04:53:03 i would like to write a learning AI but can't because i can't write into a file 2010-10-16T04:53:14 what do you mean? 2010-10-16T04:53:30 rabidus: you can train one like that before the contest 2010-10-16T04:53:33 umm... i don't know whats the real algorithm name, or... hmm.. 2010-10-16T04:53:36 i plan to train parameters like that 2010-10-16T04:53:40 antimatroid: no i dont mean that 2010-10-16T04:53:50 i mean ... hmm 2010-10-16T04:53:51 then how? 2010-10-16T04:54:06 you wont know who your opponent is during the final contest, so not updating for specific players 2010-10-16T04:54:26 a bot who plays 1000000000000 (A LOTS OF GAMES) and kinda learns from previous plays what went wrong.. 2010-10-16T04:54:29 any estimation of the opponent really needs to be done per game 2010-10-16T04:54:32 hard to tell 2010-10-16T04:54:50 honolulu has a crazy smart bot 2010-10-16T04:54:57 when you don't actually can't write english properly 2010-10-16T04:55:33 -"actually can't" 2010-10-16T04:55:43 remember, i'm on hangover drink right now 2010-10-16T04:56:38 night all 2010-10-16T04:56:49 night. 2010-10-16T04:59:38 *** bhasker has quit IRC (Read error: Connection reset by peer) 2010-10-16T05:01:34 Top 10 players: dmj111(3553), MariusCG(3465), deepblue(3420), Hazard(3356), ofrias(3315), felixcoto(3306), ruilov(3292), Alvraro(3292), savchenko(3291), rsergio(3281) 2010-10-16T05:06:54 *** bhasker has joined #aichallenge 2010-10-16T05:13:46 *** Azrathud has joined #aichallenge 2010-10-16T05:29:13 why does expandbot java file has bullybot class definition in it? 2010-10-16T05:38:12 *** hellman has quit IRC (Read error: Connection reset by peer) 2010-10-16T05:40:23 *** Frontier has joined #aichallenge 2010-10-16T05:41:21 *** Bobng has quit IRC (Read error: Connection reset by peer) 2010-10-16T05:47:18 *** mceier has joined #aichallenge 2010-10-16T05:47:56 *** choas has joined #aichallenge 2010-10-16T05:50:29 nikre: just a bug in starter package 2010-10-16T05:51:04 ok 2010-10-16T05:51:05 *** Appleman1234 has quit IRC (Ping timeout: 245 seconds) 2010-10-16T05:51:36 *** Appleman1234 has joined #aichallenge 2010-10-16T05:55:56 *** gat3way has joined #aichallenge 2010-10-16T05:57:49 where can i find game specs? 2010-10-16T05:58:27 what is the ratio of fleet increase in a planet compared to its size? or radius? 2010-10-16T05:58:42 *** yasith has quit IRC (Ping timeout: 265 seconds) 2010-10-16T05:59:10 what is the speed of fleet and the speed of fleet increase in a planet? 2010-10-16T06:00:30 http://www.ai-contest.com/specification.php 2010-10-16T06:01:34 Top 10 players: dmj111(3511), MariusCG(3471), deepblue(3426), Hazard(3369), ofrias(3322), felixcoto(3308), ruilov(3300), Alvraro(3300), Manwe56(3293), savchenko(3291) 2010-10-16T06:03:51 how is the ELO score calculated? 2010-10-16T06:04:24 *** boegel has joined #aichallenge 2010-10-16T06:05:27 nikre: The planet size is the same as its growth rate and ranges from 0 to 5. 2010-10-16T06:05:59 ok 2010-10-16T06:11:55 *** yasith has joined #aichallenge 2010-10-16T06:26:43 what was the server stats url ? 2010-10-16T06:27:08 I mean the one that displays number of games/min and the number of online servers 2010-10-16T06:28:59 http://ai-contest.com/game_stats.php 2010-10-16T06:30:14 thanks 2010-10-16T06:30:37 *** phreeza has joined #aichallenge 2010-10-16T06:30:47 *** nikre has left #aichallenge ("Leaving") 2010-10-16T06:44:28 *** BtbN has joined #aichallenge 2010-10-16T06:48:41 *** Olathe has joined #aichallenge 2010-10-16T06:57:44 *** Bobng has joined #aichallenge 2010-10-16T07:00:29 *** Migi32 has joined #aichallenge 2010-10-16T07:01:35 Top 10 players: dmj111(3509), MariusCG(3466), deepblue(3422), Hazard(3369), ofrias(3324), Alvraro(3317), felixcoto(3303), ruilov(3298), Manwe56(3295), fglider(3292) 2010-10-16T07:02:36 hi 2010-10-16T07:10:43 hi 2010-10-16T07:28:00 *** Utkarsh has quit IRC (Ping timeout: 276 seconds) 2010-10-16T07:29:37 how goes it 2010-10-16T07:30:27 a whole day of debugging 2010-10-16T07:30:32 i'm ready to rage quite the universe 2010-10-16T07:30:36 what about you? 2010-10-16T07:30:39 *** Appleman1234 has quit IRC (Quit: Leaving) 2010-10-16T07:31:01 any word on if/when the low number of games on the server would be fixed? 2010-10-16T07:31:03 the usual work day and couldn't sleep 2010-10-16T07:31:06 so fixing bugs in my bot 2010-10-16T07:31:36 delt0r_: no idea sorry, tcp <3 2010-10-16T07:31:46 *** ratatata has joined #aichallenge 2010-10-16T07:31:49 right 2010-10-16T07:31:52 i'm working on an entirely new bot, so it's still constantly crashing 2010-10-16T07:31:52 thx 2010-10-16T07:32:25 it will make a move now tough, which was better than no moves before 2010-10-16T07:34:11 *** Utkarsh has joined #aichallenge 2010-10-16T07:36:07 5 games per minute is actually quite alright for 1 server, considering if everyone would use the 1 second we'd end up with 2000 seconds cpu time, which is 33 minutes 2010-10-16T07:36:56 Migi32: yes... but with more than 1000 players... its not really enough... eps with mulitcore and all that... 2010-10-16T07:36:59 so as the contest progresses and people use more cpu time, I guess it's more logical that the number of games per minute drops 2010-10-16T07:37:56 the one server is also the web server 2010-10-16T07:38:00 running very modest hardware 2010-10-16T07:38:07 so don't blame it if it is slow. 2010-10-16T07:38:10 which is a problem 2010-10-16T07:38:16 delt0r, so what do we do? deactivate everyone who hasn't logged in for > 2 weeks? buy more servers? 2010-10-16T07:38:33 Migi32: perhaps both? 2010-10-16T07:38:46 Its not a personal attack 2010-10-16T07:38:57 I undersand universitys... i work at one 2010-10-16T07:39:06 You need to remove all bots that crash a lot 2010-10-16T07:39:21 and you need j3camero to be around 2010-10-16T07:39:26 so i can offer him a vps. 2010-10-16T07:39:31 hmmm, i just did a rather depressing calculation 2010-10-16T07:39:34 and i think some kind of "login" keep alivie counter 2010-10-16T07:39:47 antimatroid, what is that? 2010-10-16T07:40:14 if you branch 5 moves per turn, by the end of the third layer you have 5^(25) terminal nodes 2010-10-16T07:40:35 that's almost 3*10^(17) nodes 2010-10-16T07:40:42 that's never going to happen 2010-10-16T07:40:43 welcome to exponential growth!@ 2010-10-16T07:41:09 Its just at this point with so few games played, i won't prioritize my bot... and do other things instead 2010-10-16T07:41:10 5^25? Isn't it more like 5^6? 2010-10-16T07:41:12 wait, i may have been retarded 2010-10-16T07:41:25 or what do you mean by "third layer"? 2010-10-16T07:41:29 oh yeah, here did you get 5^25 from? 2010-10-16T07:41:40 yeah, nevermind 2010-10-16T07:41:50 i went over the top aha 2010-10-16T07:42:02 haha 2010-10-16T07:42:09 yea, it's 5^6, which is "only" 15625 2010-10-16T07:43:18 i am getting less than 1000 processes of states each turn atm, but i haven't gone through and fixed stupid passing of data either 2010-10-16T07:43:35 only 1000? 2010-10-16T07:43:36 i'm also only picking 1 turn per state atm too 2010-10-16T07:43:50 yes, i think it's due to passing data around stupidly though 2010-10-16T07:44:01 ah fair enough 2010-10-16T07:44:06 *** andrew-101 has joined #aichallenge 2010-10-16T07:44:14 for the first time in my life i made a tree without memory leaks 2010-10-16T07:44:23 haha 2010-10-16T07:44:49 my heart sunk when i made that wrong calculation before 2010-10-16T07:45:12 it didn't feel right though, cause of chess ai's etc. 2010-10-16T07:45:21 why on earth were you raising anything to the 25th power? 2010-10-16T07:45:34 5^(5^5) 2010-10-16T07:45:41 lol 2010-10-16T07:45:42 don't ask me why i came up with that, i have no idea 2010-10-16T07:45:44 ... 2010-10-16T07:45:49 that just raises further questions 2010-10-16T07:45:57 *** tty1 has quit IRC (Read error: Connection reset by peer) 2010-10-16T07:46:00 5^(5^5) is 5^3125 2010-10-16T07:46:14 which is a whole lot more than 5^25 2010-10-16T07:46:17 yeah 2010-10-16T07:46:25 this is what happens to pure maths students, they forget how to work with actual numbers 2010-10-16T07:46:51 i'm not even allowed calculators in most of my maths exams anymore :| 2010-10-16T07:47:05 *** tty1 has joined #aichallenge 2010-10-16T07:47:20 well, makes sense for some subjects 2010-10-16T07:47:31 yep, not really any numerical questions 2010-10-16T07:47:33 I let students have cals in my exams... but if they think they need them... they did it wrong 2010-10-16T07:47:44 delt0r_: what do you teach> 2010-10-16T07:47:46 ? 2010-10-16T07:47:56 Physics and undergrad math 2010-10-16T07:48:00 programming 2010-10-16T07:48:02 dlet0r: that's the way exams should be (possibly minus the allowing the calculator) 2010-10-16T07:48:11 yea, here too. In my first year as an undergrad, we could use computers for the exams of almost every course (and do algebraic stuff the easy way), but as time progresses we can use them less and less 2010-10-16T07:48:12 these days just population genetics 2010-10-16T07:48:18 wow, that's quite a broad range 2010-10-16T07:48:21 although i'm talking analysis and algebra classes where the calculator is just useless 2010-10-16T07:48:52 its what happens to physici folk that don't do global warming or LHC 2010-10-16T07:49:04 *** tty1 has quit IRC (Read error: Connection reset by peer) 2010-10-16T07:50:11 *** tty1 has joined #aichallenge 2010-10-16T07:51:04 delt0r, most calculators have memory to store things like formulas etc, why do you think students want them so bad ;) 2010-10-16T07:51:24 not to mention programmable ones 2010-10-16T07:51:29 Migi32: My tests (if the universty will let me) are all open book 2010-10-16T07:51:38 I don't like testing memory 2010-10-16T07:51:59 however network phones are an issue 2010-10-16T07:52:33 using friends on online networks to cheat is becoming a problem 2010-10-16T07:52:59 do calcs have wireless networks nowadays? 2010-10-16T07:53:06 Iphones doo 2010-10-16T07:53:10 etc 2010-10-16T07:53:20 you shouldn't allow iphones then 2010-10-16T07:53:25 or any phone for that matter 2010-10-16T07:53:32 well that is a university policiy problem 2010-10-16T07:53:45 they are baning them now 2010-10-16T07:53:56 when I was at the university 2010-10-16T07:54:03 a student was caught cheating 2010-10-16T07:54:13 with a very small speaker stuck into his ear 2010-10-16T07:54:19 *** tty1 has quit IRC (Read error: Connection reset by peer) 2010-10-16T07:54:25 but some students claim they need it for the "language" to English dictionary 2010-10-16T07:54:25 wtf, i have never heard of a university allowing a phone as a calculator 2010-10-16T07:54:26 he went into the hospital because he was unable to get it out of his ear 2010-10-16T07:54:41 my uni allows lecturers to specify different kinds of calculators that students are allowed 2010-10-16T07:54:52 you wont get much more than a scientific calc in a maths exam if that 2010-10-16T07:55:00 *** tty1 has joined #aichallenge 2010-10-16T07:55:08 gat3way, haha teached him a lesson :) 2010-10-16T07:55:08 delt0r_: if they need one, then it should be a paper dictionary 2010-10-16T07:55:26 anyway... university's that i worked for are pretty slack at taking undergrad cheating seriously 2010-10-16T07:55:26 having an ipone for that is completely ridiculous 2010-10-16T07:55:32 :( 2010-10-16T07:55:34 sigh: thats what i said 2010-10-16T07:56:28 the uni get the fees either from the government or from the students for each student in the class... so they don't want to throw them out 2010-10-16T07:56:41 yeah, that sucks 2010-10-16T07:56:54 same here. 2010-10-16T07:57:05 we used to have entrance exams before you could enter 2010-10-16T07:57:10 can anyone remember the link to some random forum where they had a long discussion about the planetwars planet scoring algorithms ? 2010-10-16T07:57:13 but I always make it a point to fail them and not let them into my classes... 2010-10-16T07:57:13 now they just moved those to 3 weeks into the course 2010-10-16T07:57:21 *** boegel has quit IRC (Ping timeout: 245 seconds) 2010-10-16T07:57:38 gah, it wont even crash consistently atm 2010-10-16T07:57:53 antimatroid, what language do you use? 2010-10-16T07:57:57 c++ 2010-10-16T07:58:14 but i don't have much teaching on my current contract... :D 2010-10-16T07:58:23 antimatroid, using the gcc compiler? 2010-10-16T07:59:15 antimatroid, you may already know this, but it helps a ton to define _GLIBCXX_DEBUG and _GLIBCXX_DEBUG_PEDANTIC as gcc command-line arguments 2010-10-16T07:59:45 checks for when you try to access any STL container at an out-of-scope position 2010-10-16T07:59:54 out-of-bounds* 2010-10-16T08:00:34 meh, i'm lazy at doing things like that 2010-10-16T08:00:57 have you got very far at implementing the mixed strategy nash equilibria idea? 2010-10-16T08:01:19 who's doing that? Migi32? 2010-10-16T08:01:23 antimatroid, simplex: done. nash equilibria: done. However, turns out it's way too slow... 2010-10-16T08:01:31 ah, bugger 2010-10-16T08:01:36 Top 10 players: dmj111(3521), MariusCG(3476), deepblue(3468), Hazard(3381), ofrias(3329), Alvraro(3317), ruilov(3313), felixcoto(3311), savchenko(3307), Manwe56(3301) 2010-10-16T08:01:45 simplex is too slow? 2010-10-16T08:01:51 i was still considering that for move selection 2010-10-16T08:02:04 ah yes for small problem's it's quite fast 2010-10-16T08:02:21 first move? :P 2010-10-16T08:02:22 so for nodes where each player has like < 20 options, it's fast enough 2010-10-16T08:02:38 how many games do my bot play daily on the average? 2010-10-16T08:03:00 hundreds if you play on tcp 2010-10-16T08:03:06 thousands if you kick peoples ass 2010-10-16T08:03:07 for each 50 options however, it uses about 0.15 seconds, for 150 you hit 1 second 2010-10-16T08:03:23 on tcp? 2010-10-16T08:03:31 !tcp 2010-10-16T08:03:32 antimatroid: tcp = http://www.benzedrine.cx/planetwars/ 2010-10-16T08:03:41 oh that link needs to be updated 2010-10-16T08:04:00 !learn tcp as http://72.44.46.68/ 2010-10-16T08:04:01 Migi32: Job's done. 2010-10-16T08:04:12 !give gat3way tcp 2010-10-16T08:04:13 Migi32: I'm sorry Dave, I cannot give. 2010-10-16T08:04:26 !tcp 2010-10-16T08:04:27 Migi32: tcp = (#1) http://www.benzedrine.cx/planetwars/, or (#2) http://72.44.46.68/ 2010-10-16T08:04:32 !rainbow YAY 2010-10-16T08:04:32 antimatroid: YAY 2010-10-16T08:04:45 this does not go into the rankings if I understand correctly? 2010-10-16T08:05:26 tcp has its own rankings 2010-10-16T08:05:29 i wanted to tell it to learn lemonparty as tcp, but didn't that that'd go down too well aha 2010-10-16T08:05:31 http://72.44.46.68/getratings 2010-10-16T08:05:51 !lemonparty 2010-10-16T08:05:52 Migi32: You have no gotten any error messages recently, so here's a random one just to let you know that we care. 2010-10-16T08:06:03 oldman2 is a frustrating bot 2010-10-16T08:06:16 ok thanks 2010-10-16T08:06:17 i haven't played it yet 2010-10-16T08:06:27 i hope to beat everyone once i get a bot up and running again 2010-10-16T08:10:18 ./cry this oldman2 bot keeps crushing my bot /sob 2010-10-16T08:19:39 *** VictorRG has joined #aichallenge 2010-10-16T08:21:10 is there some way to visualize the tcp games? 2010-10-16T08:21:58 go to http://72.44.46.68/getratings, find your username in the list, and click it ? 2010-10-16T08:22:05 click your bot in the ratings list, then click "view game" 2010-10-16T08:22:17 thanks 2010-10-16T08:23:54 *** JensT1 has joined #aichallenge 2010-10-16T08:29:02 krokok whats your current bots name 2010-10-16T08:29:48 hm, "krokokrusa" 2010-10-16T08:29:52 i'm not on tcp right now 2010-10-16T08:30:32 * krokokrusa is in the process of implementing tree-searching strategy 2010-10-16T08:31:04 similarly to lots of people here, it seems 2010-10-16T08:31:44 *** ratatata has quit IRC (Ping timeout: 265 seconds) 2010-10-16T08:34:01 *** jesionaj` has quit IRC (Ping timeout: 245 seconds) 2010-10-16T08:35:01 sigh: UCT is getting there... i implemented a super dumb MC strategy, which actually doesn't do extremely bad on the tcp server, so that is making me cautiously optimistic 2010-10-16T08:36:03 phreeza: cool, is that phreeza on tcp? 2010-10-16T08:36:53 sigh: nope :) i called it dimwit. elo is very low because i was debugging a lot in the beginning 2010-10-16T08:37:43 its nowhere close to being good, but it does do sniping for example, which is pretty cool since i didn't put that in explicitly 2010-10-16T08:37:47 interesting, I'll have a look at the games 2010-10-16T08:39:14 haha, very chaotic 2010-10-16T08:41:15 I would love for this bot to make it to top, everyone will wonder wtf it is doing :P 2010-10-16T08:41:55 yea that would be awesome :) 2010-10-16T08:42:51 it would be cool if there were a tcp server with say 30s timeout 2010-10-16T08:42:58 would like to see how it plays then 2010-10-16T08:43:11 so why do you even consider partial moves phreeza 2010-10-16T08:43:18 i see that your tree searching bot 2010-10-16T08:43:24 is sending a few ships at neutrals 2010-10-16T08:43:39 bhasker: yea as i said its a very dumb first shot 2010-10-16T08:43:48 hmm 2010-10-16T08:43:53 so no pruning of bad moves? 2010-10-16T08:43:54 it just draws random numbers 2010-10-16T08:43:55 i guess 2010-10-16T08:43:59 ah 2010-10-16T08:44:08 and sees what is the best one... 2010-10-16T08:44:17 yeah, next step should give smarter random moves :) 2010-10-16T08:44:25 i was just stress testing my tree implementation ;) 2010-10-16T08:44:31 :) 2010-10-16T08:44:44 just found out about transposition tables this morning though... 2010-10-16T08:44:56 but i think maybe its not worth it in this case 2010-10-16T08:45:06 I wouldn't think so 2010-10-16T08:46:26 how many game simulations does it run in a second? 2010-10-16T08:47:16 right now about 10k but i think i have a way of making it about 10 times faster 2010-10-16T08:47:28 thats on a 2GHz C2D 2010-10-16T08:47:38 my bot is too weak :( 2010-10-16T08:47:47 i see... that's a bit lower than I would have expected :( 2010-10-16T08:48:46 i was hoping for more, too... 2010-10-16T08:49:53 right now the game engine i use for simulations is just a 1:1 translation of the official Java engine... i think i can tweak it a bit, but i have serious doubts i will ever get more than maybe 200k nodes 2010-10-16T08:50:56 oh, so each fleet is an object and all that? 2010-10-16T08:51:03 yea :) 2010-10-16T08:51:22 yeah, you'll get a lot more performance if you improve your data structures 2010-10-16T08:52:28 *** otherAntimatroid has left #aichallenge 2010-10-16T08:52:35 another cool thing would be if i could reuse branches from previous rounds, but that only works if i guessed the opponents moove exactly, of course 2010-10-16T08:53:10 branching factor is so high that the benefit will be negligible, imo 2010-10-16T08:54:39 if I was running a simulation, to store fleets I would set up an array indexed [planet][turn][player]... one off allocation of array size [NUM_PLANETS][MAX_TURNS][2] 2010-10-16T08:56:03 (where the elements of the array are ship counts) 2010-10-16T08:56:37 yea that sounds about right... 2010-10-16T08:57:02 that should be way more than 10x improvement 2010-10-16T08:58:23 *** yellowseed has quit IRC (Ping timeout: 240 seconds) 2010-10-16T08:59:35 i'll probably make the whole uct thing work first and then i'll see about improving the engine. 2010-10-16T09:00:44 fair enough 2010-10-16T09:01:36 Top 10 players: dmj111(3529), MariusCG(3485), deepblue(3476), Hazard(3390), ofrias(3338), Alvraro(3326), ruilov(3322), felixcoto(3319), savchenko(3314), Manwe56(3314) 2010-10-16T09:04:34 thanks for the input :) are you planning on trying anything along those lines, too? 2010-10-16T09:07:42 I don't know... probably not, I am very lazy 2010-10-16T09:13:01 *** delt0r___ has joined #aichallenge 2010-10-16T09:14:48 *** delt0r_ has quit IRC (Ping timeout: 260 seconds) 2010-10-16T09:14:54 *** Azrathud has quit IRC (Ping timeout: 250 seconds) 2010-10-16T09:21:15 *** sigh has quit IRC (Remote host closed the connection) 2010-10-16T09:21:31 *** jorge_ has joined #aichallenge 2010-10-16T09:28:28 *** andrew-101 has quit IRC (Ping timeout: 240 seconds) 2010-10-16T09:28:38 *** Bobng has quit IRC (Remote host closed the connection) 2010-10-16T09:29:10 *** Bobng has joined #aichallenge 2010-10-16T09:30:26 sigh 2010-10-16T09:30:32 http://72.44.46.68/getplayer?player=malazan0.3 2010-10-16T09:30:35 my latest bot 2010-10-16T09:32:44 *** jorge_ has quit IRC (Remote host closed the connection) 2010-10-16T09:34:36 bhasker: its a bit inefficient 2010-10-16T09:34:38 http://72.44.46.68/canvas?game_id=517270 2010-10-16T09:34:56 yea was just watching that game 2010-10-16T09:35:00 sends ships in a triangle, routing them through 3 planets to get to a planet which is closer 2010-10-16T09:35:03 top left 2010-10-16T09:35:29 hmm hold on let me see 2010-10-16T09:35:30 from turn 56 onwards 2010-10-16T09:35:58 otherwise its pretty good 2010-10-16T09:36:18 i am not sure which routing you are talking about 2010-10-16T09:36:32 at turn 56 i have a bunch of ships headed to the top left 2010-10-16T09:36:34 planet 2010-10-16T09:36:55 yeah, the bot routes ships up to the top planet 2010-10-16T09:37:03 from the bottomish one 2010-10-16T09:37:06 through another one 2010-10-16T09:37:14 then the top planet sends them down to a smaller planet 2010-10-16T09:37:22 which is close to the original planet 2010-10-16T09:37:32 making the round trip pointless? 2010-10-16T09:37:58 ah thats because i stage to that planet 2010-10-16T09:38:03 those ships were on their way 2010-10-16T09:38:11 before i figured that i needed to save that corner planet 2010-10-16T09:38:22 ahh 2010-10-16T09:38:34 it works well mostly 2010-10-16T09:38:37 but in some split maps 2010-10-16T09:38:39 like this 2010-10-16T09:38:44 it does stupid things like that 2010-10-16T09:38:51 frankly it should have just given up the corner one 2010-10-16T09:38:57 and gone for the bigger ones below 2010-10-16T09:41:39 *** Bobng has quit IRC (Quit: Leaving) 2010-10-16T09:51:17 *** Migi32 has quit IRC (Ping timeout: 265 seconds) 2010-10-16T10:01:37 Top 10 players: dmj111(3546), MariusCG(3495), deepblue(3485), Hazard(3395), ofrias(3337), ruilov(3335), felixcoto(3329), Alvraro(3326), savchenko(3317), Manwe56(3304) 2010-10-16T10:11:09 *** hellman has joined #aichallenge 2010-10-16T10:39:37 *** Utkarsh has quit IRC (Ping timeout: 265 seconds) 2010-10-16T10:45:36 *** HakanD has joined #aichallenge 2010-10-16T10:54:11 *** yellowseed has joined #aichallenge 2010-10-16T11:01:37 Top 10 players: dmj111(3544), deepblue(3488), MariusCG(3485), Hazard(3394), ofrias(3340), ruilov(3339), Alvraro(3332), felixcoto(3325), savchenko(3319), Manwe56(3311) 2010-10-16T11:03:47 *** smellyhippy has quit IRC (Ping timeout: 265 seconds) 2010-10-16T11:06:11 *** teurastaja has joined #aichallenge 2010-10-16T11:11:50 *** GreenTea has joined #aichallenge 2010-10-16T11:12:21 hi all 2010-10-16T11:13:39 *** smellyhippy has joined #aichallenge 2010-10-16T11:13:39 *** smellyhippy has joined #aichallenge 2010-10-16T11:13:55 Hi GreenTea 2010-10-16T11:14:45 do you know something about this http://ai-contest.com/forum/viewtopic.php?f=18&t=877&start=10 ? 2010-10-16T11:16:04 *** VictorRG has quit IRC (Ping timeout: 265 seconds) 2010-10-16T11:17:47 *** JensT1 has quit IRC (Ping timeout: 245 seconds) 2010-10-16T11:25:52 *** Mathnerd314 has joined #aichallenge 2010-10-16T11:27:06 *** narnach has joined #aichallenge 2010-10-16T11:29:22 *** JensT1 has joined #aichallenge 2010-10-16T11:34:21 *** JensT1 has quit IRC (Ping timeout: 276 seconds) 2010-10-16T11:38:00 *** Titankiller has joined #aichallenge 2010-10-16T11:39:08 can somebody help me with setting up this project in eclipse / 2010-10-16T11:39:29 ? 2010-10-16T11:46:19 *** Titankiller has quit IRC (Quit: Nettalk6 - www.ntalk.de) 2010-10-16T11:58:25 *** teurastaja has left #aichallenge ("Leaving") 2010-10-16T12:01:38 Top 10 players: dmj111(3542), deepblue(3486), MariusCG(3484), Hazard(3392), ofrias(3339), ruilov(3338), Alvraro(3330), felixcoto(3323), savchenko(3318), Manwe56(3310) 2010-10-16T12:02:01 *** Blkt has joined #aichallenge 2010-10-16T12:13:34 *** davidjliu has joined #aichallenge 2010-10-16T12:17:58 *** p4p4 has joined #aichallenge 2010-10-16T12:25:02 *** tapwater has joined #aichallenge 2010-10-16T12:49:33 *** Gudradain has joined #aichallenge 2010-10-16T12:54:46 *** HakanD has quit IRC (Ping timeout: 255 seconds) 2010-10-16T12:59:12 *** p4p4 has quit IRC (Quit: ChatZilla 0.9.84 [SeaMonkey 2.0a3/20090223135443]) 2010-10-16T13:01:40 Top 10 players: dmj111(3542), deepblue(3476), MariusCG(3475), Hazard(3379), ruilov(3328), Alvraro(3314), Manwe56(3308), savchenko(3307), felixcoto(3306), ofrias(3305) 2010-10-16T13:12:30 *** hellman has quit IRC (Remote host closed the connection) 2010-10-16T13:13:38 *** Azrathud has joined #aichallenge 2010-10-16T13:14:39 well my tree now works 2010-10-16T13:14:46 i just need to get it to pick good moves now 2010-10-16T13:15:51 Which one is harder? 2010-10-16T13:16:07 the latter 2010-10-16T13:16:56 i disagree 2010-10-16T13:17:02 i've already worked out how i'm going to do it 2010-10-16T13:17:14 ak10n: haven't seen you about in a while, are you still working on a bot? 2010-10-16T13:17:23 hey 2010-10-16T13:17:32 nah, he's minecrafting, lol 2010-10-16T13:17:42 haha 2010-10-16T13:17:52 i haven't touched it in a while but yes 2010-10-16T13:17:56 *** Cyndre has quit IRC (Quit: Leaving) 2010-10-16T13:18:00 i wrote a tree-searching bot a while back and gave up on it 2010-10-16T13:18:13 i'm pretty sure this will work quite well once i get it going 2010-10-16T13:18:20 i have built the heuristic side in quite well 2010-10-16T13:18:22 in fact i've written about five revisions to my bot since 0.4.1 and abandoned them before running them 2010-10-16T13:18:51 although the new dmj bot looks damn fierce 2010-10-16T13:19:02 i had a hard enough time getting over the older versions 2010-10-16T13:21:05 it takes a lot longer to see the result of a game now 2010-10-16T13:21:14 and i'm only giving it 0.3 seconds of tree generation :| 2010-10-16T13:23:00 I just finished my first tree searching bot but it is incredibly bad 2010-10-16T13:23:10 the stuff it does makes no sense at all 2010-10-16T13:23:11 :) 2010-10-16T13:24:17 i look forward to these results :P 2010-10-16T13:25:33 how many nodes do you get with 0.3s? 2010-10-16T13:25:42 Maybe it's so incredibly artificially intelligent that you can't pick out its reasons for making the moves it does. 2010-10-16T13:26:30 Azrathud: If it has become self aware and decided to commit suicide, then it is extremely efficient 2010-10-16T13:26:50 it just has a major but 2010-10-16T13:27:38 Uh, there's not but in your statement. 2010-10-16T13:27:43 no* 2010-10-16T13:29:53 Implied but? 2010-10-16T13:33:02 phreeza: i literally just got the tree working without crashing :P 2010-10-16T13:33:21 i have it making only 2 moves per node atm, one of which is to do nothing 2010-10-16T13:33:53 Azrathud: should there be a but? 2010-10-16T13:34:14 antimatroid: ah ok, for me every node is a move 2010-10-16T13:34:20 or rather, every edge 2010-10-16T13:34:42 or you mean every node has 2 children? 2010-10-16T13:35:33 phreeza: it just has a major but <- I think 'but' should refer to an actual 'but' in a sentence. 2010-10-16T13:36:04 oh... s/but/bug/ 2010-10-16T13:36:20 Ah. 2010-10-16T13:36:59 phreeza: yeah, current branching of 2 children per node 2010-10-16T13:37:03 but that will be ramped up 2010-10-16T13:37:11 i'm not sure i can really go above 5 if even that 2010-10-16T13:37:25 i can run "terminal" nodes if necessary 2010-10-16T13:37:40 I guess vi is popular around here, unless s/but/bug/ is a general regular expression. When I typed your name out, I realized it was probably fron dragon ball. 2010-10-16T13:38:22 s/but/bug/ is originally a sed command, i think 2010-10-16T13:38:35 but works in vim too 2010-10-16T13:39:14 i've been asked that about my nick before... came up with it before i had heard of dragonball 2010-10-16T13:39:24 Ah. 2010-10-16T13:40:46 antimatroid: i grow my tree probabilistically, sounds good in theory but turns out to suck, so far at least 2010-10-16T13:42:04 we should set up a tcp server for searching bots with a 30s time limit just to see what they come up with 2010-10-16T13:42:47 Why 30s? Is the other 30 required for something else? 2010-10-16T13:42:52 i'm not touching anything like that till i get this up and running 2010-10-16T13:43:03 i am happy with 1 second for the moment 2010-10-16T13:46:39 yea tree searching :) 2010-10-16T13:46:55 I would love to have an idea how to do one right 2010-10-16T13:49:19 anyone tried genetic? 2010-10-16T13:49:42 theres a forum post, but don't know if that ever went anywhere 2010-10-16T13:50:58 I did a genetic program once before, it was to separate a series of weight equally I think 2010-10-16T13:52:17 but well, for each iteration it found a solution and used some of the best solution with some mutation to get a new one. but I just don't see how this method can be applied to this game 2010-10-16T13:52:42 would it have to run the game to the end? or just 1 or a few turn is enough? 2010-10-16T13:52:48 http://ai-contest.com/forum/viewtopic.php?f=17&t=486 2010-10-16T13:53:22 languages in the top 100: Java (42), C++ (40), Python (11), PHP (2), and 1 each of C, Javascript, Haskell, Lisp, and Perl 2010-10-16T13:53:27 i think he is doing genetic programming to develop the actual bot 2010-10-16T13:53:49 Java, shocking... 2010-10-16T13:54:14 well, C/C++ is just one behind... 2010-10-16T13:54:20 people code what they know 2010-10-16T13:54:28 and they know Java and C/C++ 2010-10-16T13:54:31 true 2010-10-16T13:54:59 java winning, lame 2010-10-16T13:55:03 both of those languages are pretty much the same to me 2010-10-16T13:55:04 i'm just happy it's not only those, that the other languages do make an appearance 2010-10-16T13:55:09 wha 2010-10-16T13:55:19 i put Java in the C-family 2010-10-16T13:55:21 go java 2010-10-16T13:55:39 i'm going to stop talking :) no need for a flamewar 2010-10-16T13:55:48 :) 2010-10-16T13:56:04 phreeza: :P 2010-10-16T13:56:33 besides, the current leader is python :P 2010-10-16T13:56:46 \o/ 2010-10-16T13:56:46 470 elo points behind first one <----- 2010-10-16T13:57:56 *** Utkarsh has joined #aichallenge 2010-10-16T13:58:23 also, i counted those stats in python :) 2010-10-16T13:58:34 * Zannick returns to working on his C bot 2010-10-16T14:00:37 I'm playing with both python and C++. 2010-10-16T14:00:46 bots 2010-10-16T14:01:39 Top 10 players: dmj111(3543), deepblue(3489), MariusCG(3452), Alvraro(3343), Hazard(3336), felixcoto(3335), ofrias(3324), Manwe56(3322), savchenko(3322), fglider(3314) 2010-10-16T14:03:22 so did those daily game dumps ever get started? 2010-10-16T14:06:27 Azrathud: you mean you have one bot written in a combination of python and c++? or you have two bots, and you use the tcp server to test them out? 2010-10-16T14:09:17 yellowseed: I have two bots, but it's more like I'm just playing around on the c++ one and working on the python one. 2010-10-16T14:15:28 i just beat my normal heuristic bot without having even done much of what i want for move selection yet 2010-10-16T14:24:11 antimatroid: You are using a search now? on just a few move? 2010-10-16T14:24:13 moves 2010-10-16T14:24:15 *** RainCT has joined #aichallenge 2010-10-16T14:26:48 it's making a tree then running minimax on it 2010-10-16T14:26:54 no alpha beta pruning or anything though 2010-10-16T14:27:46 what was the rank of your normal heuristic bot? 2010-10-16T14:27:57 it's at like 40 atm 2010-10-16T14:28:06 it has been better than that though 2010-10-16T14:28:13 nice 2010-10-16T14:28:42 is anyone here using evolutionary computation to generate bots? 2010-10-16T14:42:25 *** rebelxt_ has joined #aichallenge 2010-10-16T14:48:04 foucist: I know there are one or two ppl doing this, but I doubt they will end up with a decent bot. The game is too complex to lean back and wait for some magic evolutionary steps. But they have fun doing so and learn a lot. 2010-10-16T14:49:05 *** Frontier has left #aichallenge 2010-10-16T14:50:23 foucist: also, i recall somebody used GA to fine tune parameters of his heuristics-based bot, not sure how successful though 2010-10-16T14:50:28 *** Frontier has joined #aichallenge 2010-10-16T14:52:15 antimatroid - you got your bot up yet? 2010-10-16T14:52:42 rwa: still got lots to do before that 2010-10-16T14:52:48 *** JCS^ has joined #aichallenge 2010-10-16T14:52:50 but i'm confident now that it'll be quite good 2010-10-16T14:53:01 it kicks my heuristic bots ass and i haven't even put in half the stuff yet 2010-10-16T14:53:22 awesome 2010-10-16T14:55:39 Would there be enough turns for evolutionary computation to be successful? 2010-10-16T14:57:15 *** bhasker has quit IRC (Ping timeout: 265 seconds) 2010-10-16T15:01:39 Top 10 players: dmj111(3550), deepblue(3492), MariusCG(3459), Hazard(3378), savchenko(3351), felixcoto(3338), kcm1700(3337), Alvraro(3334), ofrias(3329), Manwe56(3321) 2010-10-16T15:03:08 *** iris1 has joined #aichallenge 2010-10-16T15:04:00 *** bhasker has joined #aichallenge 2010-10-16T15:05:53 antimatroid: is your new bot using approach other than fugure simulation and heuristics? 2010-10-16T15:06:13 future* 2010-10-16T15:07:01 rebelxt_: it's a minimax bot 2010-10-16T15:08:01 antimatroid: nice! hope you can get it to do decent depth search before 1s time limit forces you to bail out 2010-10-16T15:08:35 a minimax bot will be all about restricting the search space to plausible and useful states 2010-10-16T15:09:21 and, of course, a good strength function 2010-10-16T15:11:01 rwa, agreed. also trying to figure this out.. 2010-10-16T15:12:26 i don't expec to get very deep 2010-10-16T15:12:35 but it seems to work well anyway 2010-10-16T15:14:00 cool - can't wait to see it on tcp 2010-10-16T15:14:15 it's 6am, i might need to go to bed soon :P 2010-10-16T15:14:45 Nice. 2010-10-16T15:16:37 antimatroid: are you in Australia? operating on red bull? ;) 2010-10-16T15:18:01 *** Frontier has left #aichallenge 2010-10-16T15:18:27 antimatroid: what's a ply for your minimax bot? 2010-10-16T15:18:28 *** boegel has joined #aichallenge 2010-10-16T15:18:40 *** Titankiller has joined #aichallenge 2010-10-16T15:19:16 ply? 2010-10-16T15:19:21 yes, i'm in tasmania 2010-10-16T15:20:36 antimatroid: is it thinking in terms of orders of fleet launches per turn? 2010-10-16T15:23:58 any body doing the coding on eclipse platform ? 2010-10-16T15:26:39 *** Accoun has quit IRC () 2010-10-16T15:27:19 mega1: i'm not giving out all specifics of what i;m doing :P 2010-10-16T15:27:32 fair enough 2010-10-16T15:28:13 *** boegel has quit IRC (Ping timeout: 245 seconds) 2010-10-16T15:29:07 Titankiller: yes 2010-10-16T15:33:20 GreenTea : tell me how you setup the ide 2010-10-16T15:39:18 *** Accoun has joined #aichallenge 2010-10-16T15:40:33 *** krokokrusa has left #aichallenge 2010-10-16T15:42:08 anybody here using git and the unofficial python starter pack? 2010-10-16T15:42:34 or just git in general? i'm wondering why people have 'master' and 'integration' branches pushed to github, instead of just pushing the 'master' branch. 2010-10-16T15:44:33 *** Titankiller has quit IRC (Ping timeout: 240 seconds) 2010-10-16T15:44:48 *** Naktibalda has joined #aichallenge 2010-10-16T15:46:24 *** Bobng has joined #aichallenge 2010-10-16T15:48:52 *** Titankiller has joined #aichallenge 2010-10-16T15:50:13 yellowseed: i use git every now and then but i have no idea what the aswer to your question is :) 2010-10-16T15:57:09 *** Titankiller has quit IRC (Quit: Nettalk6 - www.ntalk.de) 2010-10-16T15:58:13 yellowseed - why not put an integration branch on github? 2010-10-16T15:58:53 i don't know. why bother though? what's the semantic significance of integration vs. master? 2010-10-16T15:59:40 to be clear, i'm trying to learn best practices by studying conventional use. 2010-10-16T15:59:48 *** yasith has quit IRC (Ping timeout: 240 seconds) 2010-10-16T16:01:12 rwa: do you use integration branches differently from master branches? do you try to cherry-pick from integration into master, or vice versa? 2010-10-16T16:01:35 yellowseed: at some level they're just names that mean whatever people want them to mean... but there is some convention, I guess 2010-10-16T16:01:40 Top 10 players: dmj111(3559), deepblue(3540), MariusCG(3467), Hazard(3386), ofrias(3361), felixcoto(3349), Alvraro(3348), Manwe56(3335), savchenko(3331), fglider(3324) 2010-10-16T16:02:48 *** tapwater has quit IRC (Ping timeout: 240 seconds) 2010-10-16T16:03:42 I think typically people develop in master and then put "release quality" changes in the integration branch 2010-10-16T16:03:53 *** bhasker has quit IRC (Read error: Connection reset by peer) 2010-10-16T16:03:57 but it's not how I work in a single developer project 2010-10-16T16:04:32 I generally don't branch unless there's something particularly speculative that I want to try 2010-10-16T16:05:13 *** Titankiller has joined #aichallenge 2010-10-16T16:09:23 *** aaa has joined #aichallenge 2010-10-16T16:10:07 hello! how to get a login and pass for connection to tcp simulator? 2010-10-16T16:10:48 *** aaa has quit IRC (Client Quit) 2010-10-16T16:11:04 *** bhasker has joined #aichallenge 2010-10-16T16:12:13 *** yasith has joined #aichallenge 2010-10-16T16:14:05 aaa: as far as i know, you just specify any password the first time; after that, anybody who wants to use the same login has to use the same password. 2010-10-16T16:19:32 *** raedwulf has joined #aichallenge 2010-10-16T16:19:38 *** rebelxt_ has quit IRC (Quit: Page closed) 2010-10-16T16:19:45 it's amazing my bot wins at all with some of these bugs I find...lol 2010-10-16T16:20:25 *** Amanieu has joined #aichallenge 2010-10-16T16:37:53 *** tapwater has joined #aichallenge 2010-10-16T16:45:45 *** amstan has quit IRC (Read error: Connection reset by peer) 2010-10-16T16:46:00 *** amstan has joined #aichallenge 2010-10-16T16:46:00 *** ChanServ sets mode: +o amstan 2010-10-16T16:55:17 *** amstan_ has joined #aichallenge 2010-10-16T16:55:17 *** ChanServ sets mode: +o amstan_ 2010-10-16T16:56:43 *** amstan__ has joined #aichallenge 2010-10-16T16:56:43 *** ChanServ sets mode: +o amstan__ 2010-10-16T16:58:34 *** amstan has quit IRC (Ping timeout: 252 seconds) 2010-10-16T16:59:28 *** amstan_ has quit IRC (Ping timeout: 245 seconds) 2010-10-16T17:01:05 *** amstan__ is now known as amstan 2010-10-16T17:01:40 Top 10 players: dmj111(3595), deepblue(3506), MariusCG(3448), Hazard(3385), Manwe56(3384), Alvraro(3357), felixcoto(3349), ofrias(3348), fglider(3330), Polsky(3329) 2010-10-16T17:07:47 *** jesionaj has joined #aichallenge 2010-10-16T17:13:02 *** delt0r_ has joined #aichallenge 2010-10-16T17:14:53 *** delt0r___ has quit IRC (Ping timeout: 255 seconds) 2010-10-16T17:15:08 *** jaspervdj has quit IRC (Quit: NEVER GONNA GIVE YOU UP NEVER GONNA LET YOU DOWN) 2010-10-16T17:20:37 *** boegel has joined #aichallenge 2010-10-16T17:23:19 *** amstan has quit IRC (Ping timeout: 252 seconds) 2010-10-16T17:35:06 not too many people on tcp this evening 2010-10-16T17:35:22 come, join the fray 2010-10-16T17:37:05 was it revealed who is CCCF? 2010-10-16T17:39:27 *** BtbN has quit IRC (Quit: Verlassend) 2010-10-16T17:41:47 *** davidjliu has quit IRC (Quit: Page closed) 2010-10-16T18:00:10 *** davidjliu has joined #aichallenge 2010-10-16T18:01:40 Top 10 players: dmj111(3639), deepblue(3532), MariusCG(3476), Manwe56(3447), Hazard(3409), Alvraro(3384), felixcoto(3371), Polsky(3365), cfaftw(3351), savchenko(3351) 2010-10-16T18:08:11 *** choas has quit IRC (Quit: leaving) 2010-10-16T18:15:36 hmm, GenericBob didnt submit a new version since 28th of September, but his got climbed up from ~250 to 127 position. 2010-10-16T18:20:59 *** Sylph has quit IRC (Read error: Connection reset by peer) 2010-10-16T18:23:32 *** boegel has quit IRC (Quit: *poof!*) 2010-10-16T18:34:00 *** GreenTea has quit IRC (Ping timeout: 265 seconds) 2010-10-16T18:36:10 Heh. 2010-10-16T18:42:24 *** flavor has joined #aichallenge 2010-10-16T18:42:56 hi all 2010-10-16T18:44:02 *** Meatkat has joined #aichallenge 2010-10-16T18:44:15 yay cloud servers! 2010-10-16T18:57:13 *** flavor has quit IRC (Quit: Yaaic - Yet another Android IRC client - http://www.yaaic.org) 2010-10-16T19:00:04 Meatkat: are they back on? 2010-10-16T19:00:21 there are 2 up now, it seems 2010-10-16T19:00:41 neato 2010-10-16T19:00:57 http://ai-contest.com/game_stats.php 2010-10-16T19:01:18 Nice. 2010-10-16T19:01:41 Top 10 players: dmj111(3654), deepblue(3547), MariusCG(3515), Manwe56(3457), Hazard(3413), Alvraro(3408), Polsky(3385), kcm1700(3378), bix0r4ever(3377), felixcoto(3374) 2010-10-16T19:06:02 *** Meatkat has quit IRC (Ping timeout: 255 seconds) 2010-10-16T19:18:29 well it pass from 6 game per minute to like 50 2010-10-16T19:18:32 a lot better 2010-10-16T19:42:57 *** mceier has quit IRC (Quit: leaving) 2010-10-16T19:44:57 CrowBot is badass 2010-10-16T19:46:17 *** amstan has joined #aichallenge 2010-10-16T19:46:17 *** ChanServ sets mode: +o amstan 2010-10-16T19:47:28 *** Itkovian has joined #aichallenge 2010-10-16T19:51:50 *** gerald512k has joined #aichallenge 2010-10-16T19:54:33 hi 2010-10-16T19:54:48 working hard on the perfect algorythm ? ;) 2010-10-16T19:56:46 *** Meatkat has joined #aichallenge 2010-10-16T20:01:42 Top 10 players: dmj111(3780), deepblue(3602), MariusCG(3600), Hazard(3589), Alvraro(3524), fglider(3491), bix0r4ever(3480), savchenko(3479), Manwe56(3464), felixcoto(3454) 2010-10-16T20:05:20 gerald512k: yep, lots of people are 2010-10-16T20:05:30 gerald512k: how are you doing? 2010-10-16T20:05:57 some are even working on figuring out if such an algorithm exists :) 2010-10-16T20:08:12 well i am thinking to my bot too lol :) i started a weekago it is very fun! 2010-10-16T20:08:16 *** RainCT has quit IRC (Quit: good night) 2010-10-16T20:08:31 and so i can learn the basics of a programing language :) 2010-10-16T20:08:38 i'm still working on stuff but here's some test games http://72.44.46.68/getplayer?player=testbot 2010-10-16T20:08:58 uploaded my current tcp version to official servers now that cloud servers are up 2010-10-16T20:09:01 lets see how high it can go 2010-10-16T20:09:13 that bot is using minimax 2010-10-16T20:09:23 http://72.44.46.68/canvas?game_id=524797 is probably the better win so far 2010-10-16T20:09:26 what is this server ? it is unofficial ? 2010-10-16T20:09:32 yes 2010-10-16T20:09:52 very nice antimatroid 2010-10-16T20:09:58 its already beating 300+ elo rated bots 2010-10-16T20:10:11 it is kind agressive 2010-10-16T20:10:15 heck it even beat honolulu-19 2010-10-16T20:10:18 #1 rated bot 2010-10-16T20:10:19 bhasker: I have no idea what you mean. 2010-10-16T20:10:35 it will still get a fair bit better 2010-10-16T20:10:44 have you a lot of experience in programing? 2010-10-16T20:11:38 kinda crazy gam 2010-10-16T20:11:58 me? i'm an undergrad student, so no 2010-10-16T20:12:56 ok 2010-10-16T20:12:59 antimatroid: nice! 2010-10-16T20:13:13 my searching one is working now, too... more or less 2010-10-16T20:13:15 i decided to make this bot because i wanted to learn a programing language so i am learning java 2010-10-16T20:13:21 looks a lot less organized than yours 2010-10-16T20:13:28 and a lot weaker 2010-10-16T20:13:31 http://72.44.46.68/canvas?game_id=524976 2010-10-16T20:14:03 but theres a lot of room for improvement 2010-10-16T20:14:52 is there a topic on the forum where i can learn how to upload my bot on test serv? 2010-10-16T20:15:20 gerald512k: have you created an account? its quite straightforward 2010-10-16T20:15:34 *** sigh has joined #aichallenge 2010-10-16T20:15:36 i have account on official serv 2010-10-16T20:16:10 ah by test server you mean the tcp server we are posting here? 2010-10-16T20:16:12 http://72.44.46.68/canvas?game_id=524976 2010-10-16T20:16:13 this one? 2010-10-16T20:16:17 yes 2010-10-16T20:20:47 *** phreeza has quit IRC (Ping timeout: 260 seconds) 2010-10-16T20:21:56 how can i supply my bot to this one ? 2010-10-16T20:23:39 *** Itkovian has quit IRC (Ping timeout: 245 seconds) 2010-10-16T20:23:53 gerald512k: http://www.ai-contest.com/forum/viewtopic.php?f=18&t=424 2010-10-16T20:26:17 you don't upload (as far as i know) but play through a network connection from your own computer. 2010-10-16T20:33:08 hmm antimatrioid but i guess it doesn't do anything if its already winning 2010-10-16T20:33:13 so it will just wait out the 200 turns? 2010-10-16T20:33:31 cause a lot of games seem to run to 200 turns 2010-10-16T20:39:54 yeah, i tried to tell the payoff function not to do that, but its difficult 2010-10-16T20:39:59 it still loses stupid games anyway 2010-10-16T20:40:21 and it's really grumpy about redistributing 2010-10-16T20:42:16 *** Meatkat has quit IRC (Quit: Leaving) 2010-10-16T20:42:31 thanks you for the info 2010-10-16T20:42:42 *** Bobng has quit IRC (Remote host closed the connection) 2010-10-16T20:43:25 *** Bobng has joined #aichallenge 2010-10-16T20:44:56 heh 2010-10-16T20:45:32 but its interesting to watch it 2010-10-16T20:45:45 it does things that seem illogical when you look at it,but it wins 2010-10-16T20:46:17 the weird part is it does partial moves 2010-10-16T20:46:21 why burn ships against neutrals? 2010-10-16T20:46:54 how does that factor into a minimax system? 2010-10-16T20:47:47 won't it be better to prune moves that don't result in any kind of increase in growth or ships? 2010-10-16T20:50:06 when it made the move it would have thought it was going to take the neutral 2010-10-16T21:01:43 Top 10 players: dmj111(3801), Hazard(3663), MariusCG(3643), deepblue(3623), Alvraro(3574), fglider(3550), savchenko(3542), Manwe56(3528), bix0r4ever(3507), ofrias(3502) 2010-10-16T21:12:37 *** shawn_ has quit IRC (Quit: Leaving) 2010-10-16T21:15:22 *** narnach1 has joined #aichallenge 2010-10-16T21:15:50 antimatroid: the game of your bot on tcp made me think of Hazard 2010-10-16T21:16:18 could he be using a minimax? 2010-10-16T21:16:32 *** gerald512k has quit IRC () 2010-10-16T21:28:33 *** bimbot has joined #aichallenge 2010-10-16T21:28:37 hi 2010-10-16T21:29:28 for the one who love funny galactic wars i have a good replay ;) http://ai-contest.com/visualizer.php?game_id=5923654 2010-10-16T21:36:08 *** deepblue has joined #aichallenge 2010-10-16T21:38:50 good evening guys/girls 2010-10-16T21:40:18 hello 2010-10-16T21:40:26 aya deepblue :) 2010-10-16T21:40:33 *** Bobng has left #aichallenge ("Leaving") 2010-10-16T21:40:48 so you are back to top deepblue 2010-10-16T21:41:54 seems like it 2010-10-16T21:41:59 @rankings 2010-10-16T21:42:01 deepblue: Top 10 players: dmj111(3822), Hazard(3727), MariusCG(3706), deepblue(3698), Alvraro(3602), fglider(3593), savchenko(3587), Manwe56(3565), ofrias(3563), DocBot(3542) 2010-10-16T21:42:22 switching places back and forth with hazard and maruis 2010-10-16T21:42:44 next version should hopefully be able to beat dmj111 again :D 2010-10-16T21:43:34 I wonder how high my bot will get, 39 right now 2010-10-16T21:44:10 do you still only use heuristic? 2010-10-16T21:44:22 yep 2010-10-16T21:44:41 it took my bot quite long to get that far up 2010-10-16T21:52:33 yes 2010-10-16T21:52:48 it is interesting that your bot can always be better :D 2010-10-16T21:52:52 you never stop play 2010-10-16T21:53:34 i love this game lol 2010-10-16T21:54:30 *** bimbot has quit IRC (Quit: Page closed) 2010-10-16T22:01:43 Top 10 players: dmj111(3843), Hazard(3748), MariusCG(3728), deepblue(3718), Alvraro(3662), savchenko(3621), fglider(3614), DocBot(3577), cfaftw(3576), ofrias(3575) 2010-10-16T22:02:19 deepblue: yay! welcome! 2010-10-16T22:02:24 your first time in here? 2010-10-16T22:03:23 *** vizier has joined #aichallenge 2010-10-16T22:04:15 oh 2010-10-16T22:04:24 amstan, i've been here a few random times 2010-10-16T22:04:46 deepblue: tip: get an account, you don't want people to impersonate you 2010-10-16T22:04:48 but i rarely use irc nowadays 2010-10-16T22:12:50 :( 2010-10-16T22:14:07 what's up azrathud? 2010-10-16T22:16:02 "... rarely use irc nowadays" 2010-10-16T22:16:22 ah 2010-10-16T22:16:24 oh well 2010-10-16T22:17:02 have stopped using that when i stopped playing CS 2010-10-16T22:18:56 deepblue: irc is much more than just a CS communication medium 2010-10-16T22:19:09 i know i know 2010-10-16T22:19:35 just tried to spend less time in front of my computer :) 2010-10-16T22:19:56 didn't work out though 2010-10-16T22:20:06 but stopped hanging around in irc channels 2010-10-16T22:20:06 meh cloud is going back down :( 2010-10-16T22:20:13 *** vizier has quit IRC (Ping timeout: 265 seconds) 2010-10-16T22:20:13 oh i lied 2010-10-16T22:20:30 last time i seriously used irc was when i helped writing that diablo 2 bot 2010-10-16T22:23:44 *** Frontier has joined #aichallenge 2010-10-16T22:28:30 *** Sorian has joined #aichallenge 2010-10-16T22:35:15 *** Gudradain has quit IRC (Quit: Page closed) 2010-10-16T22:37:36 I just realized a pretty terrible flaw in my bot 2010-10-16T22:38:35 but I'm not sayin', because I'm sure other bots have this same problem...lol 2010-10-16T22:39:33 *** deepblue has quit IRC (Ping timeout: 265 seconds) 2010-10-16T22:39:45 *** deepblue has joined #aichallenge 2010-10-16T22:48:41 Is there a way to find out why my C++ submission is failing to compile? It compiles fine locally. 2010-10-16T22:52:55 are you not getting an email? 2010-10-16T22:53:02 Nope, no email 2010-10-16T22:53:10 how are you compiling? 2010-10-16T22:53:52 Locally, I compile in VS2008. It compiles fine on my machine, but my profile page says that my submission failed to compile on their end. 2010-10-16T22:54:14 Sorian: try compiling it on gcc 2010-10-16T22:54:15 okay, the server is compiling with g++ 2010-10-16T22:54:25 g++ 2010-10-16T22:54:28 *** Appleman1234 has joined #aichallenge 2010-10-16T22:54:38 ok, Ill try that 2010-10-16T22:58:36 What does CS stand for in this case? 2010-10-16T22:58:48 computer science, presumably 2010-10-16T22:59:18 'playing computer science.' That doesn't make any sense to me. 2010-10-16T22:59:26 ahem 2010-10-16T22:59:27 counter strike 2010-10-16T22:59:27 :D 2010-10-16T22:59:34 Ah. 2010-10-16T22:59:53 the university of waterloo counterstike club 2010-10-16T23:00:02 clearly 2010-10-16T23:00:04 Heh. 2010-10-16T23:01:09 ;) 2010-10-16T23:01:14 *** pgpaskar_ has quit IRC (Ping timeout: 240 seconds) 2010-10-16T23:01:20 oh, only saw the sentence after that... 2010-10-16T23:01:25 kids these days... 2010-10-16T23:01:30 more the small town in germant counter strike LAN club 2010-10-16T23:01:34 germany 2010-10-16T23:01:41 rwa: my bot sort of works 2010-10-16T23:01:44 it's very easy to break 2010-10-16T23:01:44 Top 10 players: dmj111(3903), Hazard(3803), deepblue(3758), MariusCG(3721), Alvraro(3687), GreenTea(3684), fglider(3639), savchenko(3631), ofrias(3595), felixcoto(3595) 2010-10-16T23:01:49 and bugs are a many 2010-10-16T23:01:57 yay rank3 again 2010-10-16T23:02:11 I have to rethink a major aspect of my bot 2010-10-16T23:03:06 *** yellowseed has quit IRC (Quit: yellowseed) 2010-10-16T23:03:09 *** pgpaskar_ has joined #aichallenge 2010-10-16T23:03:09 deepblue: has your bot changed much since you were on the tcp server? 2010-10-16T23:03:28 ugh 2010-10-16T23:03:35 i am terrible at working on my bot 2010-10-16T23:04:23 the bot on the official is older than the stuff I had running 2010-10-16T23:04:34 but it's not much improved or improved at all 2010-10-16T23:04:48 working on some new things right now :) 2010-10-16T23:05:08 and still have a nice list of ideas to keep me entertained (and you guys too) 2010-10-16T23:05:09 ;) 2010-10-16T23:05:38 will probably let something run on the tcp later tonight 2010-10-16T23:09:54 *** krokokrusa has joined #aichallenge 2010-10-16T23:15:08 Fixed my issue. C++ files require a blank line at the end of the file on Unix apparently. 2010-10-16T23:18:46 yes 2010-10-16T23:18:56 that's specified on some standard I don't remember 2010-10-16T23:18:58 maybe POSIX? 2010-10-16T23:20:02 *** Sorian has quit IRC (Quit: ChatZilla 0.9.86 [Firefox 3.6.10/20100914125854]) 2010-10-16T23:21:14 more like C++ standard? 2010-10-16T23:21:46 the blank line at the end of the stuff? 2010-10-16T23:22:25 yep 2010-10-16T23:22:57 I remembered it was on some sort of standard 2010-10-16T23:23:13 also, emacs puts a blank line at the end of whatever file you create 2010-10-16T23:24:02 it has todo with the include 2010-10-16T23:24:20 if you include the file content just gets inserted there without an additional new line after the content 2010-10-16T23:24:55 so that's why there should be a new line at the end of every file :) 2010-10-16T23:25:30 *** Appleman1234 has quit IRC (Excess Flood) 2010-10-16T23:26:19 *** Appleman1234 has joined #aichallenge 2010-10-16T23:37:17 *** Blkt has quit IRC (Ping timeout: 265 seconds) 2010-10-16T23:39:07 *** Mathnerd314 has quit IRC (Ping timeout: 252 seconds) 2010-10-16T23:54:41 *** hellman has joined #aichallenge