2010-10-26T00:00:03 *** eburnette has quit IRC (Ping timeout: 265 seconds) 2010-10-26T00:00:59 but I would be OK with an NLP irc bot contest :-) 2010-10-26T00:02:23 (I doubt anybody would submit serious entries) 2010-10-26T00:04:05 (they'd do it for the Loebner prize instead) 2010-10-26T00:04:18 loebner is kinda useless 2010-10-26T00:04:37 the bot could just output random garbage.. because hey, there are humans that would do that 2010-10-26T00:08:22 not really... your goal is to *prove* you're human 2010-10-26T00:09:31 but prooving your a human is not that useful when most humans are trolls and don't obey instructions 2010-10-26T00:11:48 *** TravisD has quit IRC (Quit: TravisD) 2010-10-26T00:11:53 Top 10 players: bocsimacko(3754), dmj111(3707), felixcoto(3700), Hazard(3671), asavis(3654), GreenTea(3649), jambachili(3603), luksian(3601), sin_sun(3576), george(3575) 2010-10-26T00:11:54 Mathnerd314, I am working on an NLP irc bot now, instead of my competition entry lol 2010-10-26T00:12:27 amstan: ok, proving you're a *good* human 2010-10-26T00:12:28 yay! dmj is getting up! 2010-10-26T00:12:43 Appleman1234: yes! my secret strategy is working! 2010-10-26T00:12:47 Mathnerd314: define good human 2010-10-26T00:13:12 Mathnerd314, I started before you mentioned it 2010-10-26T00:13:16 amstan: productive. writes code or something. 2010-10-26T00:13:45 Mathnerd314: but see, a turing test won't help you in those regards 2010-10-26T00:14:29 amstan: yes it will. add conditions that the computer/person must be proficient in coding. 2010-10-26T00:15:00 Mathnerd314: but they don't have to respect any of the interviewer's wishes, that's the point 2010-10-26T00:15:10 even if they know to code anything, they can just refuse 2010-10-26T00:15:22 or even lie about their coding proficiency 2010-10-26T00:16:00 yay halting problem applied to truth determination in the form of a turing test :P 2010-10-26T00:16:07 this would be so much easier to just officially drop C# and any other mono language from being supported by the contest 2010-10-26T00:16:13 amstan: give them incentive to win, then. the loser dies if they aren't human 2010-10-26T00:16:31 janzert: lol 2010-10-26T00:16:33 (as determined by the judge 2010-10-26T00:16:41 *** eburnette has joined #aichallenge 2010-10-26T00:17:03 testing... 2010-10-26T00:17:11 janzert: this quote "We do things because they are hard, not because they are easy" has been running through my mind 2010-10-26T00:17:11 eburnette: not working 2010-10-26T00:17:43 Mathnerd314: not really, computer programmers do stuff to make their life easier 2010-10-26T00:18:38 amstan: no, they do it because they like it and/or because they get money 2010-10-26T00:19:07 i disagree 2010-10-26T00:19:13 i didn't make contestbot because i liked it 2010-10-26T00:19:19 it was enjoyable 2010-10-26T00:19:30 but i made it so it's easier for me and others to find stuff about the contest 2010-10-26T00:19:40 i'm too lazy to load rankings.php 2010-10-26T00:19:47 so.. i made contestbot do it instead: 2010-10-26T00:19:49 #rankings 2010-10-26T00:19:53 @rankings 2010-10-26T00:19:54 amstan: Top 10 players: bocsimacko(3753), dmj111(3706), felixcoto(3700), Hazard(3670), asavis(3653), GreenTea(3648), jambachili(3602), luksian(3600), sin_sun(3576), george(3573) 2010-10-26T00:19:57 :) 2010-10-26T00:20:32 *** davidjliu has quit IRC (Quit: Page closed) 2010-10-26T00:21:20 you just keep on telling yourself that :-) 2010-10-26T00:21:31 what? 2010-10-26T00:22:01 I've discovered a flaw in bocsimacko 2010-10-26T00:22:21 eburnette: cool, care to share? 2010-10-26T00:24:15 I'm not close enough to exploit it so sure. Look at http://72.44.46.68/canvas?game_id=703823 move 36, dmj swoops in for the kill 2010-10-26T00:25:45 eburnette: yeah, he left his closest planet uncovered 2010-10-26T00:25:53 dmj111: good job 2010-10-26T00:26:24 *** yasith has quit IRC (Read error: Connection reset by peer) 2010-10-26T00:27:57 I haven't figured out a way to do those multi-planet time traveling attacks that bocsimacko does so well (where everything arrives at once), any ideas? 2010-10-26T00:28:59 eburnette: assemble a list of source planets, caculate the time it takes to get to the target 2010-10-26T00:29:12 launch your farthest fleet 2010-10-26T00:29:45 let's say it arrives at target at time TT 2010-10-26T00:30:08 you have to launch your other fleets at time TT-Di, where Di is time to reach the target 2010-10-26T00:30:30 *** vizier has quit IRC (Ping timeout: 265 seconds) 2010-10-26T00:31:29 amsta bockismacko does something even more awesome actualy 2010-10-26T00:31:29 yeah, but how do you choose the target? 2010-10-26T00:31:42 *** TravisD has joined #aichallenge 2010-10-26T00:31:46 Mathnerd314: that's beyond the scope of the question 2010-10-26T00:31:56 he launches partial fleets calculating that he will take neutral at time T and that it will have X ships that he can add to the attack 2010-10-26T00:32:13 watch his start on close matches 2010-10-26T00:32:17 its pretty interesting 2010-10-26T00:32:35 Guess I need to keep state between moves, not currently doing that 2010-10-26T00:32:39 he is taking into account what neutrals will have how many ships in the future and just launching that those ships will be available 2010-10-26T00:33:49 I've seen him launch ships at a planet he owned, then those took the place of a launch one move ahead when they got there (kind of a time warp) 2010-10-26T00:36:23 He claims it's just a 2-ply minimax with a good eval function but I'm skeptical that's really it because he's doing it in lisp and there are tens of thousands of possible moves per ply 2010-10-26T00:36:43 sbcl is pretty fast. compiles to native code; pretty efficient code with type hints 2010-10-26T00:36:49 but yeah, the move generator is something special 2010-10-26T00:37:18 my goal was 1 ply minimax with a good eval function with a pretty exhaustive move generator but maybe that's not so good 2010-10-26T00:37:41 cuz i'm having a hard time defeating the tactic where you pick off a neutral right after your opponent takes it 2010-10-26T00:38:48 you could have a simple 2nd ply that does that sniping 2010-10-26T00:39:11 I found ragebot mades a pretty good 2nd ply :) 2010-10-26T00:41:18 It's hard not to be overly cautious yet still protect against rage attacks 2010-10-26T00:42:33 *** yasith has joined #aichallenge 2010-10-26T00:44:07 I wouldn't mind losing to bocsimacko so much if he didn't do it in such humiliatingly short games. last one was 35, geesh :( 2010-10-26T00:44:51 http://72.44.46.68/canvas?game_id=703988 2010-10-26T00:44:52 haha yea 2010-10-26T00:44:54 :( 2010-10-26T00:45:02 though my bot managed to beat him earlier today twice 2010-10-26T00:47:49 (shakes fist at bocsimacko) I'll get you some day... and your little bot too! 2010-10-26T00:48:51 *** eburnette has quit IRC (Quit: Page closed) 2010-10-26T00:50:13 *** hellman_ has joined #aichallenge 2010-10-26T00:51:27 *** TravisD has quit IRC (Read error: Connection reset by peer) 2010-10-26T00:52:56 *** TravisD has joined #aichallenge 2010-10-26T00:56:45 *** Mathnerd314 has quit IRC (Quit: ChatZilla 0.9.86-rdmsoft [XULRunner 1.9.2.11/20101012113537]) 2010-10-26T00:58:51 *** TravisD has quit IRC (Quit: TravisD) 2010-10-26T01:05:47 hah interesting game though i lose it 2010-10-26T01:05:48 http://72.44.46.68/canvas?game_id=704540 2010-10-26T01:05:52 complete base swap 2010-10-26T01:07:25 lol bhasker 2010-10-26T01:07:33 your algos must be similar? 2010-10-26T01:07:51 probably 2010-10-26T01:07:57 he only won by a little bit 2010-10-26T01:08:08 nah he might have won it 2010-10-26T01:08:14 at the end he had 200 ship advantage 2010-10-26T01:08:19 i had a 5 growth advantage though 2010-10-26T01:08:33 yea 2010-10-26T01:09:02 so it was how you handle it for another 40 turns :P 2010-10-26T01:11:54 Top 10 players: bocsimacko(3738), dmj111(3707), felixcoto(3657), Hazard(3645), GreenTea(3641), jambachili(3591), asavis(3591), george(3578), sin_sun(3564), Arekku(3544) 2010-10-26T01:12:57 *** delt0r___ has joined #aichallenge 2010-10-26T01:14:36 *** delt0r has quit IRC (Ping timeout: 252 seconds) 2010-10-26T01:14:53 strange all the ELOs have dropped 2010-10-26T01:17:04 yeah, it's weird 2010-10-26T01:23:12 *** boegel has joined #aichallenge 2010-10-26T01:27:35 i'm surprised bocsimacko isn't higher above everyone 2010-10-26T01:27:40 he rips the shit out of everyone 2010-10-26T01:30:54 *** boegel has quit IRC (Quit: *poof!*) 2010-10-26T01:33:41 antimatroid1: that's an old version 2010-10-26T01:34:15 damn you :P 2010-10-26T01:34:22 i have a month to pound you yet 2010-10-26T01:35:52 the old-version mega1 is pounding everyone... what aobut the new version? 2010-10-26T01:36:01 turn 3... bocksimacko wins 2010-10-26T01:36:03 :P 2010-10-26T01:36:40 turn 2... opponent flees in terror 2010-10-26T01:41:51 turn 1 player makes invalid move because it's less shameful 2010-10-26T01:42:36 :P 2010-10-26T01:42:53 fleets all leave the board 2010-10-26T01:42:56 :P 2010-10-26T01:43:08 retreat! 2010-10-26T01:43:38 *** DanielVF has joined #aichallenge 2010-10-26T01:45:08 that was a short return 2010-10-26T01:45:19 *** DanielVF has joined #aichallenge 2010-10-26T01:52:00 *** javagamer has quit IRC (Ping timeout: 252 seconds) 2010-10-26T01:52:31 *** narnach1 has joined #aichallenge 2010-10-26T01:52:31 *** narnach1 is now known as narnach_mb 2010-10-26T01:54:11 *** Accoun has quit IRC (*.net *.split) 2010-10-26T01:54:12 *** nibalize1 has quit IRC (*.net *.split) 2010-10-26T01:54:12 *** yam has quit IRC (*.net *.split) 2010-10-26T02:00:15 *** Accoun has joined #aichallenge 2010-10-26T02:02:47 *** narnach_mb has quit IRC (Quit: Leaving.) 2010-10-26T02:07:43 *** nibalizer has joined #aichallenge 2010-10-26T02:07:55 *** Palmik has joined #aichallenge 2010-10-26T02:10:45 *** yasith has quit IRC (Ping timeout: 240 seconds) 2010-10-26T02:11:54 Top 10 players: bocsimacko(3696), rsergio(3694), dmj111(3685), felixcoto(3633), Hazard(3623), GreenTea(3621), jambachili(3577), asavis(3569), george(3557), sin_sun(3543) 2010-10-26T02:20:21 *** virdo has quit IRC () 2010-10-26T02:21:27 *** hellman_ has quit IRC (Remote host closed the connection) 2010-10-26T02:21:27 *** hellman has quit IRC (Remote host closed the connection) 2010-10-26T02:22:27 *** yasith has joined #aichallenge 2010-10-26T02:22:29 !rankings 2010-10-26T02:22:39 contestbot! 2010-10-26T02:22:44 !rankings 2010-10-26T02:23:30 @rankings 2010-10-26T02:23:32 antimatroid1: Top 10 players: rsergio(3690), bocsimacko(3689), dmj111(3682), GreenTea(3627), felixcoto(3624), Hazard(3615), jambachili(3569), asavis(3564), george(3554), Arekku(3530) 2010-10-26T02:23:36 there we go 2010-10-26T02:23:38 mega1: :P 2010-10-26T02:24:00 yeah 2010-10-26T02:24:59 what puzzles me more is why 0.31 worse than 0.30 on tcp when locally it beat 0.30 2010-10-26T02:25:29 it is still running locally too :P 2010-10-26T02:31:27 *** perestrelka has quit IRC (Read error: Connection reset by peer) 2010-10-26T02:32:02 *** perestrelka has joined #aichallenge 2010-10-26T02:32:14 wow, because on tcp it loses 10-28-19 (win-draw-loss) to b0.30 2010-10-26T02:32:27 different maps distribution? 2010-10-26T02:32:33 small sample? 2010-10-26T02:32:42 there are more maps on the tcp server 2010-10-26T02:33:22 yes, but I expected the results from a 100 map test hold 2010-10-26T02:33:38 *** sigh has quit IRC (Remote host closed the connection) 2010-10-26T02:33:47 grab the next hundred and see what they do? 2010-10-26T02:34:00 there's a maps.zip file on tcp somewhere 2010-10-26T02:34:58 that's a good idea 2010-10-26T02:35:28 that brings up the question of map generation 2010-10-26T02:36:00 it would be good to fix the map generator algorithm well before the contest end 2010-10-26T02:36:16 fix = not change 2010-10-26T02:36:56 i wanted lots of human made maps added 2010-10-26T02:37:03 but people weren't keen 2010-10-26T02:37:12 the idea would be to design lots of maps to pick out potential flaws in bots 2010-10-26T02:37:43 would be cool to get some maps like that on tcp at least 2010-10-26T02:39:46 *** Commandir has joined #aichallenge 2010-10-26T02:41:30 *** Utkarsh has quit IRC (Ping timeout: 240 seconds) 2010-10-26T02:47:49 hooray poor design! 2010-10-26T02:47:52 *** Utkarsh has joined #aichallenge 2010-10-26T02:48:16 i have the same algorithm implemented three times in three different ways, and i'm going to be adding a fourth shortly 2010-10-26T02:50:47 *** num1 has joined #aichallenge 2010-10-26T02:50:48 *** num1 has joined #aichallenge 2010-10-26T02:51:52 *** wvdschel has joined #aichallenge 2010-10-26T02:52:43 Zannick: where is the maps.zip? is it on the tcp homepage or on the forum 2010-10-26T02:53:20 http://72.44.46.68/getmaps 2010-10-26T02:53:30 ah, it's a tar.gz 2010-10-26T02:53:56 *** num1 has left #aichallenge 2010-10-26T02:53:59 and there are 999 of them 2010-10-26T02:54:04 *** num1 has joined #aichallenge 2010-10-26T02:54:29 *** num1 has quit IRC (Client Quit) 2010-10-26T02:57:48 ok thanks 2010-10-26T03:03:36 are all maps generated by the same algorithm but with different seeds? 2010-10-26T03:03:47 essentially 2010-10-26T03:04:18 only essentially? 2010-10-26T03:04:40 http://code.google.com/p/ai-contest/source/browse/trunk/planet_wars/backend/map_generator.py 2010-10-26T03:05:02 the seeds aren't specified 2010-10-26T03:06:46 *** jaspervdj has joined #aichallenge 2010-10-26T03:11:55 Top 10 players: bocsimacko(3679), dmj111(3676), GreenTea(3621), rsergio(3617), Hazard(3616), felixcoto(3610), jambachili(3569), asavis(3565), george(3556), Arekku(3516) 2010-10-26T03:12:42 *** hornairs has quit IRC (Quit: hornairs) 2010-10-26T03:13:00 *** worXrillo has joined #aichallenge 2010-10-26T03:14:25 *** Appleman1234 has quit IRC (Ping timeout: 264 seconds) 2010-10-26T03:15:14 *** TheMacca has joined #aichallenge 2010-10-26T03:22:36 *** Appleman1234 has joined #aichallenge 2010-10-26T03:26:42 *** aerique has joined #aichallenge 2010-10-26T03:31:57 *** hellman has joined #aichallenge 2010-10-26T03:32:19 *** TheMacca has quit IRC (Quit: Page closed) 2010-10-26T03:35:54 *** shades has quit IRC (Ping timeout: 240 seconds) 2010-10-26T03:40:04 *** yasith_ has joined #aichallenge 2010-10-26T03:41:14 *** yasith has quit IRC (Ping timeout: 250 seconds) 2010-10-26T03:42:19 *** shades has joined #aichallenge 2010-10-26T03:48:31 *** shades has quit IRC (Ping timeout: 255 seconds) 2010-10-26T03:57:56 *** lavalamp_afk is now known as lavalamp_ 2010-10-26T03:59:01 *** hellman has quit IRC (Ping timeout: 255 seconds) 2010-10-26T04:02:53 *** n9thbit has joined #aichallenge 2010-10-26T04:04:40 *** Florian95 has joined #aichallenge 2010-10-26T04:09:32 *** perror has joined #aichallenge 2010-10-26T04:10:42 *** shades has joined #aichallenge 2010-10-26T04:11:57 Top 10 players: bocsimacko(3669), dmj111(3667), GreenTea(3625), Hazard(3600), felixcoto(3596), jambachili(3562), asavis(3552), george(3543), rsergio(3541), sin_sun(3517) 2010-10-26T04:12:14 *** tapwater has quit IRC (Quit: tapwater) 2010-10-26T04:15:52 *** Itkovian has joined #aichallenge 2010-10-26T04:23:55 http://ai-contest.com/visualizer.php?game_id=6207131 2010-10-26T04:24:08 george is really strange 2010-10-26T04:25:03 whats wrong with it mega1 ? 2010-10-26T04:25:25 that he doesn't move for several turns after the start? 2010-10-26T04:25:32 and still wins. 2010-10-26T04:25:45 good point 2010-10-26T04:25:46 :P 2010-10-26T04:25:51 althought that might have had more to do with felixcoto 2010-10-26T04:26:29 maybe very conservative 2010-10-26T04:26:36 whenever I look at it george looks like a tree search bot 2010-10-26T04:27:39 *** n9thbit has quit IRC (Quit: Page closed) 2010-10-26T04:28:22 *** boegel has joined #aichallenge 2010-10-26T04:45:49 *** p4p4 has joined #aichallenge 2010-10-26T05:11:55 Top 10 players: bocsimacko(3660), dmj111(3658), GreenTea(3615), Hazard(3588), felixcoto(3583), rsergio(3565), jambachili(3561), george(3550), sin_sun(3516), asavis(3510) 2010-10-26T05:14:25 *** Accoun has quit IRC (Ping timeout: 264 seconds) 2010-10-26T05:15:52 *** DanielVF has quit IRC (Quit: DanielVF) 2010-10-26T05:17:08 *** Titankiller has joined #aichallenge 2010-10-26T05:19:33 *** genericb1b has quit IRC (Remote host closed the connection) 2010-10-26T05:19:38 *** genericbob has joined #aichallenge 2010-10-26T05:27:03 *** p4p4 has quit IRC (Ping timeout: 252 seconds) 2010-10-26T05:27:03 *** amstan has quit IRC (Ping timeout: 252 seconds) 2010-10-26T05:27:04 *** p4p4_ has joined #aichallenge 2010-10-26T05:27:11 *** Naktibalda has joined #aichallenge 2010-10-26T05:27:16 *** amstan has joined #aichallenge 2010-10-26T05:27:16 *** ChanServ sets mode: +o amstan 2010-10-26T05:33:38 *** Azrathud has joined #aichallenge 2010-10-26T05:34:00 is the ai-contest.com server down? 2010-10-26T05:34:59 works fine for me 2010-10-26T05:35:21 *** narnach1 has joined #aichallenge 2010-10-26T05:35:21 *** narnach1 is now known as narnach_mb 2010-10-26T05:40:16 Commandir: works here too now, glitch prolly 2010-10-26T05:41:29 *** n9thbit has joined #aichallenge 2010-10-26T05:43:45 *** test23 has joined #aichallenge 2010-10-26T05:44:21 *** test23 has left #aichallenge 2010-10-26T05:51:42 *** seisatsu has quit IRC (Read error: Connection reset by peer) 2010-10-26T05:51:56 *** seisatsu has joined #aichallenge 2010-10-26T05:56:51 *** Bobng has joined #aichallenge 2010-10-26T06:08:21 *** lavalamp_ has quit IRC (Ping timeout: 265 seconds) 2010-10-26T06:11:56 Top 10 players: dmj111(3664), GreenTea(3605), felixcoto(3580), Hazard(3572), jambachili(3565), george(3536), rsergio(3523), Arekku(3518), asavis(3503), sin_sun(3492) 2010-10-26T06:15:00 *** irchs has joined #aichallenge 2010-10-26T06:42:07 *** phreeza has joined #aichallenge 2010-10-26T06:49:27 *** hellman has joined #aichallenge 2010-10-26T06:56:14 *** n9thbit has left #aichallenge 2010-10-26T06:59:15 *** Titankiller has quit IRC (Ping timeout: 240 seconds) 2010-10-26T07:09:04 *** Bobng has quit IRC (Remote host closed the connection) 2010-10-26T07:09:44 *** Bobng has joined #aichallenge 2010-10-26T07:11:56 Top 10 players: dmj111(3614), GreenTea(3593), Hazard(3552), felixcoto(3548), rsergio(3547), jambachili(3519), Arekku(3509), asavis(3506), george(3497), sin_sun(3453) 2010-10-26T07:19:41 *** Titankiller has joined #aichallenge 2010-10-26T07:29:51 *** n9thbit has joined #aichallenge 2010-10-26T07:30:36 *** j3camero has quit IRC (Remote host closed the connection) 2010-10-26T07:31:59 http://ai-contest.com/ down for anyone else? 2010-10-26T07:33:43 timeout 2010-10-26T07:37:30 yep, down. 2010-10-26T07:38:08 true 2010-10-26T07:44:25 *** narnach has quit IRC (Ping timeout: 264 seconds) 2010-10-26T07:44:56 its back up now 2010-10-26T07:45:08 offtopic: do you guys know any freeware/opensource software which makes an backup for ie. sd-card 2010-10-26T07:45:12 _full_ backup 2010-10-26T07:45:20 like an image 2010-10-26T07:45:22 /offtopic 2010-10-26T07:47:11 *** j3camero has joined #aichallenge 2010-10-26T07:49:21 *** narnach has joined #aichallenge 2010-10-26T07:52:16 *** n9thbit has left #aichallenge 2010-10-26T07:54:37 rabidus_: http://linux.die.net/man/1/dd 2010-10-26T07:54:48 in windows :| 2010-10-26T07:55:05 rabidus_: there's windows ports of that i believe 2010-10-26T07:55:18 hmm.. ok 2010-10-26T07:55:55 e.g. http://www.chrysocome.net/dd 2010-10-26T07:56:08 note the tool can be destructive if used improperly 2010-10-26T07:56:27 and i don't know how good the windows ports are 2010-10-26T07:56:37 /enddisclaimer 2010-10-26T07:56:47 aka 'i dont take any responsibility' 2010-10-26T07:56:59 yep - but that is "a tool to do that" 2010-10-26T07:57:21 yeah, thanks for help. 2010-10-26T08:11:57 Top 10 players: dmj111(3640), GreenTea(3595), rsergio(3570), Simplex(3549), Hazard(3538), felixcoto(3537), jambachili(3521), Arekku(3505), george(3472), asavis(3453) 2010-10-26T08:13:34 woah where's mega1 gone - did you update your bot? 2010-10-26T08:13:54 resubmit 2010-10-26T08:17:56 *** javagamer has joined #aichallenge 2010-10-26T08:18:03 cool ^^ 2010-10-26T08:18:30 if my bot doesn't work, i guess ill just have to take up cheerleading your bot mega1 2010-10-26T08:23:51 *** Azrathud has quit IRC (Quit: Meh) 2010-10-26T08:24:42 it's a bit early for that 2010-10-26T08:25:35 hehe ok :P 2010-10-26T08:25:48 *** Azrathud has joined #aichallenge 2010-10-26T08:27:36 *** Titankiller has quit IRC (Read error: Connection reset by peer) 2010-10-26T08:31:01 Dear developers, it seems that currently bots that have just been submitted are also doing battle against eachother. This results in slower convergence of their true rankings. Imho newly submitted bots should only play against bots that have a stable ranking. 2010-10-26T08:36:05 morning 2010-10-26T08:50:55 amstan,j3camero,janzert ? 2010-10-26T08:51:30 Not sure if this has been reported before, but I am sure that its not yet applied :) 2010-10-26T08:52:32 *** irchs has quit IRC (Quit: irchs) 2010-10-26T08:53:34 Error323: you can create an issue here http://code.google.com/p/ai-contest/issues/list 2010-10-26T08:54:46 *** Titankiller has joined #aichallenge 2010-10-26T08:59:09 Naktibalda: thnx 2010-10-26T09:00:27 this morning I was surprised to see b0.31 lose to b0.30 on tcp by 10-28-29 (win-draw-loss). The same result over all 1000 maps is 752-733-512. 2010-10-26T09:01:17 now I only need to calculate the likelihood of seeing a results as extreme as on the tcp server to gauge whether something's wrong with map selection 2010-10-26T09:06:16 mega1 your bot is awesome btw 2010-10-26T09:06:41 the first bot that i have seen that does some very funky stuff 2010-10-26T09:07:23 stop it now or I'll be too ashamed to publish to source code come the end of contest 2010-10-26T09:07:27 haha 2010-10-26T09:07:37 i like how it calculates into the future when capturing neutrals 2010-10-26T09:07:40 got a cool match to show? 2010-10-26T09:07:46 let me see if i can find it 2010-10-26T09:07:47 hold on 2010-10-26T09:08:08 its quite an interesting bot to watch 2010-10-26T09:09:13 (mega1punch (fist enemy neutral)) 2010-10-26T09:09:47 http://72.44.46.68/canvas?game_id=698634 2010-10-26T09:09:54 watch mega1's first turn 2010-10-26T09:10:00 and check the top planet with 15 ships on it 2010-10-26T09:10:06 its quite an interesting launch 2010-10-26T09:11:57 Top 10 players: Simplex(3656), dmj111(3619), rsergio(3614), GreenTea(3587), felixcoto(3544), Hazard(3538), jambachili(3508), Arekku(3484), george(3474), asavis(3455) 2010-10-26T09:12:05 do you see it Error323? 2010-10-26T09:12:27 I think it simply chooses not to attack that planet with 15 on the first turn and the rest follows 2010-10-26T09:12:28 wow indeed 2010-10-26T09:12:30 nice work :) 2010-10-26T09:12:46 it did attack it 2010-10-26T09:12:50 with 4 ships on turn 1 2010-10-26T09:12:55 on the next turn only 2010-10-26T09:13:01 hmm let me check again 2010-10-26T09:13:15 oh yea 1 turn behind 2010-10-26T09:13:18 but why send 4 ships? 2010-10-26T09:13:49 it was no way enough to take the planet unless you include the ships it got from the newly acquired neutrals 2010-10-26T09:14:04 *** delt0r has joined #aichallenge 2010-10-26T09:14:08 by that time blue's 16 ships are on their way 2010-10-26T09:14:08 *** delt0r___ has quit IRC (Ping timeout: 240 seconds) 2010-10-26T09:14:18 yes but your 4 are not enough to take it? 2010-10-26T09:14:26 you are more than 1 turn away 2010-10-26T09:14:42 in the end you needed 4+4+2 ships to take it 2010-10-26T09:14:54 what you see is a multiplanet attack 2010-10-26T09:15:18 so the initial 4 launch was either wrong or it factored the new neutrals it will have at that turn into account 2010-10-26T09:15:24 it looks better because when the first fleet is launch the other two planets are not yet taken 2010-10-26T09:15:29 yea 2010-10-26T09:15:36 *launched 2010-10-26T09:15:45 i do multi planet launch but mine doesn't take the neutrals not yet acquired into account 2010-10-26T09:15:58 need to figure out a good way to do that ^ 2010-10-26T09:16:18 though my bot did manage to beat yours twice yesterday ^ 2010-10-26T09:16:38 http://72.44.46.68/canvas?game_id=701233 :d 2010-10-26T09:16:43 i should go celebrate 2010-10-26T09:19:07 the way my code is structured it would be hard to do what your bot is doing, factoring in neutrals before acquisition 2010-10-26T09:19:46 *** jesionaj has quit IRC (Ping timeout: 240 seconds) 2010-10-26T09:20:15 *** Appleman1234 has quit IRC (Ping timeout: 252 seconds) 2010-10-26T09:21:24 why don't you start up your bot on tcp i want to see how my bot fares against yours in general 2010-10-26T09:21:37 *** Utkarsh has quit IRC (Ping timeout: 264 seconds) 2010-10-26T09:21:45 my bots elo rating is artificially inflated as most of the good bots are not running right now 2010-10-26T09:23:55 *** Azrathud has quit IRC (Ping timeout: 255 seconds) 2010-10-26T09:27:18 *** Utkarsh has joined #aichallenge 2010-10-26T09:27:42 hmm all the accoun* bots started up 2010-10-26T09:28:00 interesting 2010-10-26T09:32:21 *** Utkarsh has quit IRC (Ping timeout: 252 seconds) 2010-10-26T09:38:22 *** Utkarsh has joined #aichallenge 2010-10-26T09:40:01 *** Accoun has joined #aichallenge 2010-10-26T09:40:44 1 2010-10-26T09:42:36 1 month 1 day 2010-10-26T09:43:57 bhasker: in a few hours I'll restart them 2010-10-26T09:45:24 ah k 2010-10-26T09:51:34 *** vermapratyush has joined #aichallenge 2010-10-26T09:52:46 I used DevCpp to make a simple bot .. i want to use a simulator to see how it works.. can anyone suggest me a simulator ? 2010-10-26T09:54:22 i have GlaconBotTesting Arena but it stops responging when i start a game.. 2010-10-26T09:54:59 And planet warrior is usnble to start my bot 2010-10-26T09:55:03 please help 2010-10-26T09:56:04 vermapratyush: why not the java one? 2010-10-26T09:56:29 sprry didnt get u 2010-10-26T09:56:33 *sorry 2010-10-26T09:56:42 i have coded in c++ 2010-10-26T09:58:29 vermapratyush: there's a java tester that comes with the starter pacakage 2010-10-26T09:58:32 package* 2010-10-26T09:59:38 *** vermapratyush_ has joined #aichallenge 2010-10-26T10:00:20 raedwulf: which java tester? 2010-10-26T10:00:31 where can i download it from? 2010-10-26T10:01:19 *** vermapratyush has quit IRC (Ping timeout: 265 seconds) 2010-10-26T10:02:19 vermapratyush_: you are doing ai-contest right? 2010-10-26T10:02:34 yes 2010-10-26T10:02:36 well in the cpp starter package there's a java test program 2010-10-26T10:03:03 the PlayGame.jar file? 2010-10-26T10:03:11 how do i run it? 2010-10-26T10:03:27 yep - its shown in the tutorial/quickstart guide whatever its called 2010-10-26T10:03:33 on the website 2010-10-26T10:03:55 ok.. thanks 2010-10-26T10:07:35 *** worXrillo has quit IRC (Remote host closed the connection) 2010-10-26T10:09:58 raedwulf: it is working thanks a lot? how ever do you know any better simulator? 2010-10-26T10:10:11 raedwulf: it is working thanks a lot.. how ever do you know any better simulator? 2010-10-26T10:11:57 Top 10 players: Simplex(3819), rsergio(3644), dmj111(3628), GreenTea(3575), Hazard(3550), felixcoto(3548), jambachili(3505), asavis(3499), Arekku(3485), george(3473) 2010-10-26T10:15:07 vermapratyush_: i would use the TCP server 2010-10-26T10:16:10 *** JamesMG has joined #aichallenge 2010-10-26T10:17:37 *** smellyhippy has quit IRC (Ping timeout: 272 seconds) 2010-10-26T10:18:21 *** smellyhippy has joined #aichallenge 2010-10-26T10:18:22 *** smellyhippy has joined #aichallenge 2010-10-26T10:21:39 * raedwulf becomes pacman and eats JamesMG 's time 2010-10-26T10:22:19 got no time, this contest already stole it 2010-10-26T10:24:45 *** iNtERrUpT has joined #aichallenge 2010-10-26T10:25:38 JamesMG: and i have 2 other projects + uni work arghhh 2010-10-26T10:25:41 blargghhh 2010-10-26T10:25:56 * raedwulf has encountered a segmentation fault. 2010-10-26T10:28:54 Whoa, I think C# is broken on the official server 2010-10-26T10:29:11 http://ai-contest.com/profile_games.php?user_id=10527 2010-10-26T10:29:25 code that ran fine for almost 2 weeks is suddenly crashing out most of it's games 2010-10-26T10:32:11 happening to the top rated C# bot too: http://ai-contest.com/profile.php?user_id=8368 2010-10-26T10:35:44 JamesMG: http://ai-contest.com/forum/viewtopic.php?f=3&t=995 2010-10-26T10:37:16 Simplex is impressive 2010-10-26T10:39:31 thanks aerique 2010-10-26T10:41:41 *** Titankiller has quit IRC (Ping timeout: 272 seconds) 2010-10-26T10:41:54 mega1: Simplex? 2010-10-26T10:42:00 Why am I limited to seeing only 5 recent games for Simplex? 2010-10-26T10:42:04 look at the leaderboard 2010-10-26T10:42:11 because that's all it has? 2010-10-26T10:42:25 he got a game against george on the 3rd 2010-10-26T10:42:31 oh sorry... not up to the times.. or the minute hehe :D 2010-10-26T10:42:40 new bot? 2010-10-26T10:44:20 *** Titankiller has joined #aichallenge 2010-10-26T10:44:40 wow, his game against dmj111 is nice to watch 2010-10-26T10:47:51 pretty neat 2010-10-26T10:50:49 yep very nice defending 2010-10-26T10:52:28 *** Aragost has joined #aichallenge 2010-10-26T10:54:17 damn that was a good sleep 2010-10-26T10:54:23 pitty it's now 2am 2010-10-26T10:54:37 *** Palmik has quit IRC (Ping timeout: 264 seconds) 2010-10-26T10:55:48 xD, 2am is good :P 2010-10-26T10:55:52 5% battery 2010-10-26T10:55:54 ugh xD 2010-10-26T10:56:04 need to find some power 2010-10-26T10:56:09 i slept from about 7 30 last night, i'm well rested 2010-10-26T10:56:15 i didn't sleep the night before 2010-10-26T10:58:09 *** Sarin has joined #aichallenge 2010-10-26T11:03:16 *** narnach_mb has quit IRC (Quit: Leaving.) 2010-10-26T11:03:44 *** rebelxt has joined #aichallenge 2010-10-26T11:05:11 *** Palmik has joined #aichallenge 2010-10-26T11:06:12 *** Aragost has quit IRC (Quit: chao chao) 2010-10-26T11:08:19 *** Itkovian has quit IRC (Ping timeout: 255 seconds) 2010-10-26T11:11:57 Top 10 players: Simplex(3817), rsergio(3672), dmj111(3609), GreenTea(3542), felixcoto(3521), Hazard(3521), jambachili(3479), Arekku(3471), Xrillo(3466), george(3459) 2010-10-26T11:18:54 *** Frontier has joined #aichallenge 2010-10-26T11:21:56 *** rebelxt has left #aichallenge 2010-10-26T11:25:39 *** lavalamp_ has joined #aichallenge 2010-10-26T11:26:06 *** aerique has quit IRC (Quit: ...) 2010-10-26T11:26:51 *** Xrillo has joined #aichallenge 2010-10-26T11:27:13 *** boegel has quit IRC (Ping timeout: 255 seconds) 2010-10-26T11:29:42 JamesMG, our bots finally played last night, but it's not fair because I fixed a bunch of my bugs: http://72.44.46.68/canvas?game_id=703390 2010-10-26T11:29:55 I noticed your new code was doing very well 2010-10-26T11:31:29 *** tapwater has joined #aichallenge 2010-10-26T11:31:53 I'm just finishing up actually considering possible enemy moves 2010-10-26T11:32:21 Yeah, but then I implemented my current ideas (not in the bots currently playing) and it played worse this morning 2010-10-26T11:33:46 Ah, bummer 2010-10-26T11:34:23 I think I'm going to need to work hard on my evaluation func 2010-10-26T11:35:03 mine's being hamstrung by my move generation at this point I think 2010-10-26T11:35:56 *** xoff has joined #aichallenge 2010-10-26T11:36:10 Hi ! 2010-10-26T11:36:28 Can someone help me ? 2010-10-26T11:36:37 no 2010-10-26T11:36:44 -_- 2010-10-26T11:37:24 it's possible that someone here could help you, but you'd need to ask your question first 2010-10-26T11:37:40 JamesMG yeah you can see my minimax logic only adds 30-50 Elo to my move generator... 2010-10-26T11:38:25 *** seisatsu has quit IRC (Quit: Leaving) 2010-10-26T11:38:48 either way it takes mine clean apart lavalamp_ :) 2010-10-26T11:41:07 xoff, read /topic 2010-10-26T11:41:22 *** xoff has quit IRC (Ping timeout: 265 seconds) 2010-10-26T11:41:34 we scared him away :P 2010-10-26T11:43:35 *** TravisD has joined #aichallenge 2010-10-26T11:44:02 xoff, what is it? 2010-10-26T11:44:07 *** wvdschel has quit IRC (Quit: Leaving) 2010-10-26T11:44:16 oh he left 2010-10-26T11:44:23 :| 2010-10-26T11:46:09 *** seisatsu has joined #aichallenge 2010-10-26T11:58:50 *** Xoff has joined #aichallenge 2010-10-26T11:59:30 Is it possible to change my profile (country) in the challenge ? 2010-10-26T12:01:02 *** irchs has joined #aichallenge 2010-10-26T12:01:49 * Xoff is thinking 2010-10-26T12:05:13 read /topic 2010-10-26T12:05:16 .?! 2010-10-26T12:05:52 Xoff, there's a thread about that in the forums, I think 2010-10-26T12:06:12 ok, thank you ! 2010-10-26T12:09:07 *** Xoff has quit IRC () 2010-10-26T12:11:45 *** yasith_ has quit IRC (Ping timeout: 265 seconds) 2010-10-26T12:11:57 Top 10 players: Simplex(3811), rsergio(3670), dmj111(3609), Hazard(3545), GreenTea(3536), felixcoto(3524), jambachili(3481), Xrillo(3459), protocolocon(3457), george(3454) 2010-10-26T12:18:56 Probably just broke what little my bot had going for it, we'll see! 2010-10-26T12:21:54 *** bhasker has quit IRC (Ping timeout: 265 seconds) 2010-10-26T12:23:16 *** irchs has quit IRC (Quit: irchs) 2010-10-26T12:23:30 *** yasith_ has joined #aichallenge 2010-10-26T12:24:27 *** TravisD has quit IRC (Quit: TravisD) 2010-10-26T12:28:31 *** bhasker has joined #aichallenge 2010-10-26T12:39:12 dmj111, good job beating simplex 2010-10-26T12:40:11 *** mceier has joined #aichallenge 2010-10-26T12:41:38 * raedwulf attacks his code with fury 2010-10-26T12:44:01 *** boegel has joined #aichallenge 2010-10-26T12:45:11 *** p4p4_ has quit IRC (Ping timeout: 272 seconds) 2010-10-26T12:49:50 *** hellman has quit IRC (Remote host closed the connection) 2010-10-26T12:50:21 *** boegel has quit IRC (Ping timeout: 252 seconds) 2010-10-26T12:57:01 *** iNtERrUpT has quit IRC (Remote host closed the connection) 2010-10-26T13:11:50 morning 2010-10-26T13:11:56 i just rolled back a full morning of fixes 2010-10-26T13:11:58 Top 10 players: Simplex(3714), rsergio(3701), dmj111(3631), GreenTea(3551), felixcoto(3532), Hazard(3523), protocolocon(3485), jambachili(3481), george(3462), sin_sun(3454) 2010-10-26T13:12:01 :-\ 2010-10-26T13:12:11 strangely these bots at the top play very differently 2010-10-26T13:12:21 they don't seem to accumulate ships to take over neutrals 2010-10-26T13:12:26 they seem to stream ships at neutrals 2010-10-26T13:13:13 yea, not sure if thats advantagous or now 2010-10-26T13:13:15 not* 2010-10-26T13:13:46 it would seem that would lead to sniping, but I guess if you do the math and outproduce the opponent, or tie with them 2010-10-26T13:13:53 *** Epona666 has quit IRC (Quit: Leaving) 2010-10-26T13:13:55 then you can stream at it and he can't do anything about it 2010-10-26T13:15:00 yea its no differce either way 2010-10-26T13:15:13 bbl going home 2010-10-26T13:16:10 *** jesionaj has joined #aichallenge 2010-10-26T13:18:34 *** irchs has joined #aichallenge 2010-10-26T13:20:05 *** Naktibalda has quit IRC (Quit: ChatZilla 0.9.86 [Firefox 3.6.8/20100722155716]) 2010-10-26T13:20:09 *** O1athe has joined #aichallenge 2010-10-26T13:22:25 *** Olathe has quit IRC (Ping timeout: 255 seconds) 2010-10-26T13:23:59 *** p4p4 has joined #aichallenge 2010-10-26T13:26:21 yea not sure how that works out 2010-10-26T13:26:26 i guess you save one turn don't know 2010-10-26T13:28:20 I think streaming saves you from committing all your forces 2010-10-26T13:29:28 hmm not really 2010-10-26T13:29:33 since if the planet is far enough 2010-10-26T13:29:43 you are practically committing the same number? 2010-10-26T13:29:52 i guess one could use that as a heuristic 2010-10-26T13:29:56 for closer planets stream? 2010-10-26T13:30:05 maybe 2010-10-26T13:30:06 and further away just commit the whole thing not sure 2010-10-26T13:30:46 maybe it has to do with how close the enemy is to your sending planet... 2010-10-26T13:31:16 *** fawek has joined #aichallenge 2010-10-26T13:31:35 haha http://72.44.46.68/canvas?game_id=714421 2010-10-26T13:31:38 interesting base swap 2010-10-26T13:38:22 haha, ended up in a loop there 2010-10-26T13:38:28 just building up forces never ending.. 2010-10-26T13:40:01 malazan 11 seems to be doing quite well 2010-10-26T13:49:33 *** boegel has joined #aichallenge 2010-10-26T13:51:24 *** Accoun has quit IRC (Ping timeout: 252 seconds) 2010-10-26T13:52:55 I wonder if it's just implementation details that separate the top few bots from the rest of if they have thought of additional strategies 2010-10-26T13:53:02 *or if 2010-10-26T13:54:41 bhasker, what'd you do between malazan 10 and 11, if you don't mind my asking? 2010-10-26T13:57:47 nothing 2010-10-26T13:57:55 i was tinkering with 10 2010-10-26T13:57:58 and it got worse 2010-10-26T13:58:00 so i rolled back 2010-10-26T13:58:04 and started up as 11 2010-10-26T13:58:17 since 10's elo got completly screwed with the tinkering 2010-10-26T13:58:56 *** Florian95 has quit IRC (Remote host closed the connection) 2010-10-26T13:59:24 i am going to let this burn for a day without any tinkering 2010-10-26T13:59:27 then use the games data 2010-10-26T13:59:38 to identify and fix the bugs 2010-10-26T13:59:43 no more tinkering 2010-10-26T13:59:56 this version seems to be doing reasonably well on TCP 2010-10-26T14:02:16 *** noid42999 has joined #aichallenge 2010-10-26T14:02:27 *** noid42999 is now known as Accoun 2010-10-26T14:02:55 whats the tcp website again? 2010-10-26T14:03:02 contestbot: tcp 2010-10-26T14:03:02 !tcp 2010-10-26T14:03:03 amstan: tcp = (#1) http://www.benzedrine.cx/planetwars/, or (#2) http://72.44.46.68/ 2010-10-26T14:03:40 *** narnach has quit IRC (Quit: Leaving.) 2010-10-26T14:04:31 bhasker, lol... do you manually look at all the games you lose or do you have some system? 2010-10-26T14:05:00 i look at a few ones 2010-10-26T14:05:14 *** nann has joined #aichallenge 2010-10-26T14:05:17 don't have an automated system to figure out the mistake 2010-10-26T14:05:21 maybe i should write one ^ 2010-10-26T14:05:34 most of my bugs are of the two kinds 2010-10-26T14:05:40 a) on some maps i just pick wrong starting planets 2010-10-26T14:05:47 b) acquire the wrong neutral midgame 2010-10-26T14:06:09 I was thinking of logging something whenever my evaluation suddenly changes 2010-10-26T14:06:23 yeah I have both those problems also 2010-10-26T14:06:25 well mine is not a mini max or anything its a heuristic bot 2010-10-26T14:06:38 so sometimes the heuristics just fail 2010-10-26T14:06:46 so not easy to identify the fault 2010-10-26T14:06:53 hm good point 2010-10-26T14:06:56 have to look at the logs and then look at my code back and forth 2010-10-26T14:07:02 to figure out the bug 2010-10-26T14:07:20 i am debating writing a minimax version 2010-10-26T14:07:31 but its so much work and i don't have much experience writing minimax solutions ^ 2010-10-26T14:07:46 ...and I'm not so sure it's working out, hehe 2010-10-26T14:08:47 *** Apophis has quit IRC (Ping timeout: 272 seconds) 2010-10-26T14:08:58 yours seems to be getting there, lavalamp_ 2010-10-26T14:09:17 Thanks ^^ 2010-10-26T14:09:18 and mostly if i do write a minimax bot i might have to write it in c++ or something 2010-10-26T14:09:21 which i am not very keen on 2010-10-26T14:09:29 what is your current one in? 2010-10-26T14:09:32 python 2010-10-26T14:09:40 mostly for the speed boost 2010-10-26T14:09:43 ah 2010-10-26T14:09:44 i would need to do it in c++ 2010-10-26T14:09:49 is python that slow? 2010-10-26T14:10:07 will probably be given that minimax will require massive computation 2010-10-26T14:10:10 it will add up 2010-10-26T14:10:15 when you are simulating nodes 2010-10-26T14:10:16 *** perror has quit IRC (Quit: Bye all !) 2010-10-26T14:10:40 hm, I think I just thought of a way to cache some of my calculations to get my node count back up... 2010-10-26T14:11:06 yeah mine was doing 10-20k before my latest changes 2010-10-26T14:11:20 wow what the hell kind of map is this? http://72.44.46.68/canvas?game_id=705301 2010-10-26T14:11:23 I got mine down quite a bit this morning 2010-10-26T14:11:47 map 701 2010-10-26T14:11:49 a1k0n, that's not the only draw map I've seen lately 2010-10-26T14:11:49 yea thats a buggy map 2010-10-26T14:11:55 there are a couple more 2010-10-26T14:11:58 Top 10 players: rsergio(3729), Simplex(3666), dmj111(3616), GreenTea(3561), felixcoto(3530), Hazard(3511), protocolocon(3492), george(3465), jambachili(3458), sin_sun(3443) 2010-10-26T14:12:07 one with a 2 planets of 1 growth with 5 ships on them 2010-10-26T14:12:09 to the sides 2010-10-26T14:12:10 yeah, those must have been made before the map generator bugfix 2010-10-26T14:12:14 usually draws too 2010-10-26T14:13:40 bug fixes brought me from -550 to +50 but I'm not sure how much further I can get that way... 2010-10-26T14:14:20 bug fixes help a lot 2010-10-26T14:14:40 i still have one bug that i need to fix i don't snipe neutrals wlel i tend to launch a turn ahead of when i should ^ 2010-10-26T14:14:41 *** Ari2 has joined #aichallenge 2010-10-26T14:15:02 i just fixed up my move coordination so that it works a lot better 2010-10-26T14:15:03 bhasker, lol, I had that bug in my last bot, I think 2010-10-26T14:15:19 i played under retard on tcp just before 2010-10-26T14:15:24 was going great until Accoun 2010-10-26T14:15:27 yea been trying to fix it 2010-10-26T14:15:31 need to watch games 2010-10-26T14:15:38 my bot is doing reasonably well against accoun* 2010-10-26T14:15:47 but it fails against ObligatoryShamoo :((( 2010-10-26T14:15:48 he got my 2/3 just before 2010-10-26T14:15:50 very annoying 2010-10-26T14:15:53 but i made some pretty stupid moves 2010-10-26T14:15:56 accoun needs to fix close maps 2010-10-26T14:16:08 yeah, obligatory shampoo gives me a lot of pain too 2010-10-26T14:16:11 http://72.44.46.68/getplayer?player=retard 2010-10-26T14:16:12 o.shampoo kills me too 2010-10-26T14:16:24 except his elo rating is like 150 points below my bots 2010-10-26T14:16:27 and i keep losing to it 2010-10-26T14:16:29 same 2010-10-26T14:16:29 :-\ 2010-10-26T14:16:36 it's annoying as fuck aha 2010-10-26T14:16:42 yup 2010-10-26T14:16:44 yeah I don't understand who he loses to that he stays at that rating 2010-10-26T14:17:04 i think he might have one of those bots that isn't very good generally 2010-10-26T14:17:08 just sometimes? 2010-10-26T14:17:20 hence the elo screws up, but bothers better bots on some maps 2010-10-26T14:17:49 i guess that's a flaw with elo, you're assuming the bots are of equal "goodness" on each map 2010-10-26T14:18:36 alrigth lavalamp, lets do this 2010-10-26T14:18:56 haha did we get matched? 2010-10-26T14:19:12 yeah, i have it running as retard atm 2010-10-26T14:19:21 it was a lot more retarded yesterday, i'm ironing out bugs 2010-10-26T14:19:23 I think you'll probably win 2010-10-26T14:19:36 :p 2010-10-26T14:20:42 *** espes has quit IRC (Ping timeout: 240 seconds) 2010-10-26T14:20:44 heh 2010-10-26T14:21:10 awww, beaten by devtcp 2010-10-26T14:21:17 so antimatroid do you have staging code 2010-10-26T14:21:19 and then paired against shampoo bastard 2010-10-26T14:21:22 staging? 2010-10-26T14:21:30 like do you explicitly forward ships 2010-10-26T14:21:33 or is it just minimax at work 2010-10-26T14:21:58 i've taken minimax out for the moment, it's too complicated to consider sending my "future" ships and have the opponent moving against me 2010-10-26T14:22:16 it's just doing defense moves then a tree of individual offense and expansionary moves 2010-10-26T14:22:20 then picking what looks best 2010-10-26T14:22:23 hmm k 2010-10-26T14:22:31 and adding in redistribution where it can 2010-10-26T14:22:59 me and dev_tcp are going back and forth 2010-10-26T14:23:03 http://72.44.46.68/getplayer?player=malazan0.11 2010-10-26T14:23:08 *** Itkovian has joined #aichallenge 2010-10-26T14:23:35 8 games in a row 2010-10-26T14:23:38 4 mine 4 his 2010-10-26T14:23:41 i just got my ass handed to me twice by obligatoryshampoo 2010-10-26T14:24:05 goodbye elo score 2010-10-26T14:24:42 heh 2010-10-26T14:25:44 antimatroid1, I was doing fine until I went for that neutral at the bottom of the screen XP 2010-10-26T14:26:28 yep :P 2010-10-26T14:26:32 probably would have had me otherwise 2010-10-26T14:26:39 mcleopold just started up his antbot 2010-10-26T14:26:50 and i lost to it:( 2010-10-26T14:27:34 *** espes has joined #aichallenge 2010-10-26T14:27:53 there's still so much room for improvement on bots for everybody 2010-10-26T14:28:12 it's damn hard to get near perfect play for maps though 2010-10-26T14:28:17 with a general bot 2010-10-26T14:29:30 oh i lost by 9 ships:-\ to antbot 2010-10-26T14:29:43 11 rather 2010-10-26T14:29:44 http://72.44.46.68/canvas?game_id=715285 2010-10-26T14:29:45 :-\ 2010-10-26T14:31:17 now who is jb.exp :-\ 2010-10-26T14:34:19 *** RainCT has joined #aichallenge 2010-10-26T14:37:20 antimatroid obligatoryshampoo just destroyed your elo rating 2010-10-26T14:37:21 :-\ 2010-10-26T14:39:55 the best way to get near perfect play for each map is to have them separately in the bot, somehow 2010-10-26T14:40:21 but that won't help in general 2010-10-26T14:40:39 *** AJC has joined #aichallenge 2010-10-26T14:41:00 hey, has anyone seen j3camero around? (if not here in the forums or otherwise) 2010-10-26T14:41:16 k i have a question for this game, given infinite time and an exhaustive tree search will the bots draw if its a minmax formulation 2010-10-26T14:41:23 since maps are symmetrical 2010-10-26T14:41:49 bhasker, the graph is too big to search. 2010-10-26T14:41:59 wel thats why i said if you could do it 2010-10-26T14:42:11 bhasker, I guess so 2010-10-26T14:42:14 i know its not possible to do it in a reasonable amount of tiime 2010-10-26T14:42:18 *** amstan has quit IRC (Ping timeout: 240 seconds) 2010-10-26T14:42:27 it also depends on how well a bot can predict the other 2010-10-26T14:42:41 i got a game against dmj the muggle 2010-10-26T14:43:02 hah nice 2010-10-26T14:43:03 bhasker, there's some debate as to whether there are rock-paper-scissor situations or if it's a draw 2010-10-26T14:43:29 hmm 2010-10-26T14:43:36 has no one cooked up an example yet 2010-10-26T14:43:42 shesh, give me 10 minutes 2010-10-26T14:43:50 some people have, yes. 2010-10-26T14:43:56 there isn't a canonical one, though 2010-10-26T14:44:07 canonical as in symmetric? 2010-10-26T14:44:16 canonical as in one everybody knows 2010-10-26T14:44:27 the examples i've seen are symmetric 2010-10-26T14:44:34 my bot is set up to detect such situations and play mixed strategies in them, but it's not good enough for that to be meaningful yet 2010-10-26T14:47:00 okay, i can do it, give me more than 10 minutes though and i'll work the details out properly 2010-10-26T14:47:08 AJC: it will be a draw, because they would make identical moves (if you try a bot that has a deterministic playing style with itself, it will be a draw) 2010-10-26T14:47:31 raedwulf, but if I can predict your move I will move differently. 2010-10-26T14:47:46 yes, but both moves occur at the same time in a turn 2010-10-26T14:47:56 its not one person moves before the other 2010-10-26T14:47:59 its we have a world 2010-10-26T14:48:00 due to the fact you can't exaustively search the space, what you predict affects the behavior. 2010-10-26T14:48:29 well its whether the two bots have the same idea of "good" 2010-10-26T14:48:45 if the bots are identical, it will come to a draw because it will end up being deterministic 2010-10-26T14:49:04 OK, I agree on that. 2010-10-26T14:49:13 if there's two different bots with different ideas of "good" but search exhaustively, it may not be a draw 2010-10-26T14:50:10 *** Appleman1234 has joined #aichallenge 2010-10-26T14:50:14 ok my last line doesn't quite work but heh 2010-10-26T14:51:11 heh this looks my best iteration on the bot, its been holding 400+ elo rating for over 80 games 2010-10-26T14:51:16 normally it drops to low 300's 2010-10-26T14:51:22 after 20-30 games 2010-10-26T14:51:23 actually if you knew the entire search space... there is no way you could not draw 2010-10-26T14:51:27 bhasker, nice 2010-10-26T14:51:42 73 games played 2010-10-26T14:51:48 raedwulf, it depends if there are rps situations 2010-10-26T14:51:59 lavalamp_: rps? 2010-10-26T14:52:08 rock paper scissors 2010-10-26T14:52:29 as in, defend beats attack beats expand beats defend 2010-10-26T14:52:40 not if you knew the entire search space 2010-10-26T14:52:42 thats his premise 2010-10-26T14:52:54 i think its inflated till i play ObligatoryShampoo then i will drop lol 2010-10-26T14:52:57 search exhaustively = being able to see the entire search space 2010-10-26T14:53:00 even if you know the entire space there could still be such situations 2010-10-26T14:53:38 in other words, the best thing for you to do depends on what your opponent does on the same turn 2010-10-26T14:53:43 no, i don't believe so, because the bot would be able to know which would lead to its own loss 2010-10-26T14:54:02 if you took alternate turns with your opponent and could see his moves you would be right 2010-10-26T14:54:12 raedwulf, the fact that you move at the same time changes things, this is not chess 2010-10-26T14:54:27 hmmmm 2010-10-26T14:54:30 * raedwulf thinks some more 2010-10-26T14:54:54 i suppose that would be the case 2010-10-26T14:54:55 but as lavalamp said it's debatable whether such a mutually exclusive situation exists 2010-10-26T14:55:04 right, we don't know either way 2010-10-26T14:55:27 ok, i agree 2010-10-26T14:55:45 so... is a draw possible? 2010-10-26T14:55:57 if the bots play deterministically the same way :P 2010-10-26T14:56:06 a draw is possible yes 2010-10-26T14:56:14 we are wondering if its "always" a draw 2010-10-26T14:57:04 since no one has found an rps situation, I'm guessing they're rare if they exist-- so I'll go out on a limb and say that most maps should probably be draws 2010-10-26T14:57:16 i think this does it 2010-10-26T14:57:23 http://imgur.com/p2XGG.png 2010-10-26T14:57:29 i kept it in 1d to make things easier 2010-10-26T14:57:42 that's good, i can exhaustively search that 2010-10-26T14:57:57 you either wait, expand, or attack their neutral 2010-10-26T14:58:43 if you both do the same you draw, 2010-10-26T14:58:44 if you expand and they attack you lose, 2010-10-26T14:58:44 if you expand and they wait, you win 2010-10-26T14:58:44 if you attack and they wait you lose 2010-10-26T14:58:47 and vice versa for each 2010-10-26T14:59:21 what do you mean by attackL 2010-10-26T14:59:25 attack? 2010-10-26T14:59:36 attack the neutral closest to the enemy that turn 2010-10-26T14:59:43 ah 2010-10-26T14:59:49 so don't consider attacking the enemies planet, assume they wont send that much 2010-10-26T14:59:59 it almost goes without saying you should send 6 ships to the enemy neutral on turn 1 2010-10-26T15:00:07 but if they wait you lose 2010-10-26T15:00:36 well 2010-10-26T15:00:44 how about sending 100 ships to your neutral right away? 2010-10-26T15:00:58 if they send all their ships at that planet you lose 2010-10-26T15:01:42 d=1 isn't allowed, but numbers can be changed to represent the same situation 2010-10-26T15:02:25 am i right that there is no "best" strategy there other than to randomise? 2010-10-26T15:02:39 still thinking 2010-10-26T15:03:16 i think in reality you'd want to go for your neutral, as most bots aren't going to be trying to snipe you that move 2010-10-26T15:04:16 what if you move all 100, and then 5 for each turn after that, to your neutral? 2010-10-26T15:04:23 I'm still running the math but... 2010-10-26T15:04:33 antimatroid1 I think you have a winner 2010-10-26T15:04:33 you lose 2010-10-26T15:04:43 if they send all their ships and future ships at that neutral 2010-10-26T15:04:50 but only if they do it the first turn 2010-10-26T15:04:53 if they wait, you win 2010-10-26T15:04:54 yeah you end up behind by 4 2010-10-26T15:05:06 interesting 2010-10-26T15:05:14 oh I saw that as d=2, my bad 2010-10-26T15:05:18 yeah, ok 2010-10-26T15:05:23 so best solution would be for the bot to do nothing and draw the game ? 2010-10-26T15:05:24 it's exactly isomorphic to rock paper scissors 2010-10-26T15:05:30 sweet, in this map my bot *should* detect this and randomly chose one 2010-10-26T15:05:36 hmm 2010-10-26T15:05:38 bhasker: if they expand you lose 2010-10-26T15:05:46 it's exactly rock paper scissors 2010-10-26T15:05:58 thanks for finding this! 2010-10-26T15:06:19 it gets more complicated once you take it to 2d and players planets get closer 2010-10-26T15:06:21 way more complicated 2010-10-26T15:06:37 hmm 2010-10-26T15:07:03 i hadn't done that, i should go and solve the game in 1d 2010-10-26T15:07:06 then work on 2d 2010-10-26T15:07:10 2d makes my brain want to die 2010-10-26T15:07:11 so i wonder what the conditions are for a mutually exclusive strategy to exist... if the neutrals have 5 ships then it's a draw right? 2010-10-26T15:07:29 i think 5 is the same situation yeah? 2010-10-26T15:07:30 hmmmm what if you send 33 ships at each planet?? 2010-10-26T15:07:36 5-9 should be the same situation 2010-10-26T15:08:11 if it's 4 and you expand and they wait you lose, so you'd always wait 2010-10-26T15:08:11 well you'd gain back 5 ships before they could reach your neutral 2010-10-26T15:08:21 should 5-100 be the same, in that moving there guarantees the enemy gets the planet if he also goes there? 2010-10-26T15:08:44 oh wait hang on 2010-10-26T15:08:56 so taking your own neutral would pay for itself before your enemy's "light cone" intersects yours 2010-10-26T15:09:08 yeah, if there are more than 9 ships at the neutrals the best you can do is wait and dra 2010-10-26T15:09:09 w 2010-10-26T15:09:40 it also gets a lot more tricky when planets are changing ownership multiple times in the future 2010-10-26T15:10:07 I'd be tempted to send one ship to my neutral in this case 2010-10-26T15:10:16 just to try and trick the other player into "sniping" it for me 2010-10-26T15:10:38 i doubt that would work with anyone's bot 2010-10-26T15:10:46 shouldn't with most of them 2010-10-26T15:10:59 it just feels better than relying entirely on chance :) 2010-10-26T15:11:01 yeah ok... send 46 ships to their planet, and 27 to each neutral... 2010-10-26T15:11:54 gahh even this map is insanely complicated 2010-10-26T15:11:59 Top 10 players: rsergio(3719), Simplex(3668), dmj111(3628), GreenTea(3577), felixcoto(3536), Hazard(3510), protocolocon(3494), jambachili(3461), george(3459), sin_sun(3441) 2010-10-26T15:12:15 yeah, it's a ridiculously hard game 2010-10-26T15:12:34 there's just so many possible moves 2010-10-26T15:12:35 I think that beats waiting, expanding, and attacking 2010-10-26T15:12:57 *** Frontier has left #aichallenge 2010-10-26T15:13:13 :( 2010-10-26T15:13:39 i have an idea 2010-10-26T15:14:01 i'm going to write a program to iterate best responses to moves by the other player for simple maps like this 2010-10-26T15:14:14 and see if it can't converge on winning /drawing first strategies 2010-10-26T15:14:17 i actually wrote that program 2010-10-26T15:14:20 it may cycle on me though 2010-10-26T15:14:29 how'd it go? 2010-10-26T15:14:32 but it finds all possible moves exhaustively, and.. there are a lot 2010-10-26T15:14:43 yeah, it wouldn't be for the bot 2010-10-26T15:14:44 i had to limit depth and make assumptions 2010-10-26T15:14:49 well right 2010-10-26T15:15:19 100 ships is too many to search exhaustively 2010-10-26T15:15:20 i'm curious about getting solutions to some of the simpler maps 2010-10-26T15:15:22 it's very very hard 2010-10-26T15:15:52 meh they took a long while to even solve connect 4 2010-10-26T15:15:56 and you're trying to solve this 2010-10-26T15:15:57 :P 2010-10-26T15:16:04 what i have now makes a map with 10-n-n-10 (similar to your setup), 2 steps between all planets, n varies 2010-10-26T15:16:58 raedwulf: it's hard to solve situations with even 3-4 planets 2010-10-26T15:17:02 it generates a score for all player 1 moves 2010-10-26T15:17:19 i do have a game theory library, maybe i should try and brute force nash equilibria 2010-10-26T15:17:22 *** Frontier has joined #aichallenge 2010-10-26T15:17:25 and they're all <= 0.5 naturally; in some situations they're always <0.5 because player 2's move is conditional on player 1's 2010-10-26T15:17:30 i doubt it'd manage it 2010-10-26T15:18:10 *** hornairs has joined #aichallenge 2010-10-26T15:18:24 so basically i have the data but failed to draw any useful conclusions 2010-10-26T15:20:32 i miss the comfort of tron 2010-10-26T15:20:45 yeah that was a lot easier to wrap your head around 2010-10-26T15:21:01 yet still had a lot of details i loved it 2010-10-26T15:21:22 this is cool too, it's just such a mind musher 2010-10-26T15:22:52 *** _W_ <_W_!~w@irc.w-wins.com> has quit IRC (Ping timeout: 255 seconds) 2010-10-26T15:22:52 *** AlliedEnvy has quit IRC (Ping timeout: 255 seconds) 2010-10-26T15:23:51 *** AlliedEnvy has joined #aichallenge 2010-10-26T15:24:15 *** _W_ <_W_!~w@irc.w-wins.com> has joined #aichallenge 2010-10-26T15:25:10 *** Accoun has quit IRC (Ping timeout: 276 seconds) 2010-10-26T15:31:03 *** vermapratyush_ has quit IRC (Quit: Page closed) 2010-10-26T15:34:20 *** herschel has joined #aichallenge 2010-10-26T15:47:03 its funny a lot of the top bots lose to ObligatoryShampoo 2010-10-26T15:47:06 heh 2010-10-26T15:48:00 hmm it has 3641 games maybe the owner is updating the bot and not changing the name 2010-10-26T15:48:04 thats why the elo is so depressed 2010-10-26T15:48:11 the new version is probably much better 2010-10-26T15:49:10 I thought the server forgot the games after 24 hours 2010-10-26T15:50:59 *** davidjliu has joined #aichallenge 2010-10-26T15:51:27 it does 2010-10-26T15:53:33 *** seisatsu has quit IRC (Quit: Leaving) 2010-10-26T15:56:29 *** Olathe has joined #aichallenge 2010-10-26T15:59:53 *** O1athe has quit IRC (Ping timeout: 265 seconds) 2010-10-26T16:10:16 *** Accoun has joined #aichallenge 2010-10-26T16:12:01 Top 10 players: rsergio(3716), dmj111(3613), GreenTea(3569), Simplex(3556), felixcoto(3529), protocolocon(3507), Hazard(3478), jambachili(3458), george(3451), Arekku(3447) 2010-10-26T16:24:54 honestly right now if it were completely up to me I'd say screw it and just uninstall/disable C# 2010-10-26T16:25:49 janzert, lol... I left a note on the go issue. Do you want me to make a shell script for you? 2010-10-26T16:26:16 bah i think the only reason my bot is so high up right now is because none of the good bots are running 2010-10-26T16:26:33 and here i was thinking it might actually be better 2010-10-26T16:27:10 *** irchs has quit IRC (Quit: irchs) 2010-10-26T16:27:13 bhasker, I've noticed that the tcp server ratings vary by ~100 Elo depending on who's playing and/or time of day and phase of moon 2010-10-26T16:27:20 yea 2010-10-26T16:27:24 *** mceier has quit IRC (Quit: leaving) 2010-10-26T16:27:29 So I always run a control bot as well as the one I'm testing 2010-10-26T16:27:56 probably is as-d* and hazard and accoun bots were all up my bot will be at 350ish 2010-10-26T16:28:00 if* 2010-10-26T16:28:03 Well... I'm most comfortable using pre-made (tested) packages from a repository, I'm less comfortable using a .deb, I'm even less comfortable using a released archive, and I don't know yet if I want to touch pulling from source control 2010-10-26T16:28:50 hmmm 2010-10-26T16:28:59 but I suppose a shell script showing exactly what needs to be done would help if I get over that :) 2010-10-26T16:29:31 hm does it need to be installed for all users or does one user do the compiling? 2010-10-26T16:29:59 well, one user on the main server and one user on each worker 2010-10-26T16:30:25 remember that I have to incorporate anything into installing on each worker too 2010-10-26T16:30:32 ok so I can install to the user's home directory? 2010-10-26T16:30:45 right, do you have a script or something that sets them up? 2010-10-26T16:31:21 http://code.google.com/p/ai-contest/source/browse/branches/20100929-games-in-the-cloud/planet_wars/backend/game_server_build.sh 2010-10-26T16:31:40 *** Titankiller2 has joined #aichallenge 2010-10-26T16:32:30 fyi, the current language that doesn't use a package is node and that was set up by danielvf 2010-10-26T16:32:35 *** Titankiller has quit IRC (Ping timeout: 240 seconds) 2010-10-26T16:32:55 but then disabled from running any bots on workers with it for some reason 2010-10-26T16:33:00 ok... I can use virtualbox at home to test this, just need to download ubuntu 8 2010-10-26T16:33:06 *** boegel has quit IRC (Quit: Leaving) 2010-10-26T16:33:10 ls -l map*.txt | wc -l 2010-10-26T16:33:10 999 2010-10-26T16:33:18 for i in map*.txt; do md5sum $i; done | cut -d" " -f1 | sort | uniq | wc -l 2010-10-26T16:33:18 899 2010-10-26T16:33:25 WTF?! 2010-10-26T16:33:28 lavalamp_: yep 2010-10-26T16:34:43 would you prefer I do something like unzip an archive or use hg to download and compile like I do on my system? 2010-10-26T16:34:51 lol 2010-10-26T16:34:51 out of the 999 maps there are 100 duplicates 2010-10-26T16:35:00 mega1: lol 2010-10-26T16:35:10 if there is a release archive that would be preferable 2010-10-26T16:35:24 don't need to worry about adding hg to each worker that way 2010-10-26T16:35:32 makes sense 2010-10-26T16:35:59 ok so perhaps I should just make binaries? 2010-10-26T16:36:32 hm, it would need the packages, too 2010-10-26T16:36:43 I can prebuild it 2010-10-26T16:37:45 that way I just extract an archive and put some binaries in the right spot (or modify PATH) 2010-10-26T16:38:16 where would they be hosted? 2010-10-26T16:38:48 I'm really quite surprised the go people have no release archives available 2010-10-26T16:38:57 it's a language in flux 2010-10-26T16:39:09 there's someone who makes windows binaries 2010-10-26T16:39:42 I'll see if I have enough space at bitbucket to put it up 2010-10-26T16:40:08 *** narnach has joined #aichallenge 2010-10-26T16:40:58 the have releases marked in hg though right? 2010-10-26T16:41:00 apparently map201 - map300 are all the same 2010-10-26T16:41:31 or do we have to pick an arbitrary revision and hope it's stable 2010-10-26T16:42:12 * janzert really really doesn't want to be pulling from the head of a development branch 2010-10-26T16:43:32 but those maps are not listed on http://72.44.46.68/getmaps, so the problem is known I guess 2010-10-26T16:44:14 janzert, yeah, they tag the releases 2010-10-26T16:44:42 *** AJC has left #aichallenge ("Leaving") 2010-10-26T16:44:44 Can you bring me up to speed on C#? I'll go install a VM and work on it if it'll help 2010-10-26T16:44:50 rather than rewrite my bot... :( 2010-10-26T16:45:37 JamesMG: basically all the versions available for ubuntu 8.04 leak semaphores like crazy and eventually will crash on start up 2010-10-26T16:46:05 at this point I've given up on getting something that doesn't leak and am just trying to clean the semaphores up after it 2010-10-26T16:46:22 That's nasty 2010-10-26T16:46:22 http://72.44.46.68/canvas?game_id=717545 hahah 2010-10-26T16:46:50 I'm hopeful that I've actually accomplished that on the workers now, but letting it run for a bit before I port it back to the main server 2010-10-26T16:47:11 ok janzert, I will try to come up with something tonight or tomorrow morning 2010-10-26T16:47:27 lavalamp_: sounds good, thanks 2010-10-26T16:48:30 submitting stuff for the servers won't get my entry disqualified, will it? :) 2010-10-26T16:48:45 nope as long as you don't have access to the servers 2010-10-26T16:49:04 ok 2010-10-26T16:49:15 btw, if we are pulling an archive anyway building from source isn't going to be a big deal and may be preferable 2010-10-26T16:49:25 ok 2010-10-26T16:49:48 in that case I can just pull from the official site and update to a specific release 2010-10-26T16:50:09 unless I suppose there are a bunch of dependencies needed to build it but not run it 2010-10-26T16:50:33 There are, but I think you'll have them already for building c/c++ programs 2010-10-26T16:51:02 yeah, if it's just the basic compiler and such. if it's a bunch of libraries probably not 2010-10-26T16:51:16 no special libraries, I don't think 2010-10-26T16:51:31 ok in that case it should be pretty easy 2010-10-26T16:52:48 ok, well working from the 8.04 virtualbox letting the game_server_build script set it up should let you know what dependencies are already there or not 2010-10-26T16:53:52 exactly 2010-10-26T16:54:17 does that script have any dependancies? 2010-10-26T16:54:44 shouldn't I think I've got it so it should work from a vanilla 8.04 install 2010-10-26T16:54:51 great 2010-10-26T16:55:19 in fact using ubuntu-8.04.4-server-i386.iso in virtualbox here for testing 2010-10-26T16:56:48 oh are they 32 bit systems or 64? 2010-10-26T16:57:08 well, it won't matter for the script 2010-10-26T16:57:18 32bit 2010-10-26T16:57:46 hm maybe I should lose my float64's lol 2010-10-26T16:58:10 :) 2010-10-26T17:02:57 *** smellyhippy has quit IRC (Ping timeout: 272 seconds) 2010-10-26T17:04:46 *** Mathnerd314 has joined #aichallenge 2010-10-26T17:05:11 *** Titankiller2 has quit IRC (Quit: Nettalk6 - www.ntalk.de) 2010-10-26T17:05:28 *** Titankiller has joined #aichallenge 2010-10-26T17:07:08 *** smellyhippy has joined #aichallenge 2010-10-26T17:07:32 *** lavalamp_ has quit IRC (Quit: bye) 2010-10-26T17:12:00 Top 10 players: rsergio(3741), dmj111(3599), GreenTea(3578), felixcoto(3516), Simplex(3505), protocolocon(3505), Hazard(3472), jambachili(3466), george(3457), Arekku(3449) 2010-10-26T17:12:27 *** Itkovian has quit IRC (Quit: Itkovian) 2010-10-26T17:12:42 *** Itkovian has joined #aichallenge 2010-10-26T17:13:33 *** delt0r___ has joined #aichallenge 2010-10-26T17:15:02 *** delt0r has quit IRC (Ping timeout: 250 seconds) 2010-10-26T17:15:15 *** Naktibalda has joined #aichallenge 2010-10-26T17:15:19 Yay, my official bot got to actually play a string of games 2010-10-26T17:15:33 thanks for working on C# janzert, I do appreciate it 2010-10-26T17:15:42 hopefully they weren't all crashes :} 2010-10-26T17:15:52 nah, that's what I meant, haha 2010-10-26T17:15:57 :) 2010-10-26T17:20:03 *** p4p4 has quit IRC (Quit: ChatZilla 0.9.84 [SeaMonkey 2.0a3/20090223135443]) 2010-10-26T17:22:44 *** nann has quit IRC (Read error: Connection reset by peer) 2010-10-26T17:26:55 *** Azrathud has joined #aichallenge 2010-10-26T17:30:56 whois ObligatoryShampoo 2010-10-26T17:31:01 on tcp 2010-10-26T17:33:49 same as CaptainKangoroo was 2010-10-26T17:34:36 or there are several contestants who use the same name generator 2010-10-26T17:38:01 NameGenerator 2010-10-26T17:38:13 would be a great name for a bot 2010-10-26T17:38:47 *** Sarin has quit IRC (Remote host closed the connection) 2010-10-26T17:44:07 *** Commandir has left #aichallenge 2010-10-26T17:49:17 *** RainCT has quit IRC (Remote host closed the connection) 2010-10-26T17:49:28 *** Appleman1234 has quit IRC (Ping timeout: 276 seconds) 2010-10-26T17:51:26 ugh, just reverted all of today's code 2010-10-26T17:57:03 welcome to the club 2010-10-26T17:57:53 *** sigh has joined #aichallenge 2010-10-26T18:00:24 *** amstan has joined #aichallenge 2010-10-26T18:02:14 *** iFire has quit IRC (Read error: Connection reset by peer) 2010-10-26T18:02:42 *** iFire has joined #aichallenge 2010-10-26T18:03:01 *** perestrelka has quit IRC (Ping timeout: 264 seconds) 2010-10-26T18:06:58 *** n9thbit has joined #aichallenge 2010-10-26T18:08:45 @rankings 2010-10-26T18:08:46 amstan: Top 10 players: rsergio(3692), dmj111(3611), GreenTea(3564), Simplex(3543), felixcoto(3534), protocolocon(3503), Hazard(3478), jambachili(3459), Arekku(3458), george(3455) 2010-10-26T18:09:09 hmm, not even sure if this is a spam bot http://ai-contest.com/forum/viewtopic.php?f=17&t=1002 2010-10-26T18:09:44 ok, except that they spammed it to every forum 2010-10-26T18:12:01 Top 10 players: rsergio(3691), dmj111(3610), GreenTea(3564), Simplex(3542), felixcoto(3534), protocolocon(3502), Hazard(3478), jambachili(3458), Arekku(3458), george(3454) 2010-10-26T18:12:49 *** Appleman1234 has joined #aichallenge 2010-10-26T18:15:45 oh well, hes gone now 2010-10-26T18:15:46 *** davidjliu has quit IRC (Ping timeout: 265 seconds) 2010-10-26T18:16:10 yeah, I think I got 'em all 2010-10-26T18:18:59 *** JamesMG has quit IRC (Quit: Leaving.) 2010-10-26T18:20:25 *** yasith_ has quit IRC (Ping timeout: 264 seconds) 2010-10-26T18:24:32 *** brx has joined #aichallenge 2010-10-26T18:24:38 *** brx has left #aichallenge 2010-10-26T18:25:15 *** brx has joined #aichallenge 2010-10-26T18:25:16 *** Bobng has quit IRC (Ping timeout: 255 seconds) 2010-10-26T18:29:08 *** fairy44 has joined #aichallenge 2010-10-26T18:32:57 *** yasith_ has joined #aichallenge 2010-10-26T18:35:07 *** Mathnerd314 has quit IRC (Ping timeout: 245 seconds) 2010-10-26T18:37:08 *** Apophis has joined #aichallenge 2010-10-26T18:38:22 *** yasith_ has quit IRC (Ping timeout: 276 seconds) 2010-10-26T18:50:26 *** Palmik has quit IRC (Remote host closed the connection) 2010-10-26T18:52:09 *** narnach has quit IRC (Quit: Leaving.) 2010-10-26T18:52:20 *** yasith has joined #aichallenge 2010-10-26T18:55:55 *** fawek has quit IRC (Ping timeout: 240 seconds) 2010-10-26T18:59:08 thinking about throwing in a quick fix to the different time formats displayed on the profile page 2010-10-26T18:59:29 are people going to have a problem if they are both changes to 24H time? 2010-10-26T18:59:56 *** davidjliu has joined #aichallenge 2010-10-26T19:09:56 *** phreeza has quit IRC (Quit: kthxbai) 2010-10-26T19:11:55 *** amstan has quit IRC (Ping timeout: 255 seconds) 2010-10-26T19:12:02 Top 10 players: rsergio(3713), dmj111(3634), GreenTea(3572), protocolocon(3567), Simplex(3567), felixcoto(3546), Hazard(3489), george(3489), sin_sun(3485), rs5147(3475) 2010-10-26T19:14:23 *** genericbob has quit IRC (Read error: Operation timed out) 2010-10-26T19:14:27 *** genericbob has joined #aichallenge 2010-10-26T19:16:30 *** n9thbit has left #aichallenge 2010-10-26T19:19:57 *** amstan has joined #aichallenge 2010-10-26T19:25:49 *** amstan has quit IRC (Ping timeout: 264 seconds) 2010-10-26T19:40:45 * janzert awaits the forum outcry at having 24 hour time on games 2010-10-26T19:42:00 24 hour time on games? that's insane! 2010-10-26T19:42:14 i thought it was one-second per turn... 2010-10-26T19:42:16 ;) 2010-10-26T19:42:23 :P 2010-10-26T19:42:25 *** DanielVF has joined #aichallenge 2010-10-26T19:43:48 *** amstan has joined #aichallenge 2010-10-26T19:43:48 *** ChanServ sets mode: +o amstan 2010-10-26T19:51:08 *** McLeopold has joined #aichallenge 2010-10-26T19:53:16 Visual Studio 2010 Premium include profiler? 2010-10-26T20:01:57 *** Mathnerd314 has joined #aichallenge 2010-10-26T20:06:23 *** amstan has quit IRC (Ping timeout: 265 seconds) 2010-10-26T20:07:35 Ultimate include all 2010-10-26T20:08:27 *** hornairs has quit IRC (Ping timeout: 245 seconds) 2010-10-26T20:08:33 looks like premium includes it but professional does not 2010-10-26T20:08:35 http://www.microsoft.com/visualstudio/en-us/products#compareTable 2010-10-26T20:09:23 yes 2010-10-26T20:09:28 jast read it 2010-10-26T20:11:41 perfest debuger, like in russian java 1c.ru for buisness 8x versions 8-) 2010-10-26T20:11:54 perfect 2010-10-26T20:12:02 Top 10 players: rsergio(3759), dmj111(3678), protocolocon(3655), Simplex(3590), Slin-.-(3578), felixcoto(3578), jambachili(3528), george(3527), Hazard(3516), sin_sun(3505) 2010-10-26T20:16:03 *** virdo has joined #aichallenge 2010-10-26T20:27:22 *** rebelxt has joined #aichallenge 2010-10-26T20:30:55 *** amstan has joined #aichallenge 2010-10-26T20:30:55 *** ChanServ sets mode: +o amstan 2010-10-26T20:41:16 *** eburnette has joined #aichallenge 2010-10-26T20:59:25 I would like to see a visualizer for oprofile on Linux, so the call-graph can be expanded/collapsed and I can click on a symbol name and get the source code for it. The console output is crazy to go through. 2010-10-26T21:11:34 *** Mathnerd314 has quit IRC (Quit: ChatZilla 0.9.86-rdmsoft [XULRunner 1.9.2.11/20101012113537]) 2010-10-26T21:12:03 Top 10 players: rsergio(3766), dmj111(3689), protocolocon(3663), felixcoto(3600), Simplex(3587), Slin-.-(3586), Hazard(3557), jambachili(3544), george(3537), Xrillo(3517) 2010-10-26T21:16:39 my stupid bot refuses to make more than 1 move on the first turn 2010-10-26T21:16:47 it works well sometimes and ruins me other times 2010-10-26T21:17:51 antimatroid1: yeah it can make you lose on some maps. have you seen george's bot? he normally only makes one move on the first turn, and still manages to win most games. interesting 2010-10-26T21:19:29 my bot does a decent job at winning off one move 2010-10-26T21:19:31 but i need to fix it 2010-10-26T21:22:25 who the hell is photocolocon 2010-10-26T21:22:57 plenty of new people showing up ;) 2010-10-26T21:23:25 that's good, hopefully it's at least competitive for the top 50 this time 2010-10-26T21:23:32 i still want top 10 damnit! 2010-10-26T21:23:49 i feel like i got ripped off last time 2010-10-26T21:24:03 top 10 is going to be hard this time, but possible with lots of work 2010-10-26T21:24:15 tron - what happened? 2010-10-26T21:31:34 timed out on one game 2010-10-26T21:31:41 stupidest disqualification rule ever 2010-10-26T21:31:49 we were sitting around 10th 2010-10-26T21:31:53 i was expecting top 5ish 2010-10-26T21:32:03 before the final tournament started anyway 2010-10-26T21:32:53 you don't get disqualified for that this time thankfully, i didn't even realise that was the rule until it happened 2010-10-26T21:34:04 *** lavalamp_ has joined #aichallenge 2010-10-26T21:35:28 *** rebelxt has quit IRC (Ping timeout: 250 seconds) 2010-10-26T21:35:28 bhasker, I got lucky here: http://72.44.46.68/canvas?game_id=721533 2010-10-26T21:38:14 *** Naktibalda has quit IRC (Remote host closed the connection) 2010-10-26T21:38:40 *** rebelxt has joined #aichallenge 2010-10-26T21:54:21 *** rebelxt has left #aichallenge 2010-10-26T21:55:48 looks like my heuristic completely failed on that map 2010-10-26T21:56:07 don't know why my bot completely ignored the 3 middle planets 2010-10-26T21:56:54 oh wait i think i know why 2010-10-26T21:59:27 *** d3c1ph3r has joined #aichallenge 2010-10-26T21:59:48 quick question when my bot makes no moves it seems to crash near the end game state 2010-10-26T22:00:40 i have finishturn() at the end of the method 2010-10-26T22:06:28 without seeing the code, it may be tough to help... maybe somewhere you still try to forcefully send a fleet out against enemy that has no planets? 2010-10-26T22:10:54 enemy always has enemies want me to pastebin it for you and you can try to debug it if you dont mind? 2010-10-26T22:11:01 enemy always has planets* 2010-10-26T22:11:03 *** eburnette has quit IRC (Quit: Page closed) 2010-10-26T22:12:05 Top 10 players: rsergio(3750), dmj111(3695), protocolocon(3690), felixcoto(3617), Slin-.-(3601), Hazard(3588), jambachili(3588), Xrillo(3553), Simplex(3549), george(3548) 2010-10-26T22:14:19 bhasker: you're malzano right? 2010-10-26T22:14:22 if so i got you :P 2010-10-26T22:14:38 wait, timeouts don't count, nevermind 2010-10-26T22:14:53 *** fluffluff1 has joined #aichallenge 2010-10-26T22:15:28 *** fluffluff has quit IRC (Ping timeout: 276 seconds) 2010-10-26T22:16:13 http://pastebin.com/xf57N6pR can anybody tell me why this crashes at the end game area, i dont send any ships to planets that have fleets heading to it, my nodes use the planets that are closest to the enemy but when i actually make the attack i do it in order of the array i havent changed it yet im still debugging the tree 2010-10-26T22:18:17 *** Mathnerd314 has joined #aichallenge 2010-10-26T22:21:00 d3c1ph3r: are you logging to a file? 2010-10-26T22:21:38 do you send 2 "go"? 2010-10-26T22:23:49 d3c1ph3r: My home-brewed server tells me: "pw.player2.order: Order with 82 of 78 ships issued!" 2010-10-26T22:25:39 You expected more ships to be on your home planet (+5 = 82) than there were (78) 2010-10-26T22:25:58 *** fluffluff has joined #aichallenge 2010-10-26T22:26:24 *** TheMacca has joined #aichallenge 2010-10-26T22:28:28 *** fluffluff1 has quit IRC (Ping timeout: 276 seconds) 2010-10-26T22:32:45 *** fluffluff has quit IRC (Read error: Connection reset by peer) 2010-10-26T22:32:54 *** Cyndre has quit IRC (Quit: Leaving) 2010-10-26T22:33:03 sorry i was afk for two seconds let me catch up 2010-10-26T22:33:28 why woudl this happen 2010-10-26T22:34:01 im sending more ships my pw.planetwars should prevent that shouldn't send more ships hten source planet 2010-10-26T22:34:13 even thou i dont want that error happenign 2010-10-26T22:35:26 hehe yea anti malazan is me 2010-10-26T22:35:48 hmm i am timing out since i am running two versions on my crappy dual core laptop 2010-10-26T22:35:50 while working on it 2010-10-26T22:36:23 *** fluffluff has joined #aichallenge 2010-10-26T22:38:22 decipher are you dividing by something, probably that quantity goes to zero at end game 2010-10-26T22:38:43 oh nm doh 2010-10-26T22:38:46 ignore that comment 2010-10-26T22:38:47 beware of +0 planets 2010-10-26T22:38:50 yeah, i don't run more than one version on tcp at once 2010-10-26T22:39:16 I feel like I have to run a control bot since the ratings change depending on who's running 2010-10-26T22:39:20 *** seisatsu has joined #aichallenge 2010-10-26T22:39:46 heh, I run three versions :) 2010-10-26T22:40:13 yea i should use my desktop 2010-10-26T22:40:13 rather, three copies of the same version 2010-10-26T22:40:20 let me set it up 2010-10-26T22:40:20 its a quad core 2010-10-26T22:40:27 yeah, I'm running four versions right now 2010-10-26T22:40:42 plus whatever local testing I'm doing 2010-10-26T22:41:41 are you using a tree? 2010-10-26T22:41:53 i figure it's better to let it do its thing 2010-10-26T22:42:10 * dmj111 cats /proc/cpuinfo and whimpers 2010-10-26T22:42:30 Hi all, does anyone use delt0r's BotManager? 2010-10-26T22:42:52 dmj are you not running your production bot on tcp? 2010-10-26T22:43:24 d3c1ph3r: If you find something let me know! I can't guarante my server code is 100% correct, but it should give you a hint as what the cause may be. 2010-10-26T22:43:43 *** bhasker has quit IRC (Read error: Connection reset by peer) 2010-10-26T22:45:11 I'm trying to figure out where it saves the replays ... 2010-10-26T22:45:53 i had save to file in my original code but i cut it out in the paste bin because it looked sloppy and even with the out.flush it was still buggy wouldnt order them right ... 2010-10-26T22:46:39 damn it I have an annoying bug 2010-10-26T22:46:45 my program is a tree which recurses till no more moves and no more ships for all possible moves well almost all possible moves it doesn't look at different options on attacking a planet just using the closet planet as the first ship 2010-10-26T22:46:59 TheMacca: It is Java...probably in the user's home directory. Under Windows it is in C:\Documents and Settings\\Application Data or something like that 2010-10-26T22:47:01 *** jaspervdj has quit IRC (Quit: NEVER GONNA GIVE YOU UP NEVER GONNA LET YOU DOWN) 2010-10-26T22:47:20 contestbot later tell bhasker that was my production bot. Mine have always done better official than on tcp. I think the smaller number of official games make mine look better there. 2010-10-26T22:47:20 dmj111: Ready to serve. 2010-10-26T22:49:05 my bot wont even run on the main site atm 2010-10-26T22:49:14 i need to go through and work out what's going on there sometime 2010-10-26T22:49:19 it just crashes 2010-10-26T22:49:32 so was there any advice i saw something on +0 planets 2010-10-26T22:49:32 d3c1ph3r: It seems that your program sent a fleet of 21 ships on the first turn, then a fleet of 21 in the third turn, and a fleet with 5 too many ships in the fourth turn. Does it always send planet.ships+planet.growth? 2010-10-26T22:49:47 i'm pretty sure it's started getting better again, rather than worse 2010-10-26T22:51:16 *** hornairs has joined #aichallenge 2010-10-26T22:51:32 Frontier: Thank you, I found them. 2010-10-26T22:52:14 hm, that latest change I made seems to be working out: http://72.44.46.68/getplayer?player=lavalamp.gromble.gen.7 2010-10-26T22:53:34 *** bhasker has joined #aichallenge 2010-10-26T22:55:55 *** yasith has quit IRC (Ping timeout: 240 seconds) 2010-10-26T23:01:22 d3c1ph3r: I really get strange results from your bot! I see 2 lines saying "go" when you issue an order. Also you send the same amount of fleets several times in a row until you get disqualified because there aren't enough ships remaining to satisfy the order. 2010-10-26T23:10:03 *** McLeopold has joined #aichallenge 2010-10-26T23:12:04 Top 10 players: rsergio(3797), protocolocon(3734), dmj111(3731), felixcoto(3656), Slin-.-(3637), jambachili(3624), Hazard(3618), Simplex(3581), george(3579), sin_sun(3571) 2010-10-26T23:18:58 anyone know who Simplex is? 2010-10-26T23:19:05 *** yasith has joined #aichallenge 2010-10-26T23:21:38 *** perestrelka has joined #aichallenge 2010-10-26T23:32:34 http://72.44.46.68/canvas?game_id=723114 2010-10-26T23:32:46 Interesting starting strategy by retybok. 2010-10-26T23:33:30 *** hellman has joined #aichallenge 2010-10-26T23:33:47 i want to play mcleopold 2010-10-26T23:34:00 that always reminds me of charlie and the unicorn 2010-10-26T23:34:46 there we go 2010-10-26T23:34:48 all i had to do was ask 2010-10-26T23:34:49 *** tapwater has quit IRC (Quit: tapwater) 2010-10-26T23:36:50 That charlie?: http://www.youtube.com/watch?v=Q5im0Ssyyus 2010-10-26T23:37:07 pounded me 2010-10-26T23:37:11 it is 2010-10-26T23:37:38 jesus christ, 52 million views 2010-10-26T23:38:30 leoplaurodon 2010-10-26T23:38:32 that's what does it 2010-10-26T23:40:57 why is retybok sending 68 ships to that neutral 10+5??? 2010-10-26T23:41:42 *** jukkerknott has joined #aichallenge 2010-10-26T23:41:44 antimatroid1: which bot is yours? 2010-10-26T23:42:50 retard 2010-10-26T23:42:57 cause atm it's fairly retarded 2010-10-26T23:43:09 i'm ironing things out 2010-10-26T23:43:23 yeah, it looks like I stole your planet up top 2010-10-26T23:43:39 there are a few major issues still that i'm currently working on 2010-10-26T23:43:55 like it's making redistribution moves based off the wrong gamestate 2010-10-26T23:44:22 it also looks like you favor expansion over defense 2010-10-26T23:44:33 it shouldn't :| 2010-10-26T23:44:38 i do defense before anything 2010-10-26T23:45:20 turn 13, you go for the top 67 neutral, but you are about to lose a planet 2010-10-26T23:45:32 you did get it back, but you gave up some growth. 2010-10-26T23:46:03 it didn't know it was going to lose it 2010-10-26T23:46:19 really? the fleets are already in flight. 2010-10-26T23:46:38 no they're not 2010-10-26T23:46:44 they go through the planet with 4 growth 2010-10-26T23:46:47 4 ships** 2010-10-26T23:47:03 *** tapwater has joined #aichallenge 2010-10-26T23:47:20 atm it's just worrying about the ships you can send this turn, that's bad but otherwise he gets crippled 2010-10-26T23:47:30 at turn 14 you can see 2 fleets, a 35 and a 34 2010-10-26T23:48:08 nope? 2010-10-26T23:48:09 I'm referring to my fleets, the 17, 4 and 1 2010-10-26T23:48:16 *** yasith has quit IRC (Ping timeout: 276 seconds) 2010-10-26T23:48:27 they go straight for the neutral 10 up top after you take it. 2010-10-26T23:48:32 yeah 2010-10-26T23:48:35 * raedwulf and zombies are related 2010-10-26T23:48:48 I think you should have been able to foresee that at the time you launched for the 67 neutral 2010-10-26T23:48:50 i do future stuff for everything but you potentially taking my planet 2010-10-26T23:49:00 i am not considering the other planet atm 2010-10-26T23:49:05 i removed that as i had too many bugs 2010-10-26T23:49:16 i'm ironing things out before reintroducing the other player making moves 2010-10-26T23:49:24 other player*** gah sorry 2010-10-26T23:50:08 but, the fleets are already sent. shouldn't you be able to run the battle calcs and see that it would be lost? It's not a prediction of what I may do. 2010-10-26T23:50:48 The funny thing is, after you take that 67 neutral, you see the 12 fleet I send and protect with a 1 fleet. 2010-10-26T23:51:08 oh i see what you mean 2010-10-26T23:51:38 the problem is the expectedclosestfriend is less than or equal to closest friend atm 2010-10-26T23:51:43 i need to fix that 2010-10-26T23:52:16 ah, distance calculations getting in the way? 2010-10-26T23:52:56 if it didn't count that as a friend it never would have considered expanding there, so yeah 2010-10-26T23:52:57 now taking bets on whether bocsimacko will when 25 in a row on the official server... heh 2010-10-26T23:53:21 what's the payoff? 2010-10-26T23:53:35 that would be a terrible bet to make 2010-10-26T23:53:58 heads for 25 in a row! 2010-10-26T23:54:00 !coin 2010-10-26T23:54:11 damn... what was the command again? 2010-10-26T23:54:24 too bad the octopus died, we could've asked him 2010-10-26T23:54:26 it's taken him 16 wins to get to around 15th, so he only needs nine more now 2010-10-26T23:54:32 lol, heh. poor Paul 2010-10-26T23:55:05 @coin 2010-10-26T23:55:05 janzert: heads 2010-10-26T23:55:44 only need eight more 2010-10-26T23:55:45 @coin 2010-10-26T23:55:45 davidjliu: heads 2010-10-26T23:55:49 @coin 2010-10-26T23:55:49 davidjliu: heads 2010-10-26T23:55:51 *** jesionaj has quit IRC (Ping timeout: 252 seconds) 2010-10-26T23:55:55 @coin 2010-10-26T23:55:55 davidjliu: tails 2010-10-26T23:55:57 aw 2010-10-26T23:55:57 @coin 2010-10-26T23:55:57 McLeopold: tails 2010-10-26T23:56:49 @coin 2010-10-26T23:56:49 raedwulf: heads 2010-10-26T23:57:14 I don't think it will happen. 2010-10-26T23:57:24 I bet on 20 wins. 2010-10-26T23:57:46 @dice 1d25 2010-10-26T23:57:46 Frontier: 3 2010-10-26T23:58:17 accoun has a lot of trouble on the main site 2010-10-26T23:58:42 he's bot is geared for winning against other good bots 2010-10-26T23:59:08 in that it waits too much? 2010-10-26T23:59:16 when someone takes a wild stab at the center, he doesn't do so well. 2010-10-26T23:59:35 it depends how much i have spare whether i win or lose that 2010-10-26T23:59:39 He builds very fast, but doesn't pay attention to position. 2010-10-26T23:59:47 as much 2010-10-26T23:59:54 i took out anything about position too 2010-10-26T23:59:59 It's still a very good bot.