Xaurus: CB, can you please add some info about the tournament and a link to my website please
Xaurus: like you did with the previous match
snoopy: Awesome a new cpu tourny
Toliman4: are the openings like with R4
Xaurus: no, brand new openings
Xaurus: all 16 ply (8 moves) deep and based on games by Magnus Carlsen
Xaurus: You can download the opening PGN from the official site: http://home.halden.net/mordor/match/broadcast.php
ChessBomb: @Xaurus: Thanks for the interesting matches, feel free to give information about them here.
Xaurus: The information stands out much better with orange font :)
ChessBomb: Time Control: 120 Minutes For The First 40 Moves - 60 Minutes For The Next 20 Moves 15 Minutes For The Rest - 30 Seconds Added Per Move From Move 61
ChessBomb: Openings: Magnus_Carlsen_2005-2010.pgn - Thanks To Dennis Monokroussos For Guidance Match Length: 32 Games - Engines Play Black & White Of Each Opening CPU: Intel Core i7 920 @ 4050 MHz - OS: Windows 7 Professional x64 Executables: Stockfish 1.8 - Houdi
ChessBomb: Match operator: Martin Thoresen (Xaurus here on ChessBomb)
ChessBomb: Official match site: http://home.halden.net/mordor/match/broadcast.php
Xaurus: Thanks CB, that's appreciated :)
ChessBomb: A little information about the participants:
ChessBomb: Stockfish 1.8 is currently the strongest free, non-controversional, and also open-sourced chess engine
ChessBomb: Houdini is one of the currently strongest free, but controversional chess engines
dennis1382: what's the controverse with Houdini?
ChessBomb: There are allegations that Houdini contains code from Ippolit, which in turn is accused in having stolen code from Rybka 3
ChessBomb: Neither of these has been proven, so we've agreed to relay this match
dennis1382: ok...
schumixtreme: is this the two strongest free engine as of now?
Xaurus: most likely yes
snoopy: stockfish is the bst cause the developers are cool also thakd chessbomb and xaurus for your good works :)
snoopy: best and thanks
snoopy: firebird is supposed to be strong
snoopy: it is basically ippolito which is accused of having stolen rybka code
Xaurus: thanks for the support, snoopy
Drumworks: How can you people say SF is the best when it got hammered by Rybka?
fleiseren: What about firebird?
Xaurus: SF is free, Rybka is commercial
fleiseren: I love Rybka :P
EuropeanEmperor: stockfish is still too good for houdini
DeepRybka: I love Stockfish on my iPhone and Rybka on my PC :)
d_ahura: For Android phones there is DroidFish. A sure sign their market is catching on :)
TrojanHorse: Houdini needs a good nights sleep to improve its performance tomorrow.
LeoCalixto: ### If @Chessbomb read me, i'd like to say this kind of games make the site even more atractive, couse is allways running matches. And if it's possible, make more softwares matchs. Thx ###
Xaurus: Thanks for your support, LeoCalixto. Although I am not involved with Chessbomb I am glad they are relaying my matches.
TrojanHorse: The match is equal. Maybe Houdini is based upon Stockfish software!!
notyetagm: It is. Houdini is based on Ippolit, Sotckfish, and Fruit engines I believe.
paiter: i agree
TrojanHorse: The Ippolit part may be exagerrate. I would say this is a match Stockfish vs Stockfish.
komyt: Xaurus, have you set up any matches with TogaII as one of the contestants?
wevict: Hello Xaurus. TordRomstad has told: To say anything about strength, you need thousands of games with a wide variety of opponents. What for this match if results will mean statistically nothing?
DeepRybka: wevict: Read my reply to Tord
Xaurus: He is right, to determine exact ELO thousands if not tens of thousands of games are needed. Statistically this match means little. It's more for the fun and prestige of it.
DeepRybka: I wasnt there when he replied, so my response was posted much later. I don't think he was still there when I erplied
DeepRybka: I replied in game 9
DeepRybka: Basically, in 1 vs 1 matches u dont necessarily need to play thousands of games from a statistical standpoint
DeepRybka: but for measuring general poerformance
DeepRybka: performance u need thousands of games vs many different opponents
DeepRybka: we know that certain engines do particularly well against a specific engine
DeepRybka: so a 1 vs 1 match can be misleading
DeepRybka: thats why playing many diffeerent opponents is important
DeepRybka: as for this match, I think that if you add all the games so far for all the matches
DeepRybka: you can get a good idea for how strong SF 1.8 is
DeepRybka: although u need more games for more reliable statistical results
DeepRybka: SSDF doesnt publish the results before at least 100 games have been played
DeepRybka: SSDF only conduct matches as tournament level
DeepRybka: "as" => "at"
DeepRybka: ok, gotta go!
wevict: The problem remains. There is no statistical performance, such as CCRL or GECT, for chess engine Houdini. And the given match - entertainment which it will not give the answer how much it is strong.
DeepRybka: wevict: U can actually find hundreds of games played by Houdini if you look at certain forums
DeepRybka: so we do have a good idea of how strong it is
DeepRybka: Houdini is being rated on the IPON list
DeepRybka: however, this list is only based on blitz games
DeepRybka: and weak 1 core cpus
DeepRybka: on this list it's rated behind Rybka 4, but in second place before SF 1.8
DeepRybka: the results do correlate well with the results posted at the forums I mentioned
DeepRybka: that's also why I believed before this match started tyhat SF 1.8 and Houdini 1.03 would be quite evenly matcghed
DeepRybka: even though I think Houdini will win by a small margin
DeepRybka: However, a lot can happen in just 32 games
DeepRybka: so we'll see :)
wevict: DeepRybka: In current match it is no correlation between IPON blitz list and games. In IPON Houdini is stronger than Stockfish.
DeepRybka: wevict: This is no blitz match
DeepRybka: and the match is onlyy 32 games
DeepRybka: like i said, it's not statistcally signifcant
DeepRybka: significant
wevict: DeepRybka: I mean that no correlation here.
DeepRybka: the results I talked about was in relation tot the forum results
DeepRybka: please reread my comment :)
DeepRybka: besides, the IPON list doesnt show that big difference
DeepRybka: and like I said, I think Houdini will win this match
DeepRybka: so it doesnt contradict Houdinis higher rating on IPON
wevict: DeepRybka: What forum make tournaiment tome control? One exampe? I olso know that Houdini is stronger...
DeepRybka: wevict: I'm talking about blitz results
wevict: DeepRybka: Not tome but time, sorry.
DeepRybka: thats why I said this is not a blitz match
wevict: DeepRybka: I don't know forums with tounaiment time control. I want to find any correlation. Do you know any? You has told early: eepRybka: the results do correlate well with the results posted at the forums I mentioned
DeepRybka: I'm talking about blitz games, not tournament level games
DeepRybka: sometimes you can use blitz games as a basis for extrapolating performance at longer time controls
wevict: DeepRybka: thanks, i see
DeepRybka: I've not seen any tournament level results for Houdini 1.03 so far
DeepRybka: apart from this one
DeepRybka: this match
DeepRybka: one of the forums i talked about is www.open-chess.org
DeepRybka: but there are others that I don't have the link to at the moment
TrojanHorse: I believe that calculating ELO from computer vs computer is meaningless because of open source code sharing.
TrojanHorse: Stockfish is basically playing against itself here.
Xaurus: It's not playing against it self. Stop making such comments please.
TrojanHorse: It's no secret that Houdini is using partially Stockfish software. This makes Houdini a partial Stockfish clone.
Xaurus: Please read carefully what the programmer says: Without many ideas from the excellent open source chess engines Ippolit/Robbolito, Stockfish and Crafty (in that order), Houdini would not nearly be as strong as it is now.
TrojanHorse: I could take Stockfish - modify something unimportant an call it FishStock. FishStock would probably draw against StockFish.
TrojanHorse: Xaurus: Are you on the defence since you ask me to stop making such comments. The comments are relevant:-)
Xaurus: Whether they are relevant or not are not for me to decide, but what I do want is you to spread false information.
Xaurus: Hehe, what I _don't_ want you to. Getting late...
TrojanHorse: But the sentence you did not like was not meant to be read together with the ELO sentence - I was therefore not so careful about the exact words.
TrojanHorse: Corr: _was meant_
Xaurus: No worries, I understood what you meant. But that's a whole other discussion. If you really want to get specific about that issue then you can say that Rybka is basically a Fruit.
Aspirador: The only difference between Los Angeles and a yoghurt is that yoghurt contains less fruit -- Rush Limbaugh
Xaurus: :)
TrojanHorse: I am a computer programmer. If I had wasted time mixing up code like Houdini have done and not been able to improve the performance of Stockfish alone I would have been fired!!
bluefish: If you could program an engine as strong as Houdini,SF,Rybka you may get a raise.LOL.
fleiseren: Housini rules!
TrojanHorse: bluefish: It's easy. StockFish is open source. I could take Stockfish and modify something unimportant and call it Wizard.
TrojanHorse: Since StockFish is open source many programs use its code. What we'll se is like Linux versions - Ubuntu, Mandrake, Redhat ...
TrojanHorse: If you compare the performance of Ubuntu, Mandrake and Redhat what you'll see is a fairly equal performance.
yuriymos: Deep Rybka 4 SSE42 x64 vs Houdini 1.01 x64 POPCNT 4_CPU = 53.00%|47.00%
yuriymos: and 50 out of 50 games played - 5 Minutes/Game - Deep Rybka 4 vs Houdini 1.01= 52-48 %%
wevict: Today Houdini 1.03a after 1000 games become first chess blitz engine in IPON list: http://www.inwoba.de/index.html
yuriymos: 2010.05.28 Engine added - Deep Rybka 4 1600 games The number one has done it again and gained remarkable 52 Elo!
yuriymos: same source
EuropeanEmperor: Hi ChessBomb!Are you going to cover World Junior championship set this August?Regards...
Xaurus: CB, are you here?
bishop21: wow, stockfish won the first game..... and the next 22 no more
snoopy: but 23 is win
chess614: What chat is this?
chess614: Is this a non-specific-game chat, re. the SF-Houdini match?
Xaurus: Yes.
bishop21: @snoopy: my god... 21
snoopy: well 28 looks like an sf win
snoopy: with black
CareyHunt: What does the S-B column mean please?
MartinThoresen: See this link: http://en.wikipedia.org/wiki/Neustadtl_score