TCEC LIVE - Division I, Season 1 - All the top engines!

As in chess tournaments and matches...
Martin Thoresen
Posts: 386
Joined: Thu Jun 10, 2010 5:27 am

Re: TCEC LIVE - Division I, Season 1 - All the top engines!

Post by Martin Thoresen » Fri Jan 14, 2011 5:18 am

MichaelIsGreat wrote:Hello to All,


This Division 1 tournament is excellent. However, reviewing the rules, I believe a few of them might be better changed a little bit, not now but in the future.

At the page where the rules are stated, at http://www.tcec-chess.org/info.php , it says:
1) "However, a game can also be drawn at move 30 or later if the eval from both engines are within +0.05 to -0.05 pawns for the last 5 moves, or 10 plies. If there is a pawn advance, or a capture by any kind, this special draw rule will reset and start over."
and
2) "It will adjudicate as lost or won if both engines have an eval of at least 5.25 or -5.25 pawns for 3 consecutive moves - this rule is in effect as soon as the game starts."
---The rule 2) above might be better changed with an evaluation to 10.00 or -10.00. Just to be on the safe side, as the chess engines evaluation is far from being accurate.
---The rule 1) above might be better changed to "for the last 10 moves" instead of "for the last 5 moves", just to be on the safer side I would guess.
Just suggestions to assess for the future, as a few positions might necessitate these suggestions.

In any case, the current tournament is nothing less than excellent.

Thanks for the great guy (Martin Thoresen) who set up this initiative at his great web site, the TCEC (Thoresen Chess Engines Competition), at http://www.tcec-chess.org/ .


Kind Regards to All
MichaelIsGreat
Michael,

Thanks for your input!

In regards to #2, I think 10 moves (20 plies) is way too long. I had to set the limit somewhere, and I think most people agree that 5 moves (10 plies) are just fine.
If it's any comfort, it was 4 moves (8 plies) in the "old TCEC" system. :)

About #1, yes this could be done but not without Matthias fixing the current issue with ChessGUI in that mate scores won't be adjudicated.
If this will be fixed, I will increase this threshold to perhaps +7 or +8, but not as high as +10 in any case.

Best,
Martin
TCEC - Thoresen Chess Engines Competition
http://tcec.chessdom.com

User avatar
Gargamel
Posts: 7
Joined: Fri Aug 13, 2010 6:51 pm
Location: France

Re: TCEC LIVE - Division I, Season 1 - All the top engines!

Post by Gargamel » Fri Jan 14, 2011 11:34 am

Thank you very much Martin for this tournament. At least, it's show clearly to the chess Community (and the other) that Rybka is not the only engine to be very strong.

In this game, Critter takes an advantage in the opening. I think that Ivanhoe spend much time to move after 22.Na5

1r2kbr1/pb1p1n1p/n2Pq1p1/NpP1pp2/7P/P3P3/1B2NPP1/2RQKB1R b K - 7 22
What do you think about this ?

Martin Thoresen
Posts: 386
Joined: Thu Jun 10, 2010 5:27 am

Re: TCEC LIVE - Division I, Season 1 - All the top engines!

Post by Martin Thoresen » Sat Jan 15, 2011 5:26 pm

Thanks Gargamel!

After a maintenance of almost 24 hours the tournament has now been resumed again.
TCEC - Thoresen Chess Engines Competition
http://tcec.chessdom.com

MichaelIsGreat
Posts: 44
Joined: Thu Dec 30, 2010 3:52 pm

For Elite Match, could you allow the games to finish fully?

Post by MichaelIsGreat » Sun Jan 16, 2011 11:20 pm

Hello Martin, the organizer of the TCEC tournaments,


For the Elite Match between the top two winners of the Division 1 tournament, it would be good if you could let the games finish fully? I do not know if it is something hard to implement or not. Just a suggestion.

Hoping to see Houdini 1.5 and Rybka 4 in finale for this Elite match.

Thanks for your excellent chess engines tournaments.


Regards
MichaelIsGreat

BB+
Posts: 1484
Joined: Thu Jun 10, 2010 4:26 am

Re: TCEC LIVE - Division I, Season 1 - All the top engines!

Post by BB+ » Tue Jan 18, 2011 1:10 pm

Critter finding 20. e6! against Naum in Round 11 (not sure why Naum didn't play Qc7 rather than Rc8, though Nf5 iin response s no picnic), now it looks to be a 3-way battle for the 2nd spot.2rqr1k1/p2pbp1p/1pp3pB/n3P3/2bN2Q1/2P3P1/P4PBP/3RR1K1 w - - 4 20Perhaps not the deepest tactic, but aesthetically pleasing nonetheless.

BB+
Posts: 1484
Joined: Thu Jun 10, 2010 4:26 am

Re: TCEC LIVE - Division I, Season 1 - All the top engines!

Post by BB+ » Tue Jan 18, 2011 5:18 pm

---The rule 2) above might be better changed with an evaluation to 10.00 or -10.00. Just to be on the safe side, as the chess engines evaluation is far from being accurate.
The recent HIARCS-Rybka game had some +4 scores for Black which I think were possible fortress draws. Not sure I ever saw one with +5. I did more analysis in a different thread (when it looked to heading toward a rook endgame, again with some possible funky draws). See http://www.open-chess.org/viewtopic.php?f=4&t=988#p8880 and the following post. Even the game-line that ended in Q+P vs R+P could conceivably end in a draw, though with 5-piece TBs it should be won (for that matter, Q vs R is not always a sure win w/o TBs for some top engines).

It is certainly true that not every "+5" score is alike. Maybe somehow measure material left also?

BB+
Posts: 1484
Joined: Thu Jun 10, 2010 4:26 am

Re: TCEC LIVE - Division I, Season 1 - All the top engines!

Post by BB+ » Wed Jan 19, 2011 9:28 am

Shredder beating Rybka might be another case of the "increment" chess becoming a bit hairy. Or maybe the Shredderbases just paid off. :) In any event, I think I'm going to switch to posting at the TCEC forum about specifics, and only cross-post here for things of more general interest.

BB+
Posts: 1484
Joined: Thu Jun 10, 2010 4:26 am

Re: TCEC LIVE - Division I, Season 1 - All the top engines!

Post by BB+ » Wed Jan 19, 2011 10:42 am

1r1n3R/2p5/2P3k1/8/5PPK/p3B3/8/8 b - - 0 96Rybka played 96... Nxc6 and lost. I'm fairly certain 96... a2 draws, maybe 96... Kf7 also (engines don't know when RBP vs RN is drawn, but my guess is that White has good chances here). Mainline is 96... a2 97. Bd4 Rb4 98. f5+ Kf7 99. Rxd8 Rxd4 100. Ra8 Rd2 101. Kg5 Rc2 102. Ra7 Re2 103. Kf4 Kf6, and White can make no progress. After 96... Nxc6 it is lost I think, though there are various "null-move" questions with the Knight, and it is thus non-trivial. [And as pointed out by others, one main-line has a Knight under-promotion for White].

Jeremy Bernstein
Site Admin
Posts: 1226
Joined: Wed Jun 09, 2010 7:49 am
Real Name: Jeremy Bernstein
Location: Berlin, Germany
Contact:

Re: TCEC LIVE - Division I, Season 1 - All the top engines!

Post by Jeremy Bernstein » Wed Jan 19, 2011 12:00 pm

BB+ wrote:1r1n3R/2p5/2P3k1/8/5PPK/p3B3/8/8 b - - 0 96Rybka played 96... Nxc6 and lost. I'm fairly certain 96... a2 draws, maybe 96... Kf7 also (engines don't know when RBP vs RN is drawn, but my guess is that White has good chances here). Mainline is 96... a2 97. Bd4 Rb4 98. f5+ Kf7 99. Rxd8 Rxd4 100. Ra8 Rd2 101. Kg5 Rc2 102. Ra7 Re2 103. Kf4 Kf6, and White can make no progress. After 96... Nxc6 it is lost I think, though there are various "null-move" questions with the Knight, and it is thus non-trivial. [And as pointed out by others, one main-line has a Knight under-promotion for White].
Rybka 4 appears to analyze ...a2 as second choice for a short while, but then discards it entirely (...Kf7 takes its place). Houdini would have played ...a2. Naum 4.2 agrees with Houdini. HIARCS can't decide between ...Kf7 and ...Nxc6 for a while, although it comes around at around depth 22, discarding Nxc6 and evaluating both ...Kf7 and ...a2 at around 1.25. Rybka 3 reaches an identical conclusion as her big sister. Stockfish agrees with Rybka in this matter, as well.

BB+
Posts: 1484
Joined: Thu Jun 10, 2010 4:26 am

Re: TCEC LIVE - Division I, Season 1 - All the top engines!

Post by BB+ » Wed Jan 19, 2011 12:26 pm

Rybka 4 appears to analyze ...a2 as second choice for a short while, but then discards it entirely (...Kf7 takes its place). Houdini would have played ...a2. Naum 4.2 agrees with Houdini. HIARCS can't decide between ...Kf7 and ...Nxc6 for a while, although it comes around at around depth 22, discarding Nxc6 and evaluating both ...Kf7 and ...a2 at around 1.25. Rybka 3 reaches an identical conclusion as her big sister. Stockfish agrees with Rybka in this matter, as well.
I might point out that I analysed this with both IvanHoe and ComStock with all the TripleBases, and in each case backed up the score after 96...a2 to 0.00 (after playing forward to a draw), and the same with a +6 score after 96...Nxc6. Whether or not this is 100% reliable (due to transpositions, nullmove, and whatnot) is another issue. [With 96...Kf7, the scores hovered in the +2 range, though many of these were from RBP vs RN, which might well be drawn -- if I followed lines were pawns were not exchanged, it was still unclear, and I lost interest after knowing the results from the other two moves]. From the TCEC data, Rybka was playing at about 5 seconds per move from moves 96-100 (after using 46s at move 95 -- both had 11m left until Shredder used 8 minutes on 101. g5), and then suddenly took a 13-second think at move 102 and realised White was +3.

It really is amazing to me how many of these get decided in the later stages, though I think some of the Freestyle guys had made similar comments (that an error in the middlegame might change the score by -0.25 or -0.5, but in the endgame would simply lose).

Post Reply