Stockfish dev - Houdini 6 +45 ELO

Discussion of computer chess matches and engine tournaments.

Moderators: hgm, Rebel, chrisw

Jouni
Posts: 3281
Joined: Wed Mar 08, 2006 8:15 pm

Stockfish dev - Houdini 6 +45 ELO

Post by Jouni »

4 cores, 3m + 1s, HERT.PGN:

Code: Select all

                                    
1   Stockfish 031117 64 BMI2   +45  +21/=71/-8 56.50%   56.5/100
2   Houdini 6 x64-pext         -45  +8/=71/-21 43.50%   43.5/100

Shortest game here
[pgn]
[Event "3m+1s"]
[Site "?"]
[Date "2017.11.20"]
[Round "7"]
[White "Stockfish 031117 64 BMI2"]
[Black "Houdini 6 x64-pext"]
[Result "1-0"]
[ECO "C14"]
[Annotator "0.38;0.20"]
[PlyCount "43"]
[TimeControl "180+1"]

1. e4 e6 2. d4 d5 3. Nc3 Nf6 4. Bg5
Be7 5. e5 Nfd7 6. h4 Bxg5 7. hxg5 Qxg5 {Both last book move} 8. Nh3 {0.38/24 7}
Qh6 {0.20/24 10 (Qe7)} 9. Nb5 {0.48/21 2 (g3)} Na6 {0.28/21 5} 10. g3 {0.51/24
3} f6 {0.23/21 9 (Nb6)} 11. Qe2 {0.85/21 1 (f4)} O-O {0.55/21 10 (fxe5)} 12.
Bg2 {1.35/21 3} Qg6 {0.67/21 7} 13. Nf4 {1.41/24 6} Qe8 {1.25/20 9 (Qf7)} 14.
O-O-O {1.68/22 3} fxe5 {1.86/23 17 (Nb6)} 15. dxe5 {2.10/22 3} Nb6 {1.49/19 3
(Ndc5)} 16. Nd4 {2.48/23 4 (Rh4)} c5 {2.43/21 15} 17. Nb5 {2.87/23 3} Nc4 {2.
93/21 8} 18. Nc3 {3.09/25 10} Nc7 {2.81/21 2} 19. Ne4 {3.22/26 9 (Ncxd5)} dxe4
{2.68/19 4} 20. Bxe4 {7.15/26 8 (Qxc4)} h6 {6.52/23 10} 21. Rxh6 {7.43/27 3}
Rxf4 {8.25/25 10} 22. gxf4 {7.81/26 2 (Bg6)} 1-0

[/pgn]
Jouni
JJJ
Posts: 1346
Joined: Sat Apr 19, 2014 1:47 pm

Re: Stockfish dev - Houdini 6 +45 ELO

Post by JJJ »

100 games at 3 min + 1 means nothing, certainly not +45 elo.
tpoppins
Posts: 919
Joined: Tue Nov 24, 2015 9:11 pm
Location: upstate

Re: Stockfish dev - Houdini 6 +45 ELO

Post by tpoppins »

An interesting game. The Alekhine-Chatard Attack is my favorite line against Classical French.

Houdini's first move out of the book (8...Qh6) was already sub-optimal. On my end H6.02 finds 8...Qe7 at depth 26, switches to 8...Qh4 at d=27 and goes back to 8...Qe7 @d=33. So at LTC Houdini would find a better move than 8...Qh6.

Although SF may hold an edge over Houdini in head-to-head matches at any time control, on rating lists (where the games against the top engines are only a part of the picture) it's liable to perform somewhat worse as Houdini (and Komodo) are a lot more ruthless with second-tiers than Drawfish.
Leo
Posts: 1080
Joined: Fri Sep 16, 2016 6:55 pm
Location: USA/Minnesota
Full name: Leo Anger

Re: Stockfish dev - Houdini 6 +45 ELO

Post by Leo »

JJJ wrote:100 games at 3 min + 1 means nothing, certainly not +45 elo.
I agree. Very little meaning.
Advanced Micro Devices fan.
User avatar
pohl4711
Posts: 2433
Joined: Sat Sep 03, 2011 7:25 am
Location: Berlin, Germany
Full name: Stefan Pohl

Re: Stockfish dev - Houdini 6 +45 ELO

Post by pohl4711 »

Leo wrote:
JJJ wrote:100 games at 3 min + 1 means nothing, certainly not +45 elo.
I agree. Very little meaning.

I just finished the testrun of Stockfish 171111 with 5000 games, 3'+1“ (but only singlecore)

Stockfish 171111 bmi2 : 3439 5000 (+2262,=2479,-259), 70.0 %

Fire 6.1 popc : 1000 (+581,=402,- 17), 78.2 %
Houdini 6 pext : 1000 (+205,=690,-105), 55.0 %
Fizbo 1.9 bmi2 : 1000 (+616,=357,- 27), 79.5 %
Shredder 13 x64 : 1000 (+602,=380,- 18), 79.2 %
Komodo 11.2.2 x64 : 1000 (+258,=650,- 92), 58.3 %

http://www.sp-cc.de