Top 10 Free Chess Engine Recommendations

Sort:
Avatar of MrEdCollins

Yes, but see my comment in Post #9.

Avatar of Combezz
itsmedaniel a écrit :
MrEdCollins wrote:

Well, despite being down a pawn, the evaluation SHOULD certainly be something less than 1.00, as White already has some compensation for the sacrificed material.

It's White's move, he already has one piece developed and is about to further overtake Black in devevelopment by developing another.  He's two moves away from castling and upon doing so will develop his KR by putting it on the half open f file.

Whether all of this compensation is worth a full pawn or not has been debated for more than two hundred years... that's the King's Gambit for you. 

Other engine evaluations for that same position, some of which you're not going to like, since some are even more optimistic:

Komodo 6:  23 [-0.03]  4.Bc4 Be7 5.O-O Nf6 6.Nc3 Nxe4 7.Nxe4 d5 8.Bxd5 Qxd5 9.d3 O-O 10.Bxf4 Bf5 11.Nc3 Qd8 12.Qe2 Nc6 13.Qf2 b6 14.Rae1 f6 15.Be3 Qd7 16.Qh4 Bg4 17.h3  (158.37)

Bouquet 1.8:  21 [-0.12]  4.Qe2 c5 5.d4 cxd4 6.Bxf4 Nc6 7.Nbd2 Nf6 8.Nb3 Be7 9.O-O-O O-O 10.Nbxd4 Nxd4 11.Rxd4 Qa5 12.Kb1 Be6 13.a3 Rac8 14.Qb5 Qxb5 15.Bxb5 a6 16.Be2  (176.91)

Critter 1.6a:  22 [-0.16]  4.Qe2 g5 5.Qb5+ Nc6 6.Qxg5 Qxg5 7.Nxg5 Bh6 8.h4 f6 9.Nf3 f5 10.Nc3 fxe4 11.Nxe4 Nge7 12.Nf6+ Kf7 13.Ne4 Re8 14.Kd1 Bf5 15.Neg5+ Bxg5 16.Bc4+ d5 17.Nxg5+ Kg7  (71.98)

Houdini 3:  22 [-0.26]  4.Qe2 g5 5.Qb5+ Nc6 6.Qxg5 Qxg5 7.Nxg5 Bh6 8.h4 f6 9.Nf3 f5 10.Nc3 fxe4 11.Nxe4 Nge7 12.c3 Bf5 13.d3 O-O-O 14.Be2 Nd5 15.Nfg5 Rhf8 16.O-O Rde8 17.Bh5 Re7 18.c4  (129.14)

Deep Rypka 4.1:  20 [-0.33]  4.Bc4 h6 5.h4 Nf6 6.Nc3 c6 7.d4 b5 8.Bd3 Nh5 9.Ne2 g5 10.d5 a6 11.hxg5 hxg5 12.a4 b4 13.Qd2 c5 14.c3 a5  (347.10)

Naum 4.2:  19 [-0.34]  4.Bc4 h6 5.Nc3 Nc6 6.h4 Nf6 7.d4 Nh5 8.Nd5 Be6 9.Nxf4 Bxc4 10.Nxh5 Qe7 11.Kf2 O-O-O 12.b3 Ba6 13.Re1  (139.79)

Thought this was interesting http://en.chessbase.com/post/rajlich-busting-the-king-s-gambit-this-time-for-sure

I'm pretty sure this is an april fools

Avatar of EscherehcsE
shobhanlal wrote:

i have had a chess tournament between houdini 1.5a, komodo 5 and stockfish dd (60 moves in 10 mins). there houdini calculates at 4000-4500 kn/s, stockfish at 1200-1400 kn/s and komodo at 1000-1200 kn/s on a 4 core computer.

...which should tell you that nodes/second mean absolutely nothing. Smile

Avatar of Romancia

I agree on number one. I think it is the strongest engine right now, even stronger than houdini 4. http://tcec.chessdom.com/live.php backs me up. 24/7 engine tournament. But I have a little concern about the interface. I use arena and I find it pretty good (though I cannot play chess960 very properly). What about scid vs pc. What do you guys think are the pros and cons of arena and scid vs pc?

Avatar of xakean-ski

I think Arena and Scid vs PC are both good GUI's, but which one is better depends on what you plan to use the GUI for.  Arena is better for engine tounaments for sure, and I also prefer Arena for analysis.  I think Scid vs PC is better for databases, however.  The database functions in Arena can be difficult to use.  Overall, I would reccomend Arena, but that's mainly because Scid vs PC has given me problems, such as freezing up and not working properly, while Arena usually runs smoothly.

Avatar of MSC157

What about Winboard?

As far as it goes for phones, Stockfish is the best I guess.

Avatar of xakean-ski

I never have liked winboard, mostly because of the interface with multiple windows.  Other than the interface, I don't think it's a bad GUI.

Avatar of pfren
shobhanlal wrote:

i have had a chess tournament between houdini 1.5a, komodo 5 and stockfish dd (60 moves in 10 mins). there houdini calculates at 4000-4500 kn/s, stockfish at 1200-1400 kn/s and komodo at 1000-1200 kn/s on a 4 core computer.

Komodo and Stockfish do not autodetect the number of system cores by design- they use one, unless ordered else.

But...yes, the kilonodes per second feature is totally useless to the end user- only a programmer may have some use of it.

Next time you decide making an engine tournament, either read the manual, or ask.

Avatar of MSC157

You mean No.1 on phones?

Ah yes, Droidfish not Stockfish, or maybe it is the same, I don't know. :P

Avatar of pfren

Droidfish is a third party app, developed by Peter Osterlund, not an "official" one- still the best available for Android devices. The engine is made out of Stockfish sourcecode, optimized for that OS.

Avatar of kiloNewton

Stockfish 5 is the strongest chess engine now!

ccrl 40/4: June 2, 2014

1. Stockfish 5   3378

2. Houdini 4     3339

3. Komodo 7a   3321

Avatar of Daniel_Cohen

I believe Fritz 6 is free, and probably useful in middlegame situations but F13 is certainly over-optimistic about White's opening advantage.  Not sure if F6 would be in the top 10.

I would be careful to run the engines deeper, say to depth 25, to qualitatively evaluate the engine's quantitative positions.  Passed pawns seem to give most engines trouble (getting to d25 takes much longer) and I haven't seen an engine that recognizes that positions with no queens and bishops of opposite colors are far closer to drawish and thus should be evaluated closer to 0.00.

Caveat machina.

Avatar of mldavis617

I would think that the strongest engine would only be "strongest" against another engine and not always showing the best move for humans in a given position.  Engines have the advantage of "seeing" far more moves ahead than players, and what might be "best" at that depth might not be best against less than perfect play in earlier moves.  I usually limit the depth of my post-game analysis except for extended endgame positions, and unless I'm interested in playing engine vs. engine, (which I'm not) any engine above 3000 is going to teach me something.  Unless there is a flaw in the engine such as an endgame weakness or pawn play (as stated above), the quest for the "strongest engine" is chasing the holy grail.

Avatar of Daniel_Cohen
mldavis617 wrote:

I would think that the strongest engine would only be "strongest" against another engine and not always showing the best move for humans in a given position.  Engines have the advantage of "seeing" far more moves ahead than players, and what might be "best" at that depth might not be best against less than perfect play in earlier moves.  I usually limit the depth of my post-game analysis except for extended endgame positions, and unless I'm interested in playing engine vs. engine, (which I'm not) any engine above 3000 is going to teach me something.  Unless there is a flaw in the engine such as an endgame weakness or pawn play (as stated above), the quest for the "strongest engine" is chasing the holy grail.

Well put.  And the search for the grail continues....

Avatar of RobotDaneelOlivaw

Why don't you just play against humans on here? At least you have a chance of winning, and a smaller chance of raging (when you lose to an engine for the 420th time), unless you want to analyse a game?

Avatar of EvgeniyZh

The current rating list of free engines: http://www.computerchess.org.uk/ccrl/4040/cgi/compare_engines.cgi?class=Free+engines&only_best_in_class=on&num_best_in_class=1&print=Rating+list&profile_step=50&profile_numbers=1&print=Results+table&print=LOS+table&table_size=100&ct_from_elo=0&ct_to_elo=10000&match_length=30&cross_tables_for_best_versions_only=1&sort_tables=by+rating&diag=0&reference_list=None&recalibrate=no

Avatar of pfren

Stockfish 5 is one year old, and Gull 3 (much stronger than 2.8) is out since May.

Avatar of Daniel_Cohen

Where is Gull 3 available?  My impression was that is was proprietary during the ongoing TCEC event and would be released to the public some time after that.

Thanks in advance of your response.

Avatar of EvgeniyZh

Gull 3 is available at it's site - see first page

Avatar of capalasker1

Dont Kill Chess with this Engines