Engine Downloads

 

    Over the weekend I spent some time checking out the interesting engine Reckless. At the tine I checked it was tied with Alexandria for place 5-6 on the CCLR 40/15 rating list. Places 1-4 are Stockfish 18, PlentyChess, Torch and Obsidian. 

    Reckless is an open-source engine written in the Rust programming language. It is designed to be highly aggressive and has gained attention for its strong performance in major computer tournaments. In the beginning its estimated rating was fairly low, but newer builds have shown a dramatic increase in strength. 

    While searching for a download I discovered the interesting RWBC (Regensburg WinBoard Circuit) site which is a specialized computer chess website hosted by Gunther Simon that is dedicated to testing and rating engines that run on Windows.

    To quote the site, “You'll find here information about most ever released chess programs running under the above protocols...There are also archived results and games from various old RWBC tournaments...Moreover I have collected all kinds of interesting/useful computer chess links in the link section.” He also offers downloads of opening books. The site is definitely worth a visit. 

    As a test I conducted a tournament with the top engines I have on my laptop. The time limit was 5 minutes plus 2 seconds per move. The results are perhaps a bit surprising because the top rated Stockfish didn’t sweep the tournament. But then again perhaps not. On the CCLR rating list the other top engines have managed to hold their own against Stockfish. 

    Also, I have used a combination of Berserk and Stockfish to play four draws in correspondence games against ICCF International Masters. Because all of these top engines are so close in strength it really matters little which one you use for doing your analysis.

 

 

    In the following game Alexandria comes out of the opening with a slightly inferior position and Reckless keeps up the pressure until Alexandria’s position finally collapses. Reckless’ Rook offer at move 33 is especially nice. 

    Alexandria is another engine that has rapidly climbed the ranks to become one of the strongest open-source engines in the world. It is sometimes described as having a playing style that is less aggressive and more positional compared to engines, like Berserk for example, specifically tuned for tactical aggression.

[Event "Blitz 5 min + 2 sec per move"] [Site "?"] [Date "2026.??.??"] [Round "?"] [White "Reckless"] [Black "Alexandria"] [Result "1-0"] [ECO "D17"] [Annotator "Stockfish 18"] [PlyCount "81"] [EventDate "2026.02.07"] [EventType "tourn"] [SourceTitle "Fritz Engine Tournament"] [Source "Doe"] {D17: Slav Defense} 1. Nf3 d5 2. d4 Nf6 3. c4 c6 4. Nc3 dxc4 5. a4 Bf5 6. Ne5 Qc7 7. g3 Nbd7 8. Nxc4 e5 9. dxe5 Nxe5 10. Bf4 Nfd7 11. Bg2 g5 12. Ne3 gxf4 13. Nxf5 O-O-O 14. Qc2 Ng6 {So far all the opening moves have been main line moves that gave been played by humans. who generally castled K-side, or on occasion Q-side. Stockfish's top choice is 15.O-O which it favors ever so slightly over 15.e3} 15. e3 fxe3 {This is better than taking the g-Pawn because is draws the N back for its aggressive position.} 16. Nxe3 {Black's next move was apparently played with the idea of exchanging the B for the N which does not appear to be a good ideat. A better plan is ...a5, ...Kb8 and ...Bb4 with equal chances.} Bc5 17. O-O $1 Bxe3 18. fxe3 {White's advantage is now at about 3/4 of a P.} Nde5 19. Ne4 h5 {Initiating a K-side with this move is black's best course of action} 20. b4 {Countering on the Q-side.} h4 21. b5 hxg3 22. bxc6 Nxc6 23. hxg3 Rh5 24. Rab1 {Without having made any major errors black's a drifted into an inferior position, but the next move is a mistake that allows white to place the Q aggressively with a gain of time. A better idea for black would be 24...Qe7 preventing white's next nove. Black could the play ...Rd7 countering the doubling of white's Rs on the b-file/} f5 {A tactical mistake that loses a P.} 25. Qe2 {How is it possible that Alexandria missed this?!} Rh6 (25... Rdh8 {Allows a winning attack...at least for an engine.} 26. Rfd1 fxe4 27. Qg4+ Kb8 28. Rd7 Qc8 29. Rbxb7+ Qxb7 30. Rxb7+ Kxb7 31. Qxg6 Kb6 32. Bxe4 R8h6 33. a5+ Kxa5 34. Qg7 Ne5 35. g4 Nxg4 36. Qxg4 { This position is evaluated at about 4.5 Ps in white's favor.}) 26. Rxf5 { Black has no compensation for the P.} Nge5 27. Qf1 Qg7 28. Qf4 Rg6 29. Ng5 { Hoping for Bh3.} Nd3 30. Qg4 {An interesting development. White's pieces are clustered on the K-side and not only is his K defended, but they are poised for an attack on the Q-side.} Rd7 (30... Qd7 {might work better.} 31. Rbb5 Nde5 32. Qe4 Qd1+ 33. Rf1 Rxg5 34. Rxd1 Rxd1+ 35. Kf2 a6 {It's an odd material imbalance. It could present a problem for humans to navigate the complexities of the position, but white has nearly a 4 Pawn advantage and so it's a clear win for an engine.}) 31. Bxc6 {Destroying the K's position.} bxc6 32. Qc4 (32. Rf7 {losing a R.} Qxf7 33. Nxf7 Rxg4) 32... Kd8 {White fininhes up with a few heavy blows/} 33. Rf8+ {[%mdl 512]} Kc7 (33... Qxf8 34. Rb8+ Ke7 35. Qf4 (35. Qe4+ Kd6 36. Qxg6+ {etc.})) 34. Ra8 Rxg5 35. Rxa7+ Kd6 36. Rxd7+ Qxd7 37. Qxd3+ Rd5 38. Qe4 Kc7 39. g4 Qd6 {intending ...Qg3+.} 40. Qh7+ Qd7 41. Qxd7+ { Black resigned. Accuracy: White = 91%, Black = 61%.} 1-0