Jerome Gambit: Maia Chess Update (Part 1)

 

                                                                 

Over the last few years I have looked at a couple of chess-playing computer programs.

There is Boris-Trapsky. See:

"Jerome Gambit: Boris-Trapsky (Parts 1 & 2)"

"Jerome Gambit: Silicon Miniature (Parts 1 & 2)"

"Boris-Trapsky Makes A Return Visit"

Then, there is maia. See:

"Jerome Gambit: The Human Move (Parts 1, 2, 3, 4, & 5)"

"Jerome Gambit: I'm Really Getting Too Old For This Fast Stuff... (Part 2)"

That last post is worth quoting from:

By the way, it turns out that my opponent was Maia Chess, "a human-like neural network chess engine"

Maia’s goal is to play the human move — not necessarily the best move. As a result, Maia has a more human-like style than previous engines, matching moves played by human players in online games over 50% of the time..

Maia is an ongoing research project using chess as a case study for how to design better human-AI interactions. We hope Maia becomes a useful learning tool and is fun to play against. Our research goals include personalizing Maia to individual players, characterizing the kinds of mistakes that are made at each rating level, running Maia on your games and spotting repeated, predictable mistakes, and more...

Maia is particularly good at predicting human mistakes. The move-matching accuracy of any model increases with the quality of the move, since good moves are easier to predict. But even when players make horrific blunders, Maia correctly predicts the exact blunder they make around 25% of the time. This ability to understand how and when people are likely to make mistakes can make Maia a very useful learning tool.

The most recent version of maia chess is Maia9.

Here is a Jerome gambit "avoided" game (a line first mentioned on this blog in  "Offside!") featuring Maia5, with Maia9 in the notes.

Judge for yourself how "human" the computer program's moves are (especially the mistakes).

maia5 - baxali

15 minute rapid, lichess.org, 2026

1.e4 e5 2.Nf3 Nc6 3.Bc4 Na5 

4.Bxf7+ 

The earliest example that I have of this move in this position is in the game Sidran, Ezra - Vong, John, Compuserve email, 1992 (1-0, 9).

Of course, the game also could have continued 4.Nxe5 Nxc4 (4...Qe7 5.Bxf7+ Kd8 6.d4 Nf6 7.Bg5 d6 8.Nc3 dxe5 9.dxe5+ Nd7 10.Bxe7+ Bxe7 11.e6 Bb4 12.exd7 Bxd7 13.Be6 Bxc3+ 14.bxc3 Black resigned, Chung,M - Bonney,E, corr IECG, 1995; 4...d6 5.Bxf7+ Ke7 6.d4 dxe5 7.Bg5+ Nf6 8.Bd5 c6 9.dxe5 cxd5 10.exf6+ gxf6 11.Bh4 Qd6 12.O-O Bh6 13.exd5 Bf5 14.Re1+ Kd8 15.Qh5 Qf8 16.Qxf5 Black resigned, KnightBiker - torpocso, FICS, 2005; 4...Nf6 5.Bxf7+ Ke7 Here Black resigned in KingSaladdin - mickeyl, FICS, 2006 6.Bb3 Nxb3 7.axb3 d6 8.Nf3 Bg4 9.Qe2 Qd7 10.e5 dxe5 11.Qxe5+ Be6 12.O-O Re8 13.Ng5 Kd8 14.Nxe6+ Rxe6 15.Qa5 a6 16.Nc3 Bd6 17.d3 Kc8 18.Qg5 Rg8 19.Bf4 Bxf4 20.Qxf4Nh5 21. Qg4 g6 22.Rae1 Rge8 23.Ne4 Nf6 24.Nxf6 Rxe1 25.Qxd7+ Kb8 26.Nxe8 Rxf1+ 27. Kxf1 Black resigned, marburgcedric - HoldenHumphreys, FICS, 2008; 4...Bc5 5.Bxf7+ Kf8 6.Qf3 Nf6 7.Bd5 d6 8.Nf7 Qe7 9.Nxh8 Be6 10.O-O Bg4 11.Qf4 c6 12.Bb3 Ke8 13.Nf7 Nxb3 14.axb3 Qxf7 15.c3 Kd7 16.d4 Bb6 17.Be3 Rf8 18.e5 dxe5 19.dxe5 Nd5 20.Qxg4+ Ke8 21.Bxb6 axb6 22.Ra8+ Ke7 23.Qg5+ Ke6 24.Rxf8 Qxf8 25.Nd2 h6 26.Qg6+ Ke7 27.e6 Nf4 28.Qe4 Nxe6 29.Re1 Kf7 30.Qxe6 checkmate, wcaleb - fastcamaro, FICS, 2007) 5.Nxf7 (5.Nxc4 d5 6.Ne5 Qe7 7.f4 f6 8.Qh5+ g6 9.Nxg6 Qxe4+ 10.Kf1 Qxg6 11.Qe2+ Kd8 12.d3 Ne7 13.Kf2 Nf5 14.Re1 Bc5+ 15.Kf1 Qg3 16.hxg3 Nxg3 checkmate, KevinMontero_5STA - Roland1990, lichess.org, 2020) 5... Kxf7 6.Qf3+ Nf6 7.Qc3 d5 8.exd5 Nb6 9.b4 Qe7+ 10.Kf1 Qd7 11.a4 Nxa4 12.Rxa4 Qb5+ 13.d3 Qxa4 14.Bg5 Bd6 15.Bxf6 gxf6 16.h4 Qxb4 17.Qxb4 Bxb4 18.f3 Bc5 19.Nd2 Bd4 20.Ne4 a6 21.c3 a5 22.cxd4 Kf8 23.Nxf6 c6 24.dxc6 bxc6 25.f4 c5 26.dxc5 Kf7 27.Ne4 Kf8 28.c6 h6 (0-1, 40), Matogrosso (computer) - Jappe,K, Utrecht, 1992 

[to be continued]