Play vs computer
Play chess against an 800-rated computer
Casual Maya plays like a real 800 player. Free, in your browser, no sign-up.
Last updated
touch_appTry the opening: Casual Maya answers from what real 800 players play.
How the 800 computer plays
An 800 player develops pieces and castles some of the time, but often misses two-move tactics: a fork, a pin, a piece that becomes loose after a trade.
- 92
- average loss per move (centipawns)
- 5.8
- big blunders per 100 moves
- 3s
- average think, 10-minute game
- 2
- moves it looks ahead
For comparison, real Chess.com rapid players around 800 lose about 84 centipawns per move (measured on their own games, the same way). The bot is Casual Maya (800): see how the bots are built.
Who this level is for
You finish games and know the basic checkmates, but still lose pieces to simple tricks.
A sample game at 800
- 1.d4
- d5
- 2.Bf4
- e6
- 3.Qd2
- Nd7
- 4.Nc3
- Bb4
- 5.Rb1
- h6
- 6.Ra1
- Nf8?
- 7.Qd1
- Ba5
- 8.Qb1
- f6
- 9.Nf3
- Bxc3+
- 10.bxc3
- b6
- 11.Bc1
- Qe7
- 12.Qb5+
- Kd8
- 13.Bd2
- Rh7
- 14.Rd1
- Qa3
- 15.c4
- Bb7?
- 16.Bb4
- Qxa2
- 17.Nd2??
- c6
- 18.e4?
- cxb5
- 19.g3
- Bc6
- 20.Bh3
- a6?
- 21.Bf1?
- Nd7
- 22.h4
- Kc7
- 23.h5
- dxc4
- 24.Ke2?
- Qa4
- 25.Ba5?
- Re8
- 26.Kf3
- Kb8?
- 27.Bd3?
- Qa3
- 28.Nb3?
- bxa5
- 29.Rh2?
- cxb3
- 30.Kf4??
- bxc2
- 31.Bxc2
- Ka7?
- 32.Kg4
- Rb8?
- 33.Rd2?
- Rh8
- 34.Bb3?
- f5+
- 35.exf5
- Ngf6+
- 36.Kh4?
- Qxb3
- 37.g4
- Qf3
- 38.Rg2
- Qxg2
- 39.fxe6
- Qxg4#
- 0-1
How to beat a 800 computer
- 1Look for forks and double attacks: they win a lot of games at 800.
- 2Develop knights and bishops before moving the queen.
- 3When you're a piece up, trade pieces, not pawns.
How we measured this
Measured on 41 Chess.com players' public games, November 2026.
We picked random public Chess.com players rated 400 to 2000 and took their 5 most recent rapid games. Stockfish scored every move they played from move 5 on, comparing it with the best move, and we averaged the loss per move. We fitted a smooth trend through all the players so no single game sways a rating, then ran our bots through exactly the same measurement.
Correction: our first version used textbook reference values that overstated human mistakes; the bots are now tuned to these measured numbers.