# How do I know what the rating of a game is?

I found this game on a random scrap of paper lying on the ground. I'm not sure how strong the players are. How should I go about determining that by using a chess engine (say, the ones hosted by lichess.org)?

[]

``````[FEN ""]
1. e4 e5 2. Nf3 Nc6 3. Bb5 a6 4. Ba4 b5 5. Bb3 Nf6 6. Nc3 Nd4 7. Nxd4 exd4 8. Ne2 Nxe4 9. Nxd4 Be7 10. O-O O-O 11. Re1 Nf6 12. Nf5 Re8 13. d4 d5 14. Bg5 c6 15. Nxe7 Rxe7 16. Rxe7 Qxe7 17. Qd2 Bf5 18. Re1 Qd6 19. c3 Ng4 20. h3 Nf6 21. Bf4 Qd7 22. Bd1 g6 23. f3 Nh5 24. Bh2 Ng7 25. g4 Be6 26. Be5 a5
``````
• ask a strong player. actually I had an idea for a programming project that would estimate player strengths based on a single game, like chess.com does.
– qwr
Commented Apr 28 at 16:10
• their estimate should be taken as for fun and not seriously
– qwr
Commented Apr 28 at 17:42
• There were some rather elementary tactical errors, so the players can't have been very strong. 1) 6...Nd4 hangs the pawn on e5 (but White didn't take it) 2) 8...Nxe4? loses to 9. Bd5 forking knight and rook 3) 14. Bg5 leaves the knight on f5 undefended (but Black didn't take it), and 4) 20. h3? Qh2+ starts a king hunt that will win the game for Black. Commented Apr 30 at 4:14

I am not these days a very strong player, but I can say with confidence that these players are somewhere around 1100-1200 on most scales. They are serious enough about their chess that they keep the score, but probably not at a club. The score sheet is just a scrap of paper. Probably they meet at their homes every Wednesday or something. They intended to resume this game next week, so it was probably played quite slowly. The errors are then more serious. Both of them have seen well-played games live or in books, but have not completely understood them. They play routine moves that are well-meant but inaccurate. I feel sure they enjoy it. Thanks for the chance to play Sherlock Holmes!

A detailed analysis is not necessary. White plays an old-fashioned line of the Lopez (6.Nc3, which is OK though) and Black makes the unusual response ..Nd4, which looks dubious. White decides not to take the Pe5, which I think he could have done, but takes the Nd4 instead, and that leaves his own N misplaced. It finds a good home on f5 only because Black played 11..Nf6? rather than 11.d5! White then gains an advantage but does not take the chance to spoil the Black pawns with 16.Bxf6 !(16..Rxe1+ 17.Qxe1 Qxf6?? 18.Qe8++) After this the game is fairly level but 20.h3?? is a huge blunder allowing 20..Qh2+. There is no mate but Black is much better after 21.Kf1 Nxe3+ 22.fxe3(forced) Re6.

If the game had been better it would taken a stronger player than me to spot the mistakes and misjudgements. If it been much better it might have taken a computer. I do understand that the OP asked if software existed to make the evaluation but I cannot improve on the answers given already. Chess.com provides analysis but AFIK no evaluation. You could of course record the average move evaluation for both players, and then calibrate those readings by doing the same for games by players with known ratings.

After composing this I checked up, and chess.com agrees with me on the analysis, but this is a simple case. Chess.com is very good at spotting tactical resources but less reliable in positional judgement. You would need something better for GM games.

• If my transcription is correct, there is no 21...Nxe3+ after 20...Qh2+ 21. Kf1. Commented Apr 30 at 4:06
• You are correct and I must apologise. Even so, Black is actually even better. I think the accurate line is 21...Qh1+! 22. Ke2 Qxg2! If then 23. hxg4 Bg4+ 24. Kd3 Bf5+ 25. Ke2 Re8+ 26. Be3 Bg4+ 27. Kd3 Qe4++. Quite long and so not strong evidence Commented Apr 30 at 19:47

Lichess doesn't determine the estimate ratings of the players, only chess.com does. Do note that you can only know the game performance of each player, not their actual rating, but usually it wouldn't be that far off.

Expanding on @qwr's comment, asking a strong player would be a good idea too, because they can tell how well the players are doing. But again, you can never find the real rating of these players, for example one of them could be really sleepy and throw the game.