anybody else find this? like it will say "you missed an opportunity to win a pawn" but actually that move was setting me up to win a rook. or it'll say you should have moved the knight to a more central square, but actually it was moved there to defend against an attack
If you use Show Lines you might see why the suggested line was considered better. Sometimes one move might be slightly better and your chosen line is still good, just not as good.
Connected rooks means the rooks don't have any pieces between them and they protect each other.
yes, it has happened to me too. I played a game where game review said that I missed an opportunity to win a rook. So, I looked at the sequence. the computer was correct, but it gave like a 12 move long sequence, where I don't think two humans would have seen it or even played the entire sequence correctly.
Game review is great, but it's crucial to complement that with your own understanding of the position, strategy and long-term plans.
Quite a lot. For example it might say I missed an opportunity to win a knight and when I click what I should have done, it's a five or six-move sequence at least. Quite often a move that set up a major gain of material or even a checkmate is classified as a "mistake", and once or twice even a "blunder" had that effect though it was probably because the opponent also blundered more decisively.
Quite a lot. For example it might say I missed an opportunity to win a knight and when I click what I should have done, it's a five or six-move sequence at least. Quite often a move that set up a major gain of material or even a checkmate is classified as a "mistake", and once or twice even a "blunder" had that effect though it was probably because the opponent also blundered more decisively.
Do you have some examples where that happened and the other side didn't have a defense?
the game review is accurate enough to say this move is good this move is bad for the most part..where it fails is telling you to analyze in a way that is not taught by anyone. ie "you failed and opportunity to offer an equal trade or you missed a chance to threaten a pawn" are paraphrased examples of the so called coaches analysis as rather ...meaningless in the large picture at the very least. many moves considerd good but not the best are praised for things such as "this brings a knight off the first rank and into the action but blah blah is best"
Do you have some examples where that happened and the other side didn't have a defense?
It's happened a few times among my hundreds of games here, but going back through them to search for examples might take hours. Also I am a basic member with only the right to one game review a day. But chessterd5's description in particular strikes a chord with my experience. Basically the computer sees deeper with its silicon "mind" but even GMs would have trouble spotting the suggested move sequence, much less little old me. If it happens in future games I'll send you an example.
I have also seen things like that in Analysis, which I use more than the Game Review because it doesn't have the one a day limit. Going over a game again at a crucial move, it spits out this long string of alternative moves with the punchline often well along in the string.
The wording of the game analysis is sometimes a bit off the mark. Play through the suggested line it thinks will follow. There are times when the "suggested best lines" are really only a sequence of moves that two computers would play, and that's not going to help (unless you only play computers I suppose). The wording, to me, seems to reflect the immediate move you make (i.e. get a piece off it's starting square), but once you look a the sequence it suggests that should follow, it becomes apparent that really it was more about something like "putting your Knight prevents them from launching some attack", type thing. Often the description is somewhat helpful, but there are times when that description isn't really the reason because the "reason" is a few moves down the line.
.
Stockfish can be dumb sometimes it happens especially with inaccuracies.
An attempt to add a game I played against gjonikalamburgudar on Jan. 24. Move 23, c4, was classified as a "miss" and Nd5 preferred but I think it put the Black Queen under pressure and contributed to Black's later resignation.
ive had it tell me i “missed” getting a pawn when the move i did set me up to capture a queen
Anybody find that the game review makes any sense anytime? The Nalimov tablebase can score under 70 accuracy in an endgame.
Anybody find that the game review makes any sense anytime? The Nalimov tablebase can score under 70 accuracy in an endgame.
I don't think the review is tied into the tablebases. The site also doesn't use Nalimov tablebases; it uses Syzygy which don't generate the shortest lines to win/draw.
Anybody find that the game review makes any sense anytime? The Nalimov tablebase can score under 70 accuracy in an endgame.
I don't think the review is tied into the tablebases. The site also doesn't use Nalimov tablebases; it uses Syzygy which don't generate the shortest lines to win/draw.
I realise that. What I meant is that you can put a Nalimov v Nalimov perfectly accurate 5 man endgame into review and the losing Nalimov can come away with a sub 70 accuracy. (You can tell from the suggestions that it's not using any tablebases.)
Here's a similar organic example https://www.chess.com/analysis/game/daily/604945279?tab=review . It's a mate in 31 and as victim I stretched it to 36, so objectively I outplayed my opponent in terms of accuracy by 5 moves, but game review awards him an 8.3 higher accuracy score. It also scores me a blunder for a move that every man and his dog knows is the only accurate move. (At least I scored better than Nalimov.)
anybody else find this? like it will say "you missed an opportunity to win a pawn" but actually that move was setting me up to win a rook. or it'll say you should have moved the knight to a more central square, but actually it was moved there to defend against an attack