Coaching Mistakes

Here, Coach says this move is dubious.

But, in the same position three moves later, it is the best move.

how is it the same position three moves later?

Same question 😅🙈

I played 6.Bc4+ and "coach" gave me an inaccuracy with a 93% accuracy score for the 11 move game: https://www.chess.com/analysis/game/live/145253629310/reviewAfter, the computer's recommended 6.Qd5+ Ke8 7.Qh5+ Ke7 8.Qe5+ Kf7, it prefers 9.Bc4+ as the best move.

Here's another coaching error from my most recent game:

A move that turns an elementary draw into a clear loss is a blunder, not a mistake. I frequently struggle to get an advantage against players much lower rated until they make a horrendous move in a simple endgame. These are blunders, whether they are simply terrible at endings or not.

In contrast to the dubious comments and often wrong analysis of chessdotcom's game analysis, ChessBase gets to the heart of the matter with less fluff and better accuracy. One could argue that 3...fxe5 is the losing move. 6...Kg6 is certainly a mistake, as checkmate now comes quickly. After the necessary 6...d5, White still has a decisive advantage, but not a forced checkmate. I've even lost once from the resulting position by allowing my queen to get trapped.

I think I know the answer to the first one.

Qd5+ and Bc4+ is practically the same, as we have seen. After Qd5+ the best play is to return to the position and play Bc4+. So the engine analyses them simultaneously, but at different depths, as one of two identical positions is further in the future. That's why the evaluation is not exactly the same, but they are always very close, and they fluctuate around each other.

Now comes the juicy part. Chess.com's Game Review script looks at the engine evaluation at a moment when the Qd5+ line is 0.1 points better. And it confidently declares that Bc4+ is an "inaccuracy".

Of course, we can always expect Game review to make incredibly dumb statements very confidently.