Analysing Blunders

Is it possible to know WHY a particular move is a blunder?

1. Compare the engine line of the blunder and of the best move suggestion and try to understand the difference
2. Ask a stronger player (some are willing to help in the Analysis forum here)

Does it show a variation for the blunder move? Like how the opponent manages to secure an advantage? The site only shows a move is a blunder, and tells the best move.

Sorry, I meant turn on the engine and then follow it.

@Knightrider29 said in #3:

Does it show a variation for the blunder move? Like how
the opponent manages to secure an advantage? The site
only shows a move is a blunder, and tells the best move.
At lichess.org/forum/general-chess-discussion/how-can-i-improve-positional-chess-i-keep-losing-because-i-get-myself-into-a-bad-position?page=4 , in the game, httpscolon//lichessperiodorg/FQfZFcX8 , after 1 e4 e5 2 Nf3 Nc6 3 Bc4 f5 4 exf5 d5 5 Bb5 e4 6 Bxc6+ bxc6 7 Ne5 , the machine apparently characterized 7...Qd6 as a "blunder", believing that 7...Qh4 would have preserved a "0.63" advantage for Black, whereas 7...Qd6 gave a "2.88" advantage to White. The machine attempted to indicate the advantage of 7...Qh4 by providing the line, 7...Qh4 8 d4 Bb4+ 9 c3 Bd6 10 Nxc6 Nh6 11 g3 Qh3 12 Bxh6.
Apparently, by replying to 7...Qd6 with 8 d4, White did not properly exploit Black's blunder. The machine attempted to show a better reaction to 7...Qd6, 8 Qh5+ Ke7 9 Qf7+ Kd8 10 Nxc6+ Qxc6 11 Qxf8+ Qe8 12 Qxg7 Ne7 13 Qe5 Bb7 14 d4 exd3, securing the advantage for White. I guess 15 O-O would be next.

https://lichess.org/FQfZFcX8

The machine does not always make such an attempt to be helpful.

From what I understand, do you mean to say that engine was wrong here? Or that what was flagged off as a blunder is not so much of a blunder afterall? Btw, how can I get such an engine analysis.. like the (2.88 -> 0.74) thing

If you're not hanging the piece you just moved then it is usually leaving another piece unprotected or compromising king safety. Or you had a capture that you missed.

I hate losing material so sometimes I won't capture a piece if I don't want to lose that piece that soon. But the engine says WRONG. Bc it can see perfectly the nxt 32 moves. Well, we can't lol.

And importantly the computer is guessing opponents responses the nxt 32 moves which may only seriously be accurate above 2300. Under 1600 anything weird can happen at any time.

Don't hang a piece
King security
And take free captures after checking for traps. They could be pulling you away from defending a square also.

@Knightrider29 said in #6:

From what I understand, do you mean to say that engine was wrong here? Or that what was flagged off as a blunder is not so much of a blunder afterall?

No, the engine was doubtless right on both counts.

Incidentally, you have been flagged by the site for some dubious activity (hopefully not cheating, given the question); so you might want to get that taken care of first.

The engine is right, but it's important to keep in mind that blunders are simply flagged automatically if your position becomes so many centipawns worse than the recommended best move. For example, you can be -12 as black, and if you drop to -6, you probably didnt make a great move, but blunder, it depends...This is not necessarily bad if your aim is simplification, and still end up in a completely winning position.. but often if you're crushing your opponent the computer would prefer that you use your attack to the maximum extent, mate them if possible, or win as many pieces as can be forcibly done, all while restricting their counterplay maximally.

You still have to be able to use your judgement at the end of the day to do analysis. If you have bad judgement, your analysis will also probably be bad, which seems like a catch 22, but the key in my opinion is to continue to question yourself critically. It's not just about looking at the result, but the whole situation. Blunders happen, but if your position is absolutely ridiculous and hanging on by a thread, maybe the better question is to ask why you're in a situation where the comp might say you're equal, but have a house of cards that will fall with the slightest breeze.

You wont come up with all the right questions to ask--you'll probably be solving things that are irrelevant, like preventing Qh5+, instead of asking why none of black's pieces are not developed on move 7 with king in the center and an open diagonal to it from throwing his F pawn down the board or something. Eventually you'll either realize you're asking the wrong questions, or find your way to right ones. It takes time, but doing so will improve your analysis, which will improve your gameplay, and the two things can create a positive feedback loop. You'll have less and less irrelevant questions to ask as you play better, so it can lead to a situation where you have to ask better questions, which improves your analysis, etc etc. We all hit plateaus, and there's ways to accelerate this process, coaching, chess educational content, but it's this process that's essential for improvement I believe.

In that example Black's position completely falls apart after Qd6. Why? It's very clear from white perspective that they want to go Qh5+, Qf7+.. Qd6 does nothing to prevent that, and has the added bonus of un-defending the black bishop on f8. White's Queen will get deep into Black's position for free, their king will be perma-exposed, and they will probably lose even more material as the queen forks or forces his pieces into passivity around the king who will lose the right to castle.

Personally as a human, I'd probably play Nf6 to stop Qh5 because it helps you castle faster, and develops a piece. Losing the c6 pawn is just a reality that happens because black did not develop their pieces. Qd6 would be a reasonable move if the knight was already out.

@kindaspongey said in #5:

At lichess.org/forum/general-chess-discussion/how-can-i-improve-positional-chess-i-keep-losing-because-i-get-myself-into-a-bad-position?page=4 , in the game, httpscolon//lichessperiodorg/FQfZFcX8 , after 1 e4 e5 2 Nf3 Nc6 3 Bc4 f5 4 exf5 d5 5 Bb5 e4 6 Bxc6+ bxc6 7 Ne5 , the machine apparently characterized 7...Qd6 as a "blunder", believing that 7...Qh4 would have preserved a "0.63" advantage for Black, whereas 7...Qd6 gave a "2.88" advantage to White. The machine attempted to indicate ...
@Knightrider29 said in #6:
From what I understand, do you mean to say that engine
was wrong here? Or that what was flagged off as a blunder
is not so much of a blunder afterall? Btw, how can I get
such an engine analysis.. like the (2.88 -> 0.74) thing
As a mere former USCF ~1500 player, I am habitually cautious in my statements, but I would guess that the machine's opinion of 7...Qd6 was correct (although, perhaps, 7...Nf6 would have been a better improvement than 7...Qh4). My point was to indicate the sort of explanation that one MIGHT get from the machine as justification for its assessments. I am not sure when the machine will deign to provide details. Perhaps, it is necessary to post a link to the game (that is what I did in this case) and the machine automatically inserts some details.