Blunders, mistakes and inaccuracies

I analyzed my last game (about 6am Central time in America on 9-11-26) and my opponent's last pawn move blundered checkmate. However, the game analysis didn't call that last move a blunder, but only a mistake. How can you make a move that gives your opponent forced mate in 1 and not call it a blunder????

I guess it is this game.

So, what is the difference between a blunder and a mistake? I just assumed that a blunder is a bigger mistake. And I can't think of a bigger mistake than a move that loses the game on the spot. I guess, like you said, because I was already winning.

I decided to google it. Apparently, a blunder drops your winning chances by 30%, a mistake drops 20% and an inaccuracy drops 10%.

To me a blunder turn a winning position into a losing position,
while a mistake either turns a drawing position into a losing position, or a winning position back to a drawing position.

that is the lichess defitions. there no generally agreed definition of blunder or mistake. One typical is blunder is bad move which you would not do under normal circumstances. That hard to turn in objective code. Unlike the above definitions. Ofcourse the mapping from engine output to probabilities is based on statics from strong players so even that is not really objective.

@BGW5422 said in #4:

I decided to google it. Apparently, a blunder drops your winning chances by 30%, ...

So, by this definition, if one's "winning chances" are ALREADY below 30%, it is impossible to make a blunder?

@kindaspongey said ^

I decided to google it. Apparently, a blunder drops your winning chances by 30%, ...

So, by this definition, if one's "winning chances" are ALREADY below 30%, it is impossible to make a blunder?

I think you could still blunder by lowering your chance to 21%. I say that because 30% of 30% is 9%, so if you dropped from a 30% chance of winning to a 21% chance of winning, you just lowered your chances by 30%.
Or, another way to look at it is: If the eval bar has you at +3 and you make a move that lowers you advantage to +2, that could be seen as a 33% drop in your advantage?? I have no clue if this has anything to do with LiChess' criteria, so it's a mystery. lol

@BGW5422 said in #4:

I decided to google it. Apparently, a blunder drops your winning chances by 30%, ...

@kindaspongey said in #7:

... So, by this definition, if one's "winning chances" are ALREADY below 30%, it is impossible to make a blunder?

@BGW5422 said in #8:

I think you could still blunder by lowering your chance to 21%. I say that because 30% of 30% is 9%, so if you dropped from a 30% chance of winning to a 21% chance of winning, you just lowered your chances by 30%. ...

It seems to me that going from 60% to 42% should either be described as
(A) a drop by 30% OF 60%, or
(B) a drop by 18%.

In general accuracy is only strict in equal positions. The reason is that a 0.0 eval to a 0.75 eval changes win rate quite a bit. However a 5.0 eval to 4.25 or 5.75 eval won't have as much impact on the result.

That's also the reason why if in a game one player blunders a queen early on then both players will have near perfect accuracy as it takes an incredible blunder to overcome that deficit.

Usually I find mistakes occur most often in positions where you're around 2.5 eval better and play a poor move that loses a lot of your advantage but not all of it so it's maybe down to 1.1 eval. They're also common in hopeless positions where you've simply speeded up the inevitable outcome.

Accuracy tends not to care of a player misses a mate in 1 if there still remains a mate in 10.