There are no excellent moves in computer analysis. The concept doesn't make sense. A "brilliant" move cannot be deemed any better by a computer than a forced capture. It can however, be deemed as worse.
Computer analysis - Legends
Sort:
iBotelho wrote:
My question is: How does the computer analysis conote? In other words, what is considered a mistake, blunder, innacuracy, good, and an excellent play.
I would also like to know what is the avg diff
I know that .33 of a pawn is considered 1 tempo.
As far as what is considered a blunder, and such, that would depend on the threshold you have your engine set at. Mine is set at .33, so any move worse than that gets another line of analysis.
If youre just talking in general, i have no idea.
ChessOath wrote:
There are no excellent moves in computer analysis. The concept doesn't make sense. A "brilliant" move cannot be deemed any better by a computer than a forced capture. It can however, be deemed as worse.
Maybe there is not how to fully say a move is brilliant, but my question is about the game analysis of this website. Sometimes I feel like I need a better understanding in what the computer based on to say my move was good, excellent, etc
iBotelho wrote:
ChessOath wrote:
There are no excellent moves in computer analysis. The concept doesn't make sense. A "brilliant" move cannot be deemed any better by a computer than a forced capture. It can however, be deemed as worse.
Maybe there is not how to fully say a move is brilliant, but my question is about the game analysis of this website. Sometimes I feel like I need a better understanding in what the computer based on to say my move was good, excellent, etc
That will depend on what the move accomplishes.
Oh on this website specifically? Don't trust it. I can probably play chess better than that thing and that is truely saying something. Also, if it says moves are brilliant then that makes no sense at all. I can't imagine that's true.
Diakonia wrote:
iBotelho wrote:
My question is: How does the computer analysis conote? In other words, what is considered a mistake, blunder, innacuracy, good, and an excellent play.
I would also like to know what is the avg diff
I know that .33 of a pawn is considered 1 tempo.
As far as what is considered a blunder, and such, that would depend on the threshold you have your engine set at. Mine is set at .33, so any move worse than that gets another line of analysis.
If youre just talking in general, i have no idea.
Hm, maybe Im asking gerenal wise. Dont you know how do differenciate inacuracy, blunders or a mistakes moves? Like, how you set the level to separate them?
ChessOath wrote:
Oh on this website specifically? Don't trust it. I can probably play chess better than that thing and that is truely saying something. Also, if it says moves are brilliant then that makes no sense at all. I can't imagine that's true.
So any computer analysis arent reliable or only this one? Any suggestions to see where the mistakes were done in the chess games?
iBotelho wrote:
Diakonia wrote:
iBotelho wrote:
My question is: How does the computer analysis conote? In other words, what is considered a mistake, blunder, innacuracy, good, and an excellent play.
I would also like to know what is the avg diff
I know that .33 of a pawn is considered 1 tempo.
As far as what is considered a blunder, and such, that would depend on the threshold you have your engine set at. Mine is set at .33, so any move worse than that gets another line of analysis.
If youre just talking in general, i have no idea.
Hm, maybe Im asking gerenal wise. Dont you know how do differenciate inacuracy, blunders or a mistakes moves? Like, how you set the level to separate them?
My first piece of advice is not to get caught up in terminology like "inaccuracy", "blunder", and such. You need to understand the "why", and not the label attached to the word.
In the computer analysis, what's the difference between inaccuracy, mistake, and blunder?
When you are reviewing the output from a Computer Analysis of a game, it can be useful to know how terms like "inaccuracy," "mistake," and "blunder" compare to each other.
Assuming the position is roughly even (no big advantage for either white or black), it works like this:
"Inaccuracy" - The computer evaluates that this move resulted in a position that is at least 0.3 points worse than the position resulting from the best move available
"Mistake" - This move is at least 0.9 points worse than the best move available
"Blunder" - This move is at least 2 points worse than the best move available - pretty bad! :)
These values are somewhat relative to existing/previous advantages held by one player or another. For instance, if you are already massively ahead - say, by 30 points - and you make a move that drops your evaluation to +28, you're still massively ahead; relatively speaking, the size of your advantage hasn't changed significantly. Therefore, such a move would not be labeled a blunder.
Why don't you just look at the numbers it gives and work it out for yourself? Of course we're talking about an analysis system that once gave me different numbers in two indentical positions (following a single repeat)... So yeah, like I said, don't trust it. You can download very strong chess engines and the software to make use of them on-line for free. Using chess.com's analysis feature isn't something worth doing.
ChessOath wrote:
Why don't you just look at the numbers it gives and work it out for yourself? Of course we're talking about an analysis system that once gave me different numbers in two indentical positions (following a single repeat)... So yeah, like I said, don't trust it. You can download very strong chess engines and the software to make use of them on-line for free. Using chess.com's analysis feature isn't something worth doing.
While the chess.com engine is weak, it is better on V3. I do think it would serve a valuable purpose if it used a graph, instead of terms like: blunder, inaccuracy, etc. With a graph you can get an idea where youre going wrong in games, as you will see a pattern.
I like the idea of a graph (and a better engine, obviously). Anyway, I think post 9 answered the OP perfectly and I'm probably being harsh on the engine. It is very handy for spotting any tactical errors in your games if you don't have an engine installed on your computer.
ChessOath wrote:
I like the idea of a graph (and a better engine, obviously). Anyway, I think post 9 answered the OP perfectly and I'm probably being harsh on the engine. It is very handy for spotting any tactical errors in your games if you don't have an engine installed on your computer.
<simulated wiping of hands> Well our work is done here...
My question is: How does the computer analysis conote? In other words, what is considered a mistake, blunder, innacuracy, good, and an excellent play.
I would also like to know what is the avg diff