poor analyses

Submitted my last two games for computer analysis (sic) to find that in the first which ended in checkmate in 23 moves on my part, and the latter, which had the same result in 30 moves - both were (noticably) in error according to the 'analyses.' I beg to differ. Although not an historical student of chess, it has been my pleasure to enjoy the game some 50 years, during which, like many of us, my caliber of play has been cyclic.

But trying different tactics and strategy during a game; that's the challenge and much of the enjoyment in play, to me. That my response cannot be named after some apparent Italian who lived a few hundred years ago, and started much the same game, and then to have shoved down my throat a completely DIFFERENT game than my own; followed the moves, and it became readily apparent this wasn't my game at all, and perhaps because it wasn't the 'analysis' just burped that mine didn't follow the prescribed or previous.

Makes me wonder, when said computer analysis found 45 errors and 3 goobers in a game that lasted a total of 23 moves, and then went on to improve my failings (sic) in the latter game to merely 14% of the moves. And i thought i was doing pretty good until then.

Don't know the 'average' number of moves for most chess games, having heard that 70 or so was top-end, and 4 an embarrassment; seems the fewer moves taken to checkmate the better; guess that's wrong according to the computer's 'analyses.' Perhaps every possible game of chess has been played, and a record retained; great, my game didn't follow the previous, so must be wrong. That provides little but confusion when trying to better understand the game, in my opinion.

Better to evaluate singly, as it's possible the evolving game might indeed be different, though not wrong. Anyway, got that off my chest, and would be curious if anyone else has been equally disappointed

The computer analisis in the site is one of the few almost useless features. I have heard if you are a premium member you get a stronger analisis, but I still feel the computer on the site isin't a good analizer.

You can't record every possible sequence of moves in a chess game because there are more possible moves in a 40-move game than there are stars in the universe. ... Or something to that effect, I'm bad with quotes.

When did the analysis deviate from your game? I'm sure it at least didn't complain with the first few moves. 

The computer is not claiming that its suggestions were perfect, it evaluates each position and gives it a rating and suggestions, based on the strength(rating) of that particular chess engine and how many moves ahead it can calculate without error. 

But computers aren't perfect and can't grasp certain long-term positional goals and etc... mainly use it to catch your tactical mistakes.

And the game you won wasn't only about how good you played, but also with how well you reacted to your opponent's mistakes. 

Ok, so I took the liberty of having the chess.com computer analyze one of my own games for the first time.

Apparently, this computer gives you two lines of analysis every time it thinks you made an inaccuracy/mistake/blunder. The first analysis line starts with the move the computer thinks you should have played. The second line is its own rendering of what would happen if you played YOUR move. Obviously, the computer thinks differently than you or your opponent and will probably give you a line that isn't close to what is actually played later on.

I would say at most levels mating someone in 23 moves often says more about an opponents poor play than your genius.  The computer finds the "best" moves for you, but it also does so for your opponent.  A "mistake" or "blunder" should still be pointed out even if your opponent didn't take full advantage.