Ignore the performance rating. It's some strange algorithm that's based on your own rating. Accuracy is a better rough idea at how well you played, but again accuracy isn't everything, you can play very simply and get a high accuracy.
How is the estimated Elo from the game reviews so inaccurate?

Ignore the performance rating. It's some strange algorithm that's based on your own rating. Accuracy is a better rough idea at how well you played, but again accuracy isn't everything, you can play very simply and get a high accuracy.
My accuracy is most of the time about 70-80%, so I guess it is estimated more consistently than the elo estimation. Would you say premium membership is worth it for the game review or should I not bother with it, considering that most extra stats you get aren`t really that accurate anyway?

Because you did not blundered
That is another funny thing, the Analysis tells me I blundered once and made one mistake, while the game review tells me that I had 0 blunders and 0 mistakes. Kind of contradicts itself there as well.

Ignore the performance rating. It's some strange algorithm that's based on your own rating. Accuracy is a better rough idea at how well you played, but again accuracy isn't everything, you can play very simply and get a high accuracy.
My accuracy is most of the time about 70-80%, so I guess it is estimated more consistently than the elo estimation. Would you say premium membership is worth it for the game review or should I not bother with it, considering that most extra stats you get aren`t really that accurate anyway?
Accuracy is definitely 100% consistent and helpful, however it doesn't give the full picture, that's all.
As for premium, I mostly like premium for the unlimited puzzles and lessons , other sites have infinite free puzzles but I enjoy how chess.com's puzzles get harder as you improve and your puzzle rating increases. I also have benefited a lot from chess.com's lesson videos and the puzzles at the end that test your learning. If you are only considering buying premium for the game review I would not buy it, I would instead paste your game into an alternate site that has a chess engine analysis tool but for free, there are plenty of them. You "share" your chess game from chess.com and paste the given PGN/FEN elsewhere.

Ignore the performance rating. It's some strange algorithm that's based on your own rating. Accuracy is a better rough idea at how well you played, but again accuracy isn't everything, you can play very simply and get a high accuracy.
My accuracy is most of the time about 70-80%, so I guess it is estimated more consistently than the elo estimation. Would you say premium membership is worth it for the game review or should I not bother with it, considering that most extra stats you get aren`t really that accurate anyway?
Don't get premium unless you want the lessons. Go to lichess for puzzles and a stronger engine for free.

Ignore the performance rating. It's some strange algorithm that's based on your own rating. Accuracy is a better rough idea at how well you played, but again accuracy isn't everything, you can play very simply and get a high accuracy.
My accuracy is most of the time about 70-80%, so I guess it is estimated more consistently than the elo estimation. Would you say premium membership is worth it for the game review or should I not bother with it, considering that most extra stats you get aren`t really that accurate anyway?
Accuracy is definitely 100% consistent and helpful, however it doesn't give the full picture, that's all.
As for premium, I mostly like premium for the unlimited puzzles and lessons , other sites have infinite free puzzles but I enjoy how chess.com's puzzles get harder as you improve and your puzzle rating increases. I also have benefited a lot from chess.com's lesson videos and the puzzles at the end that test your learning. If you are only considering buying premium for the game review I would not buy it, I would instead paste your game into an alternate site that has a chess engine analysis tool but for free, there are plenty of them. You "share" your chess game from chess.com and paste the given PGN/FEN elsewhere.
They don't censor lichess anymore.

Yeah I already do my puzzles on lichess to have unlimited puzzles, so I wouldn`t get premium for them and you get at least 1 lesson for free each week anyway, which is not much but better than nothing. Interesting to know you can use lichess to analyze your games on other sites as well. Thanks for your help everyone.

hola,soy español.y nuevo en el mundo del agedrez.me pueden enseñar un poco?
Lo siento, pero no hablo español, sólo inglés y alemán. También soy novato, así que no puedo enseñarte mucho en ajedrez.

Accuracy is definitely 100% consistent and helpful, however it doesn't give the full picture, that's all.
On the contrary. Accuracy is a bad joke. It is nothing. Accuracy is just an invention of chess.com that measures nothing, and it is bad at even doing that.
There is already a good and consistent way to measure how well someone plays: rating.

Hikaru played a game of chess, and he got a very high estimated elo. Another lower rated guy and his friend recreated the game to see what their estimated elo was. It was much much lower, even though they played the excact same moves. So the higher rated you and your opponent are, the higher chess.coms elo estimate will be.
I`m a beginner in chess and currently only 400 elo and I played against Nelson bot 1300 rated and Komodo8 1200 rated, but the game review told me twice that I somehow played like a 1100 (pretty sure I`m nowhere near that level and just a week ago lost against players of my rating) and that Nelson and Komodo are 600 and 550 instead of 1300 and 1200. I played Komodo8 which is supposedly rated 1200 here https://www.chess.com/game/computer/89427847
The game review showed that I was supposedly a 1100 rated player with 81.1% accuracy
and the 1200 engine only played like a 550 rated player with 74.9% accuracy.
I maybe played a 100 chess games in my entire life, studied no openings so far and so I`m pretty sure I`m nowhere near a 1000 and how come chess.com rates their own 1200 rated bot as a 550? Are the elos of the bots and the estimated elo of the game reviews pretty much useless and I should ignore it in the future? I played Nelson bot just before that here: https://www.chess.com/game/computer/88641713
and he too was only rated 600 instead of 1300 with 76.1% accuracy and I again was somehow rated as a 1100 with 79.3% accuracy. I liked the accuracy and estimated elo ratings of the game reviews and thought about going premium to see them any time but if they are that far away from being accurate, it kind of misses the point. What do you think about the accuracy and elo reports in the game reviews? Did you experience the same?