lichess.org
Donate

Learn from your mistakes.

Analysis used to be unlimited, and someone (hugely) abused it.
Wait a second,who abused it? and how do u abuse this?? anyways, the new feature is atleast 50,000,000 times better than any other website for chess. Also, this is sorta off topic but i started a website on html, any experts here?
Amazing stuff.
Since the analyses are already done with masters book integrated, is there hope for an improved opening training from it?
And if you let me be greed, a training session of only positions from own games would be awesome.

Thanks Thibault! You guys are simply KILLING the competition. Since you asked for more feature ideas, what would it take to take a stab at a new "learn an opening" feature? With the current games library, I bet you could find the most popular openings played across all of the ECO codes and compile a set of 10-12 moves for both black and white and let the user play either color, with the computer taking the opposite color and playing the most popular (and also accurate) moves, allowing learning through repetition. It would be great to scroll through the list, pick an opening and add it to a personal library (like a study) that you can practice with whenever you want. Lichess could also track the player's accuracy in learning them, perhaps highlighting openings and moves which have been giving them trouble.
What an amazing feature. Probably one of the best made on this site, which has a bunch of it. One suggestion: a training based only on your games. So instead of only training with one selected game, you get puzzles chosen from all the games you played. It would be awsome especially for people who played a lot of games here. Thanks a lot @thibault and all lichess developers and patrons.
Thank you for your response @thibault
I understand it's impossible to allow unlimited analysis due to abuse. Please raise the limit of number of games that can be analyzed every day. I hope it's possible.
Can't you guys just use the Analysis Board? As far as I know it is not limited, and uses local resources.
@Gerurururu I have the same "if I were greedy" wish as you do.

I could see how it would be pretty computationally intensive, though (need to first StockFish the game to identify mistakes, then it seems like it's calculating evaluations for each entered response before telling me if it's a good move or not).

It might make life easier to include only games that the user has requested analysis and store the list of relevant moves in advance. Realistically, analyzing a game is probably a reasonably good signal that the user found a position either interesting or serious and would avoid the dilution of having dozens of "you hung your queen in a 1+0 bullet game" examples.

This topic has been archived and can no longer be replied to.