I might have gone on a rampage of blitz games today, and I like to analyze every single game. Turns out that now "I am not authorized to do that" when I request a computer analysis. Is there a limit to how many games one can request analysis per day or something like that? Never happened before, but then again I never played over 40 games in a day (I know, I shouldn't have).
The Max is either 25 or 30 a day, and 100/120 a week.
I would block blitz and bullet games analysis.
It got increased a while ago. The new limit is 35/day or 160/week, whichever is hit first. Also, if you already used over two-thirds of your weekly limit it goes down to 20/day so that you don't accidentally use up your whole weekly limit and can't analyze for multiple days in a row.
Would be nice to have the possibility to do the analysis locally once one hits the limit and have it uploaded to the server. Not sure if that would be too slow in one's computer/mobile app.
#5
You can analyse locally, there is a switch when doing the analysis board in the top right for "Stockfish 13+NNUE" which does it locally in your browser.
EDIT> Rereading your post I think you already know what I just wrote, and you're right it doesn't upload to the server to assist with future analysis. Maybe a future feature :-)
--------
My question is what is the difference between "Deep Stockfish 14+ server analysis" (which you have a 35/day limit) and "Cloud engine analysis" (which is unlimited) as to how we use each one?
From what I can tell Stockfish is run on the Lichess servers themselves, and Cloud is run on volunteer computers running Fishnet. But is it possible to use the Cloud analysis in lieu of Stockfish on the Lichess servers, or is the Cloud engine used for a different purpose?
lichess.org/features
You can analyse locally, there is a switch when doing the analysis board in the top right for "Stockfish 13+NNUE" which does it locally in your browser.
EDIT> Rereading your post I think you already know what I just wrote, and you're right it doesn't upload to the server to assist with future analysis. Maybe a future feature :-)
--------
My question is what is the difference between "Deep Stockfish 14+ server analysis" (which you have a 35/day limit) and "Cloud engine analysis" (which is unlimited) as to how we use each one?
From what I can tell Stockfish is run on the Lichess servers themselves, and Cloud is run on volunteer computers running Fishnet. But is it possible to use the Cloud analysis in lieu of Stockfish on the Lichess servers, or is the Cloud engine used for a different purpose?
lichess.org/features
They are entirely different things.
Server analysis is an evaluation of a whole game by Stockfish, either run on Lichess' servers or on a volunteer's computer through fishnet.
"cloud analysis" is something completely different. It consists of stored analysis for a given position. When you analyse a position locally in the browser, Lichess may decide to save the analysis you computed for later so that it doesn't get computer over and over again when other people want to evaluate the same position. Of course, Lichess can't possibly store analysis for each position that has ever been player or analysed by anybody so it's mostly limited to openings and popular/frequently analysed positions, e.g. maybe from a current high-profile game or when a Reddit post goes viral.
So if you see the green "cloud analysis" label, it simply means a stored analysis for that position was found on the server and shown to you. There wasn't any real computational power involved at that moment.
Server analysis is an evaluation of a whole game by Stockfish, either run on Lichess' servers or on a volunteer's computer through fishnet.
"cloud analysis" is something completely different. It consists of stored analysis for a given position. When you analyse a position locally in the browser, Lichess may decide to save the analysis you computed for later so that it doesn't get computer over and over again when other people want to evaluate the same position. Of course, Lichess can't possibly store analysis for each position that has ever been player or analysed by anybody so it's mostly limited to openings and popular/frequently analysed positions, e.g. maybe from a current high-profile game or when a Reddit post goes viral.
So if you see the green "cloud analysis" label, it simply means a stored analysis for that position was found on the server and shown to you. There wasn't any real computational power involved at that moment.
Thanks for the insight Ben! Since you seem to be involved in Lichess development, do you think "volunteering" one's own computing power to analyze one's games after the limit is reached (or even more frequently if it's not that slow) would be a reasonable feature to have? If it makes sense it would help reduce server load.
This topic has been archived and can no longer be replied to.