What in the world is that about and are the moves next to it still a reference to mistakes made during the game?
Welcome to the community Noire_Diaspora
Yes, it is still a reference to mistakes made during the game. The word blunder was causing confusion because many times blunder means a game changing mistake and so the terminology was changed to 10+% meaning the number of moves each player made that lost them greater than or equal to 10 percentage points. This is the level it was always set at, but many people thought 10% was too low to be called a “blunder”.
I guess the problem is that the winrate scale is not “linear”.
If you are e.g. at around 50% winrate and make a blunder, which loses 50 points, your winrate might go down to 0.5%. If you then dump another group worth another 50 points, your next winrate will be maybe 0.1%.
Thus, both blunders are equal in points but one has -49.5% winrate change, the other one only -0.4%. It is therefore difficult to judge from the number of “blunders” about who will win the game…
Correct because it doesn’t matter if you lose by 10 points or 60 if you’re still losing. However, all 19x19 games also have the % win prob at the end of the game which is helpful for people who like to resign to get confirmation whether they were really behind or not and also if you become site supporter you get access to full game review not just blunders.
No where on interface it says that 10+% means 10+% decrease in chance. Anyone can think it means increase or just number of Top Game changing moves of this amplitude.
There is no such thing as a move that increases your chance to win. It is simply the move you make after your opponents blunder, assuming you do not also blunder.
Sorry to interupt, but this is only true if we assume perfect analysis, in which case the winrates would be 1 or 0 in each position.
And in a more practical sense: The analysis tool as it’s implemented at the moment, showcases positive percentage point changes for many moves.
There is, when it’s a move from your opponent. So somewhere I understand that people get confused.
I think that one thing thats also causing some confusion comes from not quite realizing what leela thinks is game changing. You can easily have bunch of >10% moves in a row, if there is some unsettled group / urgent move / ko on the board, but neither player wants to play and thus they keep playing elsewhere.
ps. i feel like like >10% is cleaner way than 10+% for telling that something has affected the winrate negatively
YES! Make it this!
I dont understand what the +win percentage is. It is certainly not a move after your opponent’s blunder like bhydden said, cuz i have seen +percentage moves during josekis, where neither side were making mistakes.
Maybe someone choose wrong/bad joseki for the overall whole board situation?
Or maybe there was some urgent move somewhere, that can cause leela to regard every move in the joseki as a mistake, because both players should tenuki from it and play the urgent move. You can easily rack up 20 “≤-10pp” moves from a single joseki this way.