Forums

what is "the frequency of the best move"

Sort:
hlkkln

because it is written in turkish I translated this line expressed as % value on the computer analysis statistics... it may be different in the english version. But it is always below than the proportion of the perfect moves given at the top line of the statistics to the total number of move at that game... if a move is not the best move how it can be deemed as perfect? 

RonaldJosephCote

  Te frequency of the best move refers to the number of grandmasters who would play it.  Any number of other moves would be very very good, but maybe not ALL masters would play it.

hlkkln
RonaldJosephCote yazdı:

  Te frequency of the best move refers to the number of grandmasters who would play it.  Any number of other moves would be very very good, but maybe not ALL masters would play it.

 

thank you for your answer but as I said I see these statistics in Turkish and it is written "mükemmel hamle" (perfect move) ... as far as I know "perfect" is something stronger than very very good and at least equivalent to the best. Now I would like to know if they have made a translation mistake. So Can you tell me if itis written "perfect" in english version?

RonaldJosephCote

   I'm sorry, I can't help you with "perfect" english, I can barely spell.  Its like NASCAR. On any given Sunday, everyone has a "perfect" car.....but only 1 can win. wink.png

hlkkln
Postafi yazdı:

Is this what we're talking about in the Turkish version 

 

Ok there is a little translation problem that perfect is different from excellent... never mind... but but do you agree with mr.cote's expalanation to distinguish what is excellent and what is the best?  

MickinMD

The % best move, means the % of time the player's move matched the best move Stockfish 8 calculated for the position at the depth the analysis is done.

Here, for example, are data from a game I won as White this month (it's here: https://www.chess.com/daily/game/187192962) and the chess.com "maximum" (11 minutes to calculate) indexes. It says each of us picked the same move Stockfish 8 would have picked around 60% of the time. It also says that, on average, we picked a move that was 0.35 or 0.37 Pawn Equivalents weaker that Stockfish's choice:

null

Here is a deeper calculation of each move, made on Lucas Chess with Stockfish 8, at a deeper level of 20 ply and 15 move depth, that took an hour to do on a 3.2 Ghz, quad-core processor. Each of the moves in blue match the move Stockfish would have made:

null

Here are the more-detailed indexes Lucas Chess, with Stockfish 8, produced for that game, though they are certainly too generous about ELO ratings: each of us made one very bad move in the game. The ELO stuff is a new feature that will surely change in the future:

null