How would you count? Any estimate is surely a complete guess.
You hear numbers but I think one person said them and nobody ever questioned it.
How many patterns does a master know?
Sort:
Nice position. I've seen that idea many times before but it's definitely missable when short of time, pressure of facing an opponent, and an energy-sapping long game almost played out.
Oh and for anyone reading Rxh6 Ra6 skewer Kd5 is the idea and if the rook is taken stalemate.
There is an old commercial for Post Raisin Bran.
The little cartoon kid asks the fictitous Mr. Raisin Man, "...Gee Mr. Raisin Man how many raisins are there in Post Raisin Bran..." Then the little Kid's Black Horn Rim Glasses turn into an old fashioned cash register and ridiculously large numbers begin to scroll by in the lenses of his hornrim glasses.
Does anyone have maybe a rough estimate of how many patterns they would know?
As I recall Adrian de Groot said something like in the 10,000 range for masters and 100,000 range for grandmasters.
Bottom of first post here (link below) says as much.
http://www.chess.com/forum/view/general/the-effects-of-speed-on-skilled-chess-performance
sources at bottom of OP in link.
Bishopsmorethanrooks..........I believe it is 200 . People are going to RUSSIA to learn these 200...even GrandMasters.
Mika_Rao wrote:
As I recall Adrian de Groot said something like in the 10,000 range for masters and 100,000 range for grandmasters.
Bottom of first post here (link below) says as much.
http://www.chess.com/forum/view/general/the-effects-of-speed-on-skilled-chess-performance
sources at bottom of OP in link.
Must have taken a while to count.
Scottrf wrote:
How would you count? Any estimate is surely a complete guess.
There are wild ass guesses done by idiots and wild ass guesses with a little thought and work behind them.
For me personally, I could imagine there being a few thousand for a master. 10,000 seems a little bit much. Maybe with a broad opening repertoire leading to dissimilar middlegames. 100,000 seems ridiculous... but then again GM accuracy/knowledge also seems ridiculous.
I don't really agree with the chess by memory idea in the link however. Not only from experience, but also the argument has some problems. Kids have a real trump in their strong calculation. Higher rated OTB players don't always do better in blitz. There is more calculation in blitz than they seem to think. And they didn't account e.g. stronger players being stronger in OTB because factor X which also happens to correlate with known positions.
To be fair, I probably know a thousand positions with K&P vs K. 10,000 therefore doesn't sound impressive.
In fact, tens (perhaps hundreds) of thousands of positions with K&B vs K that I can evaluate as a draw.
ummmm... over 9000
Haha 
For sure counting certain ways would hugely bloat the number.
It may be another 10,000 number, but even if the number isn't too interesting, the discussion is worth something.
At least in the land of pretend trolls vs spam bots it's fresh air.
Mika_Rao wrote:
Haha
For sure counting certain ways would hugely bloat the number.
It may be another 10,000 number, but even if the number isn't too interesting, the discussion is worth something.
At least in the land of pretend trolls vs spam bots it's fresh air.
E.g. I put a king on one square. The other king has a minimum of 53 squares to go on, and the bishop on any of the remaining.
So 64*53*62=210,304 positions.
Double for the bishop of the other player. Double to include the calculations for a knight.
841,216 positions even the worst beginner will be able to evaluate instantly.
I was watching two masters (2200) play a uscf rated game under standard conditions where both had less than ten minutes left. The position resembled the diagram with black to move. However, Black, instead of playing the natural Rxh6 which easily drew the game, he played Rh3+ because he didn't know the stalemate pattern.
I heard somewhere that chess is mainly pattern recognition. Thus, how many patterns does a master know?