Calculation Is Never Just Calculation
I usually don’t write about the same position twice, but the position I showed last week sparked a lot of interesting discussion. It’s not that there’s anything magical about this position. You face many decisions like this per game, that’s the whole point. But chess has a hologram-like quality where every part contains the whole.
Here again is the position:
The best sequence for Black is 10… cxd5 11. exd5 Nxd5, when the best White can do is 12. Nxb5, resulting in a position where Black has an extra pawn and no real problems.
Posting on Twitter confirmed what I already noticed: for players below master level, finding the right sequence was very tricky.
This included even quite strong players. I showed it in a class at the US Chess School, which includes kids from around the US. These are seriously talented kids, up to and including masters. Yet even in this strong group, less than 50% landed on the correct sequence.
But I also noticed something else: players seemed to either find the right sequence immediately, or not at all. For instance, when John Bartholomew encountered this position in a blitz game, he found and played the best move in six seconds:
(Did you know you can search for positions on YouTube using chessvision.ai?)
This suggests that the position is not, as I suggested last week, about calculation. Or at least not only about calculation. Very strong players are able to find the right move, not by crunching variations, but by applying some combination of heuristics, experience, and know-how.
Last week I briefly sketched my own thought process, but this week I’m going to try to unpack in more detail the principles strong players are using under the surface to arrive at the best decision. There are three big ones:
King Safety. Like me, John had an immediate negative reaction to opening the e-file with White’s rook lined up with the Black king. What makes this principle more difficult to apply is that, to keep the king safe, you have to give up something else: the queenside pawn chain. It seems that stronger players are more likely to prioritize king safety. Another factor is that in the initial position, the queenside pawn chain is the more salient feature; Black’s king appears quite safe. Only after the sequence of exchanges (which requires visualization) does Black’s king become a target. So strong players might also be more likely to appreciate changes to the position and trade one advantage for another.
Fischer’s Golden Rule. According to various online sources, Bobby Fischer said that whenever an enemy piece enters your territory, you need to neutralize it as soon as possible. I can’t find any direct evidence of Fischer saying this, but it’s a good rule nonetheless. Masters were more likely to remove the intruding pawn on d5 immediately. This helped simplify the process: they focused on deciding between the two captures, whereas weaker players became overwhelmed thinking about a slew of moves like 0-0, e5, b4, Nb6, Nc5, etc.
Forcing moves. Or limiting your opponent’s options, which is another way of saying the same thing. Many players were attracted to the knight move 10… Nb6. But White has a strong reply in 11. d6, when after 11… Bxd6 12. e5, White wins a piece (or similarly, 11… Qxd6 12. Qxd6 Bxd6 13. e5). The same problem applies to 10… Nc5. The combination with d6 isn’t too hard to find, but it’s not the reason I would reject Nb6. Rather, I wouldn’t consider Nb6 in the first place, because I can sense that it gives White too many options. You might notice that in the position after White wins the piece, Black in fact has three pawns for it, which is equal in material count. But in a position like this where all of White’s pieces are active, the extra pawns are just sitting ducks. So evaluation always plays a role as well.
Some players asked me how to work on getting better at decisions like this. The good thing is that these decisions are everywhere, so you don’t need to do anything special to find training material. In US Chess School, Greg Shahade suggested doing Solitaire Chess (aka Guess-the-Move) with games between 2600 and 2300 players.
I like this suggestion, and you can adjust it to your level. According to the Elo predictions, a 200 point rating gap corresponds to an expected score around 75%, and a 400 points to a score around 90%. Whether these predictions hold up in practice, the point is, there are players a little ahead of you who are beating players at your level very consistently. How do they do that? You can find out by studying their games.