Reviewing the answers to our daily exercises in The Chess Gym has given me the chance to see what kinds of positions most players can usually solve, and which they struggle with.
Broadly speaking, most players do well when dealing with:
Narrow variations (few alternatives)
Good moves for themselves
They tend to struggle with:
Broad variations (many alternatives)
Good moves for the opponent
Playing off of the “tree of analysis” popularized by Alexander Kotov in Think Like a Grandmaster, GM Josh Friedel calls narrow variations sticks. That is, it’s a long, straight path with few branches. I call the opposite kind of tree, where there are many short branches, a bush.
One problem is that many tactics books focus on stick-y variations, where there is a long, forced line that leads to the desired outcome; but in my experience, in practical chess, bush-y variations are more common. In most positions, there are many possible moves. These positions don’t always fit cleanly into a puzzle format, but they’ll determine whether you win or lose many of your real games.
Let’s look at a few examples.
Here’s a puzzle that most people got right.
Black has a straight shot to a forced checkmate.
32... Qe1+33. Kh2 Bd6+ 34. Nxd6 (34. g3 Bxg3+) Qg3+ 35. Kg1 Qxg2#
All of White’s moves are forced, except for the option of blocking with the pawn on move e4, but this is easily dealt with.
In contrast, here is a puzzle that almost no one got.
White has sacrificed a pawn for a piece, but is counting on the pin on the d-file to regain the piece. To do this, they need to add another attacker to the knight, and there seem to be three reasonable tries: e4, Qf5, or Qc4. It turns out that only one of these succeeds in getting the knight back without allowing Black to get a pawn in return.
18. e4 Nxe4! 19. Qxe4 Rxc1 20. Bxc1 Nc3! 21. Rxd8 Nxe4 leads to an endgame with equal material.
18. Qf5 Nxb3! 19. axb3 Rxc1 20. Bxc1 Qc8! again allows Black to escape thanks to the loose bishop on c1.
18. Qc4! doesn’t offer Black any tricks. White is just going to take the knight next move. For example, 18… Nxb3 19. Qxb3 Rxc1 20. Bxc1 doesn’t do anything now, because with the queen on b3 there’s no Qc8 trick.
What makes this so much harder than the previous puzzle? Rather than a single branch, there are three to deal with. Although in fact, many players did not consider Qc4 as a candidate at all. At first glance, Qf5 looks better, because it gets the queen off the line with Black’s rook.
There’s also the fact that finding the correct line hinges on finding defenses for the opponent, rather than forcing moves for oneself. The psychological aspect of this is often emphasized: most of us enjoy finding good moves for ourselves more than finding them for the opponent. But there’s also an algorithmic difficulty: if you suspect your opponent may have a resource, but can’t see it, how long do you keep looking?
It’s a little like looking for a parking spot in your car. If you’re driving away from your destination, there’s a very easy algorithm that guarantees you park in the closest spot: park in the first spot you see. But if you’re driving towards your destination, every open spot presents you with a difficult choice: take the spot, or keep driving and hope for a better one?
In the chess scenario, there’s not an easy way around this. The solution as presented in A, B, C form masks the difficulties of actually solving the position at the board, because it presents everything in a tidy format with no wrong turns. This rarely happens in real life.
Nonetheless, imposing some structure on your thought process can help. When you’re in the situation of looking for a win when you have multiple options, I would argue the thought process should look like this:
Start with the most promising option.
Try very hard to refute it.
If you can’t, play it.
If you can, proceed to the next most promising option.
What do you think? Have you made any changes to your thought process that helped you calculate more effectively?
Want to simplify your chess training? Check out The Chess Gym, my online chess learning community, where we focus on the most impactful techniques for leveling up your game.
And if you're interested in 1:1 coaching, send me a message at coaching@natesolon.com.