How to Make Opening Knowledge Stick

Using progressive question and answer to expand your knowledge

’m running a rapid practice tournament Saturday afternoon with my friend Martin Justesen from the Say Chess newsletter. If you want to get some great practice games, join the Lichess team and tournament!

---

Last week I wrote about why rote memorization doesn’t work in the opening. But what does work?

One technique I love is rewriting your knowledge of a line from memory. Another is progressive question and answer. When you’re studying an opening, write down all the questions you have. Then answer these questions using any sources available: books, courses, coaches, engine, database, your own analysis. Periodically review your questions and answer them, referring to your notes if you can’t remember. Build on your previous answers to ask more nuanced and detailed questions, and answer those.

This technique is all about asking why. This is always a great idea, but the question and answer method crystallizes it into a technique that’s repeatable.

I’ve been working on a data-driven Black repertoire vs. 1. d4 using the Lichess opening explorer to find lines that score well in practice at the club level. It turns out that when you poke around the database you can find some lines that score absurdly well. The backbone of the repertoire is a line of the Benko Gambit that absolutely crushes in practical games.

In one of the main lines, White’s most common move in the masters database is pawn to a7, voluntarily giving up a pawn. (This is almost never played in amateur games. One of the reasons this line is so effective is that White’s best moves are counter-intuitive.)

It seems strange for White to move this pawn just to force Black to take it. Here’s where we can use the question and answer method. I wrote out three initial questions I had at this point:

• What is the point of a7?

• How does a7 work out in terms of tempi? Is someone gaining or losing a move here?

• Why does White play it on exactly this move? Would it matter if they played it one move earlier or later?

I used the engine and database to answer my questions:

What is the point of a7?

It lures the rook to a7, where it will get hit by a later Nb5, gaining a tempo for White. This happens in the main line:

8. a7 Rxa7 9. Nf3 e6 10. Be2 exd5 11. exd5 d6 12. O-O Na6 13. Nb5

How does a7 work out in terms of tempi? Is someone gaining or losing a move here?

In the sequence a7 Rxa7, both sides invest one move, and two things about the position change:

• The pawn on a6 disappears.

• The rook moves from a8 to a7.

This means Black won’t have to spend a tempo later to capture on a6, but it’s almost impossible to develop without moving a knight or bishop to a6 anyway, so Black would have gotten this pawn for free at some point anyway. So the only real difference is the rook. In effect, White can choose whether the rook is on a8 or a7, and this move argues that it is worse on a7.

Why does White play it on exactly this move? Would it matter if they played it one move earlier or later?

One move earlier:

This seems to transpose to the main line without giving either side any important additional options.

7. a7 Rxa7 8. e4 O-O 9. Nf3 e6 and we’re back to the main line.

One move later:

Now it’s too late! Black wins by ignoring the pawn.

9. a7? Nxe4! 10. axb8=Q Rxb8

White is up a piece, but can’t defend against all the threats on the queenside. Gukesh lost to Firouzja by mixing up the move order in this way!

At this point I feel like I already have a much richer understanding of the position and I’ll be more likely to remember my lines if they come up in a game, as well as having more context to find the best path if my opponent deviates from the main lines.

I could now add some follow-up questions building on what I’ve learned so far:

• By playing a7 White sets up Nb5 with tempo, but what’s so great about Nb5 anyway? Wasn’t that knight doing a good job on c3 defending the d5-pawn?

• After 8. a7 would there be any point in continuing to ignore the pawn and playing 8…Na6?

• 8. e5 hitting the knight and grabbing lots of space in the center looks like an attractive option. Is this a good move? How does Black deal with it?

Give this technique a shot and let me know if it works for you!