Move Repetition Decisions (Part 1)

Jul 13, 2017, 11:18 AM |

A threefold repetition of the position implies an interesting decision for one or both players - whether or not to deviate. The following is a position from a game I (with the Black pieces) played in 2015.

White (2048)- Black (2198)

Before continuing, I suggest you pause and think about the position for a bit. Black is an exchange for a pawn down, and there is a perpetual available by checking back and forth on d3 and h3 (27...Qd3+ 28 Qe2 Qh3+ etc.). You have about ten minutes remaining on your clock. Make a decision.


 For context, I had earlier played a sacrifice I was quite happy with. In fact, I expected it to lead to a quick win. However, my opponent defended well, and the above position arose. Black is an exchange for a pawn down, without any immediate mate just yet. I should mention that in general I'm not one to sacrifice material, so I was naturally a bit sceptical of my position. Without much clock time left to make a decision, I decided I didn't have enough resources to checkmate the White King, ended up taking the perpetual.

I didn't think much of this until I eventually ran the game through a computer, which gave Black a serious advantage in the final position (nearly -1). This was surprising to me at first, but it makes a lot of sense. Once you realize that Black does not, in fact, need to give mate, and he has a huge positional advantage for a relatively small amount of material, it becomes obvious that Black should play on.

In White's Favour In Black's Favour
Material Quantity (one extra "point" of material) Far superior king safety
  Material Quality (excellent B/d4; ...Nd7-c5; Queen attacking as opposed to defending, etc.)
  Pawn structure (b2 is hanging, and f4 can potentially drop off)

Given all this, objectively the correct decision was to continue play with 27...Qh5! Below I've put an analysis file detailing some possibilities of how the game could go.

I suppose the moral of the story is that sometimes you just have to trust that your positions various plusses will end up compensating for the material in the (relatively) long-term, even if there's no immediate mate. I would hope in the intermittent year and a half since this game was played, I've learned my lesson and would play more ambitiously next time a similar situation arose in one of my games.


Appendix A: Solely for my amusement, I ran an engine match from the position after 29...Qh5. Here are the results:

(Houdini A had White in all six games; Houdini B had Black.)