[FEN "r3k1nr/pp1b1ppp/4p3/B7/8/8/PPP2PPP/R3KB1R w KQkq - 1 13"]

This position is from ScorpioNN-Minic, in the 23rd TCEC season, after 12...Bd7. ScorpioNN evaluated the position as +2.18, while Minic gave +1.84.

I don't see where such a large advantage is coming from. Yes White has the bishop pair, but surely the bishop pair can't be that large an advantage. Black has no weaknesses, material is equal, White barely has a development advantage. The only explanation I can think of is that Minic and ScorpioNN, being strong-but-not-top engines, might be misevaluating the position - but kibitzing Leela's +0.55 is fairly substantial (by Leela's standards). Leela's Win/Draw/Loss score of [44.4% W | 46.7% D | 8.9% B] indicates White has almost as much chance of winning as of drawing.

  • Black’s knight is not developed, and black cannot castle right now. Sep 15, 2022 at 17:18
  • Stockfish gives +0.61, this seems a little "under" to me by (very different!) human standards, I'd say it's closer to 1. There's nothing cryptic here. (The bishop pair /can/ and /often is/ that large an advantage! But it's not here, at least not just yet.) For any more detail, you'd have to introspect the evaluation engines and/or hundreds of megabytes of game tree.
    – BadZen
    Sep 15, 2022 at 19:06
  • Bishop pair + a small hedge in development. That's not really "large".
    – Déjà vu
    Sep 16, 2022 at 3:06

1 Answer 1


Stockfish 15 evaluates this position at around +1 (from the first graph in the readme of https://github.com/vondele/WLD_model, this roughly seems to be a little over a 20% chance of winning), which seems reasonable to me. The bishop pair is a nice asset, and White's development advantage is perhaps more potent than it might seem. First of all, it's his turn. But second, after say 0-0-0, the d7-bishop will be attacked. It's likely Black will need to spend an additional tempo moving the bishop (leaving it defended by an f6-knight seems very flimsy). And then, a black rook won't be able to immediately move to d8, due to the a5-bishop. So Black may have to play ...b6 too, which will slightly weaken the queenside pawns. Since they're on dark squares, they may become a long-term target for White's dark-squared bishop later on. The a6-square would also be weakened, and conveniently this is a square that White's light-squared bishop already controls.

There's also the problem of the black knight. It doesn't really have any good outposts, and so it lacks a way to stably be active. Sure it can sit on f6, but if White plays a move like f3, the knight will be denied control of the e4- and g4-squares (a pawn on f3 would also limit a black bishop on c6). If the knight moves to d5, White always has the option of just kicking it away with c4. In open positions like these without a fixed pawn structure, a knight often does poorly due to there not being many squares that only its side controls.

One more thing I'd like to add is the opposing pawn majorities. If the pawn structure were symmetrical (say, 3v3 on the kingside and queenside), White would still have some edge, but not as much. Since White has a 3v2, he has the ability to create a passed pawn. While the same applies to Black with his 4v3, White will probably have an easier time in creating a passer due to his various advantages already discussed.

  • 3
    Only thing InitialIgnorance forgot to mention (explicitely) is that White probably will also grab control of the only open file d. I think +1 quite nails it, I'd hate to play Black here. Sep 16, 2022 at 7:07

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.