 |
| Black has just captured on e8 in Spanton (167) - Yasser Tello (161), London League Division 3 2018. Who stands better, and by how much? |
*****
*****
*****
*****
*****
*****
*****
*****
*****
White has two isolated queenside pawns, while Black has doubled f pawns. The knight is unchallengeable on the d4 square. The analysis engines Stockfish12 and Komodo12.1.1 reckon White is winning, but if it were Black to move, they reckon White would only have a slight edge.
27.Kb3 f5
*****
*****
*****
*****
*****
28.f4?!
I thought it was useful to blockade the black pawns, where possible, on light squares, but the engines reckon centralising the king by 28.Kc4 is much more important.
*****
*****
*****
*****
*****
28...Kf6?!
Neither engine likes the text, but they split on what Black should play. Stockfish12 at first reckons 28...Kh6!? leaves White only slightly better, while 28...f6!? gives White a winning advantage. Komodo12.1.1 reckons 28...f6!? is only slightly better for White, while 28...Kh6!? gives White the upper hand. However, given enough time, Stockfish12 comes to vary between reckoning 28...f6!? leaves White winning or with just a slight edge, finally, it seems, settling on the latter, whereas Komodo12.1.1 comes to view 28...Kh6!? as only slightly better for White.
One point about ...f6 is that it allows ...Bf7, while ...Kh6 allows kingside counterplay.
29.Kc4 b6?!
This probably makes it harder for Black to defend his queenside.
While I was thinking about my next move, YT offered a draw.
30.Kd5 Ke7 31.Nc6+ Kf6 32.Ne5 Ke7 33.Nc4 f6 34.Nd6
Even stronger is 34.Nxb6 Bf7+ 35.Kc5 Bxa2 36.Nd5+ Kf7 37.Nb4, winning the a6 pawn.
34...Bd7 35.a3 Kd8
While I was thinking about my next move, YT offered a draw.
36.g3 Ke7 37.h3 Kd8 38.Nc4 Kc7 39.Ne3 Bc8 40.Kd4 b5
Black gets a lost pawn-ending after 40...Be6 41.Nd5+ Bxd5 42.Kxd5.
41.Nd5+ Kd6 42.Nxf6 bb7 43.Nh7 Bd5 44.Ng5 a5 45.h4 a4 46.Nh3 Bb3 47.Nf2 Bc2 48.Nh3 Bb3 49.Ng5?
This allows a drawing resource, so best is 49.Nf2.
*****
*****
*****
*****
*****
49...Bg8?
Black draws after 49...Bd5, the point being 50.Nh3 allows a draw by threefold repetition of position (after 50...Bb3). If instead White plays 50.Nh7, then 50...Bc4 51.Nf6 Bf7 leaves White with no way of improving his position, eg 52.g4 Ke6 53.Nd5!? fxg4 54.Ne3 Be8! 55.Nxg4 Kf5 etc. Finally, 50.Ke3 is met by 50...Kc5.
50.Nf3 Bf7 51.Ne5 Be8 52.Nd3
The game was adjourned here.
YT sealed ...
52...Bd7
... but resigned without resuming. A simple continuation would be 53.Ne5 Be8 54.c4 bxc4 55.Kxc4 Ke6 (what else?) 56.Kc5 Ke7 57.Kd5 Kf6 58.Kd6 Bb5 59.Nd7+ Kf7 60.Nb6 Kf6 61.Kc5 bb8 62.Kb4 etc.
Hi Tim, would you recommend the study of knight vs bishop endgames? Are there common patterns and techniques like in rook endgames? Or are they more like king and pawn endgames, where a lot of precise calculation is needed and from a practical sense it's easy to turn a win into a draw or even a loss.
ReplyDeleteWhat this series show is there certainly are common patterns and techniques.
ReplyDeleteOther things being equal, these include:
1. The more pawns there are on the board, the more the knight is favoured.
2. Blocked positions in which the pawns cannot move are particularly favourable to the knight.
3. Positions with rival pawn-majorities, eg 2v1 on the kingside against 2v3 on the queenside, favour the bishop.
4. Mobile positions in which pawns are free to advance favour the bishop, especially if the knight-player has less space.
None of this is revelatory - endgame books usually state as much.
What this series also shows, and perhaps is not always emphasised in books, is:
a) The starting positions of the kings can be crucial and can easily override the relative merits of the minor pieces.
b) There are often hidden resources that can only be found (at least by humans) with careful study.
c) Analysis engines sometimes have trouble correctly evaluating bishop-v-knight endings, with a marked tendency to underestimate the strength of the knight.
d) Theory and practice do not always match. In particular the bishop-player has to beware positions that are theoretically equal, but in which the bishop has no targets because enemy pawns are on squares the bishop cannot attack. In such positions the knight-player can continue manoeuvring, knowing it is almost impossible to lose (but see the following point) while needing only an apparently slight slip by the bishop-player to garner the full point.
e) Careless slips and outright blunders are much more common in these endings, at least at club level, than might be supposed.