Tony Kosten, Winning Endgames
Monday, 28 September 2026
Sunday, 27 September 2026
Hull 4NCL Round Five
Andy Corbyn (1713 ECF/1652 Fide) - Spanton (1953 ECF/1997 Fide)
*****
*****
*****
King's Indian Attack
1.Nf3 Nf6 2.g3 d5 3.Bg2 c5 4.0-0 Nc6 5.d3
Staying in KIA territory, but Stockfish18 and Dragon1 reckon 5.d4!? gives White a slight edge.
5...e5 6.Nbd2 Be7 7.e4 d4 8.Nc4 Qc7 9.a4 0-0
 |
| Now both sides have castled, how would you assess the position, which occurs 205 times in ChessBase's 2026 Mega database? |
*****
*****
*****
*****
Black has more space in the centre, which is something the engines normally seem to rate highly, but here they give White at least the better part of equality, perhaps thanks to the well-placed queen's knight, and the possibility of a kingside attack in King's Indian style with the push f4.
10.h3 Be6 11.b3 h6 12.Nh2!? Qd7 13.g4?!
The engines strongly dislike this, claiming White is slightly better after 13.f4.
13...Nh7!? 14.f4 exf4 15.Bxf4 Bxc4 16.bxc4 Bd6 17.Bxd6
Preserving the bishop-pair does not make much sense in this closed position, especially as White's dark-square bishop would have to take up a passive position on d2 or c1.
17...Qxd6 18.Nf3 g6!? 19.Qd2 Kg7 20.Rae1
The engines are happy with this, but also like 20.e5!? Nxe5 21.Nxe5 Qxe5 22.Bxb7 Rab8 23.Rae1!? Qd6 24.Bd5, claiming equality.
20...Ne5 21.Nxe5 Qxe5 22.Qf4 Rae8 23.Qxe5+ Rxe5
 |
| How would you assess this late-middlegame? |
*****
*****
*****
*****
Black's knight is better than White's bad bishop, but White's rooks will be more active, thanks to having two half-open files to work on, especially the b file. The engines reckon chances are equal.
24.Rb1 b6 25.a5! Re6
25...bxa5?! 26.Rb7 is what White wants.
26.axb6 axb6 27.Rb2 Nf6 28.Rfb1 Nd7 29.Ra1 Rb8 30.Ra7 Ne5 31.Ra6?!
The engines suggest 31.Kf2 or 31.Rb1!?
31...g5?!
The engines like 31...h5, eg 32.gxh5 gxh5, when the white king is looking a little exposed. Dragon1 at first reckons Black is winning, but it soon comes closer to Stockfish18's assessment of a slight edge for Black.
32.Bh1 Kg6 33.Bg2?!
White should be activating the king with 33.Kf2 or 33.Kh2, according to the engines.
 |
| Black to play and gain what the engines reckon is a large advantage |
*****
*****
*****
*****
33...Kg7?!
The engines give 33...h5!? 34.gxh5+ Kxh5, with ...Kh4 to come, when again White's king looks vulnerable.
34.Bf1 Kf8 35.Be2 Ke8 36.Rb3 Kd8 37.Kg2 Kc7 38.Kg3 Ng6 39.Bf3 Nh4 40.Bh1 Rf6?
Black should mark time with 40...Ng6 or 40...Kd8.
41.e5 Rf1
Even worse is 41...Re6? 42.Ra7+ Kd8 43.Rxf7, when 43...Rxe5 44.Be4 is winning, eg 44...Re6 45.Ra3 Re7 (White was threatening to double on the seventh) 46.Rf6 etc.
42.Ra7+ Kd8 43.Bd5?!
Probably stronger is 43.Be4, eg 43...Ke8 44.e6! fxe6 45.Rh7, reaching a position similar to one that will be arrived at in the game.
43...Ke8 44.Rba3
Again the engines reckon 44.Be4!? is winning.
44...Re1??
Inexplicably unprotecting f7.
45.Be4?
Inexplicably not capturing on f6. White's advantage after 45.Rxf7 is worth much more than a queen, according to Dragon1, although Stockfish18 is slightly less effusive. Nevertheless, even after the text, White is winning, according to the engines.
45...Re3+?!
Black can probably put up more of a fight after 45...Ke8, but 46.Kf2 Rc1 47.Rd7 Rxc2+ 48.Kg3 Rc1 49.Raa7 Rf1 50.Bd5 is convincing enough for White.
46.Kh2?
Correct is 46.Kf2 Rxh3 47.Bc6+!? Kf8 48.Bd5 etc.
46...Re2+?
Black should play 46...Nf3+, as comes later.
47.Kg3 Re3+ 48.Kh2? Nf3+
This cuts White's advantage to a minimum.
49.Bxf3 Rxf3 50.e6!? fxe6?
50...Rf6 holds.
51.Rh7 Rf7
There is nothing better.
52.Rh8+
This does not spoil anything, but 52.Rxh6 is strong, eg 52...Ke7 53.Ra7+ Kf8!? 54.Rh8+ Kg7 55.Rxf7+ Kxh8 56.Kg3, after which Black has too many weak pawns.
52...Rf8 53.Rh7
Not 53.Rxh6? Ke7, after which Black should hold.
53...Rf7 54.Rh8+? ½–½
It was not too late to play 54.Rxh6, transposing to the note at move 52.
Chess Tip Of The Day 539
The general objective every chessplayer should have during the game is to secure a harmonious cooperation between his pieces and pawns.
Alexey Sokolsky, The Modern Openings In Theory And Practice
Hull 4NCL Round Four
Spanton (1953 ECF/1997 Fide) - Roger Greatorex (1758 ECF/1754 Fide)
*****
*****
*****
*****
*****
*****
*****
Spanish Cozio
1.e4 e5 2.Nf3 Nc6 3.Bb5 Nge7
For quite some time this is Dragon1's second choice, but the engine comes to agree with Stockfish18 that the best continuation is 3...Nf6, with 3...a6 second-best (!?).
4.Nc3!?
More popular in ChessBase's 2026 Mega database are 4.c3 and especially 4.0-0, but the text is fashionable, for a reason I will explain after Black's next move.
 |
| How should Black proceed? |
*****
*****
*****
*****
4...Ng6!?
The 'normal' Cozio move, 4...g6, is out of favour because of 5.d4 exd4 6.Nd5!? Bg7 7.Bg5, with the upper hand for White, according to the engines.
5.d4 exd4 6.Nxd4
6.Nd5?! makes much less sense without a hole at f6.
6...Bb4 7.0-0 Bxc3?!
Black smashes White's queenside, but the resulting structure is not easy to attack, at least not before an ending, and in the meantime White has the bishop-pair and a lead in development.
8.bxc3 0-0
 |
| White has the upper hand (Stockfish18) or possibly a positionally won game (Dragon1, but the evaluation fluctuates, and eventually more-or-less settles on plus-over-minus) |
*****
*****
*****
*****
9.f4!
The engines agree this is best. White takes control of the e5 square, while preparing a kingside attack.
9...a6!?
Apparently a novelty, and preferred by the engines over 9...Nxd4, which was played in games featuring a 2200 and a 2295 as Black.
10.Bd3 d6 11.Rb1 Re8
Not 11...Rb8?? 12.Nxc6.
12.Qh5!? Nxd4?!
Not 12...Qh4? 13.Qxh4 Nxh4 14.Nxc6 bxc6 15.f5, after which the black knight has no retreat squares. But, rather than the text, the engines much prefer 12...Qe7.
13.cxd4 Nf8!?
This looks horribly passive, but it is probably best.
14.f5 Qf6
Perhaps a better way to stop White playing f6 is for Black to play 14...f6, or 14...Nd7.
15.c3 h6
White threatened to win the black queen with 16.Bg5.
16.g4
Perhaps 16.Rf3 is strongest.
16...Nh7 17.h4 g5!? 18.fxg6 Qxg6 19.Qxg6+ fxg6
 |
| Black has foiled the mating attack, but is losing a pawn |
*****
*****
*****
*****
20.Bxh6 Bxg4 21.Rxb7 Re7 22.Bc4+ Kh8 23.Bxa6??
White is roughly the equivalent of being a rook up after 23.Rf7, according to the engines.
23...Bc8 24.Rfb1
The engines reckon 24.a4!? Bxb7 25.Rxb7 Rxa4 is equal, White having a pair of bishops and a pawn for the exchange. Note that 24...Rxa6!? 25.Rb8 Re8 26.Rf7 may slightly favour White, according to the engines.
24...Bxb7
But here the engines reckon 24...Ra6!? 25.Rb8 Re8 gives Black at least the better part of equality.
25.Rxb7
The engines reckon 25.Bxb7 Rb8 26.Rb3 is completely equal.
25...Rxa6!?
Black has a slight edge after 25...Rxe4, according to Dragon1, but Stockfish18 disagrees.
26.Rb8+ Nf8 27.Bxf8?!
Probably better is 27.Rxf8+ Kh7 28.Bg5.
27...Rxe4 28.Bxd6+ Kh7
Black has at least a slight edge after the somewhat counter-intuitive 28...Kg7!, according to the engines.
 |
| White to play and equalise |
*****
*****
*****
*****
29.Be5!
After 29.Bxc7 Rxa2, Dragon1 reckons Black is winning, but Stockfish18 gives Black just a slight edge. The text threatens mate.
29...g5 30.h5!? Rxe5
Forced.
31.dxe5
 |
| How would you assess this rook-and-pawn ending? |
*****
*****
*****
*****
White is two pawns up, but it is Black to move, and the black king is more active than its white counterpart. The endgame is completely equal, according to the engines.
31...Rc6?
But not any more. Three moves seem to hold the balance: 31...Rxa2, 31...Re6 and 31...Kh6
32.Rb4 Kh6 33.Re4 Re6
 |
| The central passed pawn is blockaded, and h5 is about to fall, but the black rook has become passive - White's advantage is again worth about a rook, according to the engines |
*****
*****
*****
*****
34.a4?
The king should be activated, eg 34.Kf2 Kxh5 35.Kg3, when Black is in zugzwang, a possible continuation being 35...Re8 (35...Kg6 36.Kg4; 35...g4 36.Kf4) 36.a4 Re6 (36...Re7 37.e6) 37.a5 Kh6 38.Kg4 Kg6 39.Re2, when Black has to give way.
34...Kxh5 35.a5
Activating the white king also only draws.
35...Kg6 36.a6?! Rxa6 37.e6 Ra8 38.Kf2 Re8
 |
| Black has the less-vulnerable passed pawn, and the more-active king, but the ending is still drawn |
*****
*****
*****
*****
39.Kf3 Kf6 40.Rc4 Re7 41.Kg4 Kxe6 42.Kxg5 Kd5 43.Rd4+
The only move to draw, but not so hard to find.
43...Kc5 44.Kf4
Again an only-move, according to the Syzygy endgame tablebase.
44...c6 45.Rb4 Re1 46.Re4!? Rxe4!?
The draw is easy after this, whereas care is still required after 46...Rc1, but one drawing line runs 47.Ke3!? Rxc3+ 48.Kd2, reaching a well-known drawn position, although there is nothing stopping Black playing on.
47.Kxe4 Kc4 48.Ke5 ½–½
48.Ke5 was also an only-move, after which play might have continued 48...c5 49.Kd6 Kxc3 50.Kxc5.
Saturday, 26 September 2026
Hull 4NCL U2000 Round Three
FACED a Nigerian.
*****
*****
*****
Kristian Usifoh (1910 ECF/1953 Fide) - Spanton (1953 ECF/1997 Fide)
Catalan
1.c4 e6 2.d4 d5 3.Nf3 Nf6 4.g3 Bb4+ 5.Bd2 Be7!?
This by-now well-known manoeuvre sees Black spend a tempo misplacing White's dark-square bishop, the general consensus being the bishop is better developed at b2. However, 4...dxc4 and especially 4...Be7 are more common in ChessBase's 2026 Mega database.
6.Bg2 0-0 7.0-0
 |
| With both sides castling early, how would you assess the position, which occurs 10,378 times in Mega26? |
*****
*****
*****
*****
Black has a typical QGD setup, while White, thanks to the fianchettoed bishop, hopes to pressure the black queenside. Stockfish18 and Dragon1 give White the better part of equality.
7...Nbd7 8.cxd5!?
Playing for a Minority Attack is unusual in this variation, but has been tried by grandmasters. White normally continues developing with 8.Qc2.
8...exd5 9.Nc3 c6 10.Qc2 Re8 11.a3 a5 12.Rab1 b5!?
Probably a novelty. Known moves are 12...Bd6 and 12...Bf8, while Stockfish18 suggests 12...Nb6!?, and Dragon1 12...h6.
13.e4!? dxe4
The engines reckon Black is at least equal after 13...b4.
14.Nxe4 Nxe4 15.Qxe4 Bb7 16.Qf4 Bf6
Perhaps best is 16...c5.
17.Rfe1 Nf8?!
Black equalises with 17...Rxe1+ 18.Rxe1 c5, according to the engines.
18.Rxe8 Qxe8 19.Qc7!? Qd7 20.Bxa5 Rd8?
The engines much prefer 20...Qxc7 21.Bxc7 Ne6 22.Be5 Bxe5 23.dxe5 c5, reckoning White has only a slight edge (Stockfish18) or perhaps the upper hand (Dragon1).
21.Qxd7 Rxd7 22.Ne5!?
 |
| White has a winning advantage, according to the engines |
*****
*****
*****
*****
22...Rxd4 23.Bxc6 Ba6?!
The bishop is horribly passive on a6.
24.Re1 Ne6 25.Bc3 Rd6 26.Bg2 h6 27.h4 Nc7 28.Bb4 Re6 29.f4 Re8 30.Bd6 Ne6 31.Bd5 Bc8
Not 31...Rd8? 32.Nxf7!
32.Bb4 Rd8 33.Ba2
 |
| After much manoeuvring, White remains a pawn up and has denied Black any significant counterplay |
*****
*****
*****
*****
33...Bb7? 34.Nxf7! Kxf7 35.Bxe6+ Kg6 36.b3 Rd3
Black does now get some play, but only after dropping a second pawn.
37.Kh2 Bd4 38.f5+?
This throws away almost all of White's advantage by seriously weakening g3, as well as allowing the black king relative security on f6. The engines reckon 38.g4 and 38.Bg4 maintain a large advantage.
38...Kf6 39.Bd7!?
This may be best. The alternative is 39.Re2, eg 39...Rd1 40.Be7+!? Kxe7 41.Bd5+ Kd6 42.Bb7, but the engines agree 42...Rd3 gives complete equality.
39...Be5 40.Bxb5 Rxb3?
Drawing is 40...Bxg3+ 41.Kg1, and now 41...Rxb3.
41.Be7+! Kxf5??
After 41...Kxe7 42.Rxe5+ Kd6 43.Re6 Kc5 44.a4, White is two pawns up, but play is not simple, and Dragon1 (but not Stockfish18) reckons White may only have the upper hand, rather than being winning.
42.Bd7+ Kg6 43.Rxe5 Rb2+ 44.Kh3
Also winning is 44.Kg1.
44...Bg2+ 45.Kg4 h5+!? 46.Kf4 Rf2+ 47.Ke3 Rf3+ 48.Ke2 Rf7?!
Black's position is hopeless, but this does not help, as is also the case with 48...Rxg3?! 49.Kf2.
49.Be8 1-0
Chess Tip Of The Day 538
When there is a structural imbalance, opposite-coloured bishops tend to favour the attacker because the defender is often powerless to prevent the infiltration of his position.
James Rizzitano, Understanding Your Chess
Hull 4NCL U2000 Round Two
FACED a junior (born 2014)
*****
*****
Spanton (1953 ECF/1997 Fide) - Dominik Pitel (1818 ECF/1762 Fide)
Sicilian Bb5(+)
1.e4 c5 2.Nf3 d6 3.Bb5+ Bd7 4.Bxd7+ Qxd7 5.0-0 g6 6.c3 Bg7 7.d4 cxd4 8.cxd4 e6!?
Stockfish18 and Dragon1 like this alternative to the mainline 8...Nf6, reckoning the two moves are of roughly equal value.
 |
| How should White proceed? |
*****
*****
*****
*****
9.Nc3
There is nothing dramatic to be done, although 9.d5!? has some interesting points, as the engines reckon the reply 9...e5!? leaves White with the upper hand. They prefer 9...exd5!?, meeting 10.exd5 with 10...Ne7, and 10.Qxd5 with 10...Nf6.
9...Ne7 10.Qe2 Nbc6!?
This apparent-novelty is the engines' top choice. Known moves are 10...0-0 and 10...d5, although the engines dislike the latter because White has 11.Ne5.
11.d5!? exd5 12.exd5 Nb4?
Going after the d5 pawn is too greedy. The engines give 12...Ne5 13.Bf4!? Nxf3+ 14.Qxf3 0-0, when they agree White is slightly better.
13.Re1 Bxc3?
Consistent, but bad. Best, according to the engines, is 13...Kf8!?, but White has much the better game.
14.bxc3 Nxd5 15.c4 Nc7
This was DP's idea. The queen's knight can hop into e6, unpinning the king's knight. But Black is horribly weak on the dark squares, and after ...
16.Bb2 Rg8
... loses kingside castling rights, and has too many weaknesses to defend.
17.Ng5 Ne6
There is nothing better.
18.Nxh7 0-0-0!?
Again, there is nothing better, but now Black loses the exchange.
19.Nf6 Qc6 20.Nxg8 Rxg8
 |
| How would you assess this middlegame? |
*****
*****
*****
*****
White is up the exchange, has the only bishop on an open board, and has the safer king, advantages worth close to +4.00, according to the engines.
21.Rad1 Nf5 22.Qe4 Qb6 23.Rb1 Qa5 24.Bf6!?
The engines like this post for the bishop, where it is relatively free from harassment, covers d8, prevents black knights using g5 or h4, and continues to stop ...Rh8.
24...Qa6 25.Qd5
Threatens 26.Rxe6!
25...Re8 26.Red1 Kc7?
The engines suggest 26...Nf4 or 26...Ne7!?, but agree White is winning.
27.c5! Rf8
Not 27...dxc5? 28.Qd7+, nor 27...Nxc5? 28.Qxf7+.
28.cxd6+Nxd6 29.Be5
Technically even stronger is 29.Rxb7+!, provided White sees 29...Nxb7 can be met by 30.Be5+.
29...Rd8 30.Bxd6+ Rxd6 31.Qxd6+ 1-0
29...Rd8 30.Bxd6+ Rxd6 31.Qxd6+ 1-0
The ending after 31...Qxd6 32.Rxd6 Kxd7 33.Rxb7 is hopeless for Black.
Friday, 25 September 2026
Hull 4NCL U2000 Round One
Steve Crow (1772 ECF/1763 Fide) - Spanton (1953 ECF/1997 Fide)
*****
*****
*****
*****
*****
*****
*****
*****
Sicilian Closed
1.e4 c5 2.Nf3 g6 3.d3!? Nc6 4.g3 Bg7 5.Bg2 d6 6.0-0 e6
This is a popular continuation, but even more common are 6...e5 and 6...Nf6.
7.Re1 Nge7 8.a4 0-0
 |
| Now both sides have castled, how would you assess the position? |
*****
*****
*****
*****
This exact setup only occurs 25 times in ChessBase's 2026 Mega database, but it is typical of the type of position that arises from the Closed Variation of the Sicilian, and is equal, according to Stockfish18 and Dragon1.
9.h4!?
This may be a novelty.
 |
| How should Black respond? |
*****
*****
*****
*****
9...e5!
Moving the e pawn a second time in the opening, but the point is that with White no longer able to play h3, the g4 square becomes an inviting post for Black's light-square bishop. The text is the engines' top choice.
10.h5!?
The engines prefer 10.c3, or developing the queen's knight, to either d2 or c3.
10...Bg4?!
Black has the upper hand after 10...gxh5, according to the engines.
11.h6 Bh8 12.c3 Qd7 13.Nbd2 Rad8 14.Qc2 f6!? 15.Nf1 d5 16.N3h2!?
The engines like 16.a5, claiming White has the upper hand.
16...Be6 17.Bd2 dxe4
The engines prefer 17...b6 or 17...d4.
18.dxe4 Na5 19.Nf3 Bb3 20.Qb1 b6
Not 20...Bxa4?? 21.Qxa2+ etc.
21.Ne3
 |
| Can Black now safely capture on a4? |
*****
*****
*****
*****
No.
21...Bxa4?
After 21...f5 or 21...c4 the game is completely equal, according to Dragon1, although Stockfish18 slightly prefers White.
22.b4 cxb4 23.cxb4 Nb3
 |
| How should White proceed? |
*****
*****
*****
*****
24.Qa2
Even stronger, according to the engines, is 24.Ra3!? Nxd2 25.Nxd2, when Black is still a pawn up, and with the bishop-pair, but the engines reckon White is winning, eg 25...Rfe8!? 26.Bf1!? (the simple 26.Qa2+ Kf8 27.Rxa4 Qxd2 28.Qxd2 Rxd2 29.Rxa7 is also good) Bc6 27.Bc4+ Kf8 28.b5 Ba8 29.Rd1, with massive pressure. If Black tries to hold on to the extra pawn by defending the light-square bishop with 25...b5, the engines reckon 26.Nb3 is very strong for White, and certainly Black's somewhat exposed king, and sidelined bishops, do not make a good impression.
24...b5 25.Rab1 Qe6 26.Rb2 Rd6?!
The engines much prefer 26...Qf7!?
27.Bc3
Even stronger seems to be the engines' 27.Nd5!?, eg 27...Nxd5 28.exd5 Rxd5 (28...Qxd5? 29.Nxe5) 29.Rxb3 Bxb3 30.Qxb3 Rd6 31.Qxe6+ Rxe6 32.Be3 leaves White with bishop and knight for rook and two pawns, and, more importantly, much the better coordination, including having the bishop-pair.
27...Rc8?
Black should probably play 27...Rd3 or 27...Rfd8, although the engines are unsure which move is better.
 |
| White to play and gain a large advantage |
*****
*****
*****
*****
28.Rc2?
Good is 28.Nd5 as 28...Nxd5 29.exd5 Rxd5 can be met by 30.Nd4, but best is 28.Bxe5! fxe5 29.Ng5, giving White an advantage worth about a minor piece, according to the engines.
28...Nd4!? 29.Qxe6+ Nxe6 30.Rcc1 Rdc6
 |
| Black has escaped the complications, while remaining a pawn up, but White apparently has sufficient compensation |
*****
*****
*****
*****
31.Nd5?
The engines give 31.Bb2 or 31.Bd2, eg 31.Bb2 Rxc1 32.Bxc1!? (they reckon it is important for White to keep a pair of rooks on the board) Kf7 33.Nd5, claiming complete equality.
31...Nxd5 32.exd5 Rxc3 33.Rxc3 Rxc3 34.dxe6
 |
| Black to play and win |
*****
*****
*****
*****
34...Kf8!?
Obvious enough, perhaps, but it is the only winning move, according to the engines.
35.Nd2
The engines suggest 35.g4!?, but have no good answer to 35...Ke7. Another possibility, pointed out by Dragon1, is 35.Bf1!?, when 35...Rxf3?? 36.Rc1 seems to allow White to draw, but again 35...Ke7 wins.
35...Ke7
35...f5 seems even better.
36.Ne4 Rc4 37.Rb1?!
White's best try is 37.Bf1!, when Black should reply 37...Rc7!, keeping control of the c file, as 37...Rxb4 38.Rc1!! Rxe4 39.Rc7+ Kxe6 40.Rxh7 may even win for White, according to the engines.
37...Kxe6?!
There is no rush for this, so better is the consolidating 37...a6, and the engines give as best 37...f5!?, eg 38.Ng5 e4 39.Nxh7 Bc2!? 40.Rc1 Bb2. But note that the tempting 37...Bc2?! runs into 38.Ra1!, with strong counterplay.
38.Nc5+ Ke7 39.Bh3?
This is easily met, and with a move Black wants to play anyway. However, after the engines' 39.Bd5 Rd4 40.Bg8, White more-or-less has full compensation for two pawns.
39...f5 40.Bf1 Rc3 41.Re1 e4 42.Nd3 Kd6 43.Nc5 Bd4
 |
| White is already two pawns down, and is losing more material imminently |
*****
*****
*****
*****
44.Rxe4?!
Desperation.
44...fxe4 45.Nxe4+ Ke5 46.Nxc3 Bxc3 47.f4+ Kf6 48.Bg2 Bb3 49.Bc6 Bc4 0-1
Chess Tip Of The Day 537
The conditions for a successful wing attack are either an advantage in the centre or at least a solid, albeit passive central position.
Luděk Pachman, Complete Chess Strategy 3: Play On The Wings
Subscribe to:
Posts (Atom)