[Event "English version"]
[Site "?"]
[Date "2023.04.22"]
[Round "?"]
[White "Gert, Legemaat"]
[Black "Julian, Oskam"]
[Result "0-1"]
[ECO "C65"]
[Annotator "HenkBos"]
[PlyCount "96"]
[EventDate "2023.??.??"]
1. e4 e5 2. Nf3 Nc6 3. Bb5 Nf6 4. d3 Bc5 5. O-O Nd4 6. Nxd4 Bxd4 7. c3 Bb6 8.
Qe2 c6 9. Bc4 O-O 10. Bg5 h6 11. Bh4 d6 12. Kh1 g5 13. Bg3 Bg4 14. f3 Be6 15.
Nd2 Nh5 {The position is equal.} 16. Rfd1 $2 {But this is to slow. The
dreamed-of centre break is never going to happen.} Nxg3+ 17. hxg3 d5 $1 {
Black comes first!} 18. Bb3 Qf6 19. g4 Kg7 {A surprise attack along the h-file
from Julian. Although Gert has the half-open h-file, he has to lose valuable
tempi to mobilise his rooks.} 20. g3 $2 $19 {Wrong choice. Now it is going
from bad to worse.} (20. exd5 cxd5 21. Re1 {is the only way to keep the
position together...} Bc7 22. Nf1 $1 $17 {but black has a clear advantage.})
20... Rh8 21. Kg2 h5 22. gxh5 Bh3+ $1 {[#] Wow! A nice but necessary blow!} 23.
Kxh3 Rxh5+ 24. Kg2 Rah8 25. Rh1 Qh6 $1 {Triplication at the h-file.} 26. Qe1
Rh2+ 27. Rxh2 Qxh2+ 28. Kf1 Qh1+ 29. Ke2 Qg2+ 30. Kd1 Rh1 31. Kc2 Rxe1 32. Rxe1
{[#] The smoke has cleared, but the ruin is still smouldering. The position
should be lost for Gert.} d4 {A good move, but....., (spoiler alert!) I think
Julian will have regretted afterwards that he didn't just capture the pawn on
g3} 33. g4 Ba5 34. a4 Qf2 35. Rb1 a6 36. Ba2 $5 {Sneaky! Gert plays his last
practical trump card.} Bb6 $6 ({Well..., should I give this move a question
mark? Stockfish still rates the position -3. But} 36... dxc3 $1 37. bxc3 b5 38.
axb5 cxb5 $18 {is better. The bishop now can go to b6 and more important: the
b-file is closed.}) 37. c4 $1 {Smart move. then defend from behind the
battlements. The white strategy is getting clearer....} c5 {Computer says....:
"Best move".} 38. b4 $1 cxb4 39. Bb3 $3 {And Gert has a fortress. A fortress
is a spectacular phenomenon in chess literature. A fortress exists when the
King of the party at a material disadvantage cannot be approached. It is
actually a kind of dynamic stalemate in which ultimately the moves can only be
repeated. A famous quote by Carlsen in 2016: ‘I don't believe in
fortresses!’ By this he meant to say that although they do exist, most
fortress-like positions can still be demolished and thus are pseudo-fortresses.
The (pseudo)fortress also has played an interesting role in the development of
computer chess. Even the strongest chess programmes have, until recently, had
a lot of trouble valuating interlocking pawn structures. With
fortress-positions, this is often the case. If there really was a fortress,
until recently, the computer gave the same valuation in all cases. However,
that score was not 0.00, but for example -5.02 for the side that is at a
material disadvantage. With the arrival of the latest generation of chess
programs (A.I.), this shortcoming has been remedied. A real fortress is valued
at 0.00: in other words, a draw. A pseudo-fortress, to which Carlsen refers,
is valued with, say, -5.02. I do not call this valuation arbitrary; Stockfish
16.1 values the diagram position in this game on this. Too bad for Gert and
pleasant for Julian. But how to demolish this pseudo fortress? There are only
two strategies here. First, Julian can play on the breakthrough b5. But try
getting inside his head. The white bishop is promoted from pawn to bishop with
a panoramic view, the a-line opens up for the white Rook, he gets a weak pawn
on f7, the black (unequal!) Bishop is a primal bad defender and, on top of
that, the white Knight gets a beautiful square on c4. White is absolutely not
inferior in most variants. It is understandable that he did not dare to do so.
that leaves only a rammei. But that one looks even more unlikely by human
standards. And what does Stockfish think....? Even now he/she is struggling
to make it clear to us what the fastest win for black is. After rattling
around for a while, in all the variants after 39. Lb3!!! the move f5!!!
appears somewhere. And also, following on from this the black moves g4, Ld8
and Lg5 with eventually at an appropriate moment the exchange on d2. This is
the only method to penetrate the white position. In the grandmasterly,
virtually digital world, Carlsen is right: ‘I don't believe in
fortresses!’. In that world, it continues with the ‘Hidden
Move’‘’:} f5 $3 {A beautiful Hidden Move!} ({In the analogue, ordinary
real world, it continued} 39... Ba5 40. Ra1 Bd8 41. Rh1 Qg2 42. Re1 Qh2 43. Rd1
{Here, Gert offered and Julian accepted the offer. Good for Gert, bad for
Julian. In this world Gert is right: "I believe in fortresses!"}) 40. gxf5 (40.
exf5 $2 Qe2 41. Rh1 e4 42. fxe4 Bc7 43. Rd1 Qxg4 $18 {de vrijpion is te sterk.}
) 40... Bd8 41. Rf1 Qh2 42. Rd1 g4 43. fxg4 Bg5 44. Kc1 {White tries to
maintain his fortress.} Qe2 45. Kc2 Bxd2 {The demolishing starts} 46. Rxd2 Qe1
$1 {and black invades at a1 or c3} 47. Ba2 (47. c5 Qa1 {Threathening Dc3} 48.
Bd5 b3+ $1 {and white loses material.}) 47... Qa1 48. Bb3 Qc3+ {[#] A blasting
Royal Visit by the black Queen. It is mate in 18 moves.} 0-1