1.d4 Nf6 2.c4 c5 3.d5 b5 4.Nf3

4.Nf3 is the classic and most popular way of declining the Benko Gambit, although it is in breach of my favourite chess maxim, that the way to refute a gambit is to accept it. This game was played in a charity simul in support of a St. George's Hospital charity, and competitors were encouraged to attract sponsorship on the basis of the number of moves they could survive in the lists against the charity's GM champion. Whether or not Ray Keene (an admirable giver of his chess talents to worthy causes) in general played in a way to maximise our financial contributions is beyond my power to judge, but he certainly did not do so against me. None the less the thought does add an additional complexity to the strange world of the simul.

In the Benko Gambit accepted, 4.cxb5 is immediately followed by the offer of the rook's pawn with 4....a6 from which huge complications follow.

Over many years my usual response to 4.Nf3, declining the gambit, as indicated above, has been to enter the Blumenfeld Gambit with 4...e6, ignoring the sacrificed pawn (which White clearly hopes Black will grab) and counter-attacking in the centre. My plan this time, if my fluid hopes could so be dignified, was to get out of the book early and calculate through positions that would be broadly familiar to me because they are associated with the Benko Gambit accepted rather than the gambit declined, which Keene prefers. As John Fedorowicz puts it in The Complete Benko Gambit, White in the BGD "retains the option of accepting the gambit pawn under more favourable conditions". I wanted to encourage Raymond to accept the pawn, and lure him into "unfamiliar" territory. What a stupid thought, I soon realised.

4....d6 5.Nbd2 last book move 5....g6 6.e4

Allowing such an easy e4 is not a good idea, particularly as in the event I did not even get around to castling.

6....Nbd7 7.cxb5

With 7.cxb5, at last the gambit pawn drops and I look around for some standard Benko Gambit tactics, hoping to play in some order moves like a6, Qa5 or Bb7 as well as castling out of trouble.

7....Qa5 8.Bd3 a6 9.b6

I was more than a little shaken by 9.b6, which I had not been expecting although I have seen it in other lines, and which immediately put me on the wrong end of some fine tactics. Although 9....Nxb6 is playable, this is only because of the temporary pin and Black's position is now distinctly fragile.

9....Nxb6 10.0-0 Nbd7 11.Nc4 Qc7 12.Re1 Bg7 13.Bd2 Ng4

14.Ba5 Qb7 15.Rb1 Turning up the wick Nge5 16.Nfxe5 Nxe5

17.Nxe5 Bxe5 18.b4 Qa7 19.bxc5 dxc5 20.Bb6 Qb8

21.Bxc5+-

GM Keene by this move arrives nice and early at a clearly winning position, the annotation gratis Fritz, but also my thoughts at the time. He celebrates with a little show of stylised aggression, using his bishop like a miniature croquet hammer to knock the pawn off its square into his cupped hand. Fair enough. But at least I have arranged the previous exchange to give me a long shot at h2, a square "chunking theory" might suggest could lie outside the immediate "chunked" vision.

21....Bxh2+

[If 21....Qc7 22.Bd4 0-0 23.Bxe5 Qxe5 24.Qf3 with advantage.]

I think this sudden emergence of ostensible counter-play took him by surprise as he pondered for over a minute to decide between 22.Kf1 and 22.Kh1. As always in these circumstances vulture spectators appeared in order to investigate the reason for the delay, and some stayed around waiting for tasty entrails.

22.Kh1 Qf4 23.e5??

This move, played quickly, involves an error in calculation although it might qualify as "the kind of move we make in situations like this". The problem is that the rook which this clearance move allows into the defence is useless, since following 23....Qh4 the bishop can block its line at the same time as discovering check. Not a good decision, argues Fritz explaining the award of the double question mark, "as now the opponent is right back in the game". [Fritz offers 23.g3 Bxg3 24.fxg3 Qxg3 25.Re3 Qh4 26.Kg2 as a way of hanging on to much of the advantage.]

23....Qh4 24.Rb4??

Keene, however, perseveres with his original plan, apparently not realising that the light at the end of the tunnel has become an approaching express. Something of White's deteriorating position might still have been salvaged by 24.Qf3 Bxe5+ 25.Kg1 Qh2 26.Kf1

24....Bf4+ The corpse lashes out. 25.Kg1 Qh2+ 26.Kf1 Qh1+

27.Ke2 Bg4+ 28.f3 Qxg2+ 29.Bf2 Bxf3 Checkmate

The final position, which in my view would make a slightly naff but proudly wearable T shirt, is given below.

Actually, to be honest, I never got to play 29....Bxf3 as GM Raymond Keene bowed out with a smile and generous handshake just as I was reaching to pick up the bishop to play it. But "checkmate" is more rhetorically powerful than "White resigned", so I have rounded off the score sheet to reflect my preference. Raymond (this was his only loss) was an absolute sweetie in this simul defeat, presenting me with what he considered an appropriate signed book.

It was on the French Defence.