Newmarket

AM playing at the East Anglian Chess Union congress, being held in the King Edward VII Memorial Hall, Newmarket.
Martin Walker (FM2179) - Spanton (1880)
Open Spanish
1.e4 e5 2.Nf3 Nc6 3.Bb5 Nf6 4.0-0 Nxe4 5.d4 a6!?
Heading for a Berlin Wall with 5...Nd6 is overwhelmingly more popular, but the text has been tried by quite a few strong players, including Carlsen and Ivanchuk. Also popular - more so than the text - is 5...Be7.
6.Ba4
White's commonest choice, according to ChessBase's 2019 Mega database, but also critical is 6.Bxc6, with Ivanchuk this time being on the White side along with Topalov.
6...b5 7.Bb3 d5 8.dxe5 Be6
Via a slight transposition, we have reached the main tabiya of the Open Spanish.
9.Nbd2 Nc5 10.c3 Be7
Winning the bishop-pair with 10...Nxb3!? is generally frowned upon here, because, I believe, White gets good control over the d4 square with 11.Nxb3! (rather than half-opening the a file with 11.axb3). Nevertheless, 10...Nxb3!? has been played by Short and Sokolov.
11.Bc2 0-0
Most popular is 11...Bg4, as seems to have been first played by Levenfish against Alekhine at Carlsbad 1911.
12.Re1
MW's suggestion of 12.Nb3 looked strong in the postmortem, but the text has been played by Anand.
12...d4
A thematic move in the Open Spanish when White does not take control of the d4 square.
13.Ne4?!
Probably not the best, although again it has been played by some strong players. Fellow Battersea club-member Aldo Camilleri played the superior-looking 13.cxd4 in a win against me in a Central London League match (Battersea 1 v Battersea 2) earlier this year.
13...dxc3 14.bxc3
14.Nxc5 looked better in the postmortem, but Stockfish10 and Komodo9 reckon the two moves are roughly equal, in both cases slightly preferring Black.
14...Qxd1 15.Rxd1

Black to make his 15th move

*****
*****
*****
*****
*****
15...Rad8?!
I thought it was important to rule out a pin on my c6 knight, but it seems even stronger was 15...Bf5. Black wins a pawn after Stockfish10's 16.Nf6+ Bxf6!, while Komodo9's 16.Re1 Rfd8 certainly looks nicer for Black.
16.Be3?!
A natural-looking move, but MW's postmortem suggestion of 16.Bg5 appears to keep any Black edge to a minimum. The engines give 16...Nxe4 17.Bxe4 h6 18.Bxe7 Nxe7 19.Nd4, when White is probably OK despite his isolated queenside pawns.
16...Rxd1+ 17.Rad1 Nxe4 18.Bxe4 Rd8
This may have been the move MW missed in his calculations.
19.Rxd8+
White offered a draw.
19...Nxd8 20.Bb1 c5
Played to prevent 21.Nd4, but now White wins the bishop-pair.
The engines give 20...Nc6! For quite some time, Stockfish10 reckons the move wins for Black, although it eventually calms down a bit (Komodo9 gives Black a slight edge). The strength of the move lies in that it prevents both 21.Nd4 and 21.Ng5.
21.Ng5
A complete equaliser. The game finished:
21...Bxg5 22.Bxg5 Nc6 23.Be3 Nxe5 (as MW pointed out, White is better after 23...c4? 24.f4) 24.Bxc5 Nc4 25.Bc2 g6 26.Kf1 Kg7 27.Ke2 Kf6 28.Bd4+ Ke7 29.a4 Kd6 30.axb5 axb5 31.Bd3 Kd5 32.Bxc4+ ½–½