KEG: Despite serious errors by both sides, this is a wonderful game to play over and analyze. Brody blundered in the opening and lost a pawn. Theoretically, he was lost. But Brody's swashbuckling play brought him back into contention in this contest thanks to some complacent play by Marco (whose Queen Knight was mired on b1 until move 26) and thanks to Brody's never-say-die attitude. However, Brody's wild style that lost him a pawn in the opening and then brought his game back to life ultimately was his undoing when he unleashed a "brilliant" combination that--while forcing Marco to give up the exchange to avoid a mating trap involving the sacrifice of Brody's Queen (!)--had a fatal flaw that Marco alertly spotted.
Brody's 7...0-0 in the opening was, as Marco correctly stated, a "very bad mistake" (Brody should have played 7...d6 as recommended by Marco and NOT Rosenthal's proposed 7...exd4 which loses to 8. e5).
After 7...0-0 8. Nxe5, Brody was probably already lost and should have tried 8...Nxe4. Marco says he would have responded 9. Bd5 but Brody might then have clawed his way back into the game with 9...Nf6. The real problem with 8...Nxe4 is that Marco could then have played 9. Bxf7+ or 9. Nxf7.
In any case, Brody's actual move, 8...Bb7 was hopeless. He was now down a pawn with a bad game. Had Marco played the simple 9. NxN and then, if 9...dxc6, 10. e5 (or if 9...BxN 10. d5) Brody would have had little chance.
But Marco got sloppy and Brody fought his way back into the game. Marco's 9. f3 was far inferior to 9. NxN, and after 9...Na5 Marco's 10. c3 (instead of 10. Nd3) forfeited most of his advantage, especially after Brody's excellent 10...c5.
Marco faults his 11. Bc2 and says he should have played 11. d5, but then after 11...Bd6 or 11...Qc7 Brody would have had counterplay well worth the lost pawn. Comparatively best at this stage for Marco would have been 11. dxc5, but his winning edge would still have been gone.
Marco is correct, however, that after 11. Bc2 Brody would have had a decent game despite his pawn minus with 11...d5. Brody's 11...d6 was bad, and after his 13...Qb6 he was probably lost again (he should have played 13...Rc8).
Perhaps recognizing he was lost, Brody went all out for the attack. His 14...Nc4 was a desperate bid for complications (14...Nh5 was perhaps theoretically best but practically hopeless). And after Marco's 16. e5, Brody might as well have sallied forth with the madcap 16...Ne4 as Fritz suggests. This loses a piece but the game gets wild, whereas after Brody's 16...Nd7 he was just dead lost.
Marco severely criticizes his 18. b4 (which allowed Brody to re-post his Knight at c4), but the move wasn't really all that bad. Had Marco followed up with 19. Nc3 he would have been comfortably up a pawn with the game well in hand. But Marco kept his b1 Knight dangerously buried with 19. Nc5.
Marco still probably had the game in hand, but he failed to notice the poison lurking in Brody's 20...Qh6.
The move that truly spoiled Marco's game was 21. a4. As Rosenthal demonstrates in the Tournament Book, Marco should have played--and probably still had a win with-- 21. Re1. By contrast, Marco's 21. g3 would have relinquished most of his advantage after 21...f6. (21. Nc3 was bad because of 21...Ne3, but it would not have been a losing move as Marco mistakenly claims so long as Marco had answered 22. Qc1 and not the awful 22. BxN that Marco considers that of course gets demolished after 22...QxB+).
Marco's 22. g3 was also weak (22. Re1 was best) as was his 23. e6 (23. exf6 as recommended by Rosenthal would have left the game about even).
After 23...f5 24. Re1 Rf6 25. f4, Brody could have gotten the upper hand with 25...g5. His 25...Raf8 was not best, but Brody was pursuing a fascinating but flawed plan as will be seen.
After 26. Nd2 Rxe6 (winning back the lost pawn) 27. Bb3 (27. a5 was better) Brody played the tricky 27...Kh8?! I agree with Rosenthal that 27...Na5 was better, and perhaps 27...Rf7 was the best move available, but Brody had an ingenuous idea that gives this game its interest. After 27...Kh8?1, the position was as follows:
Marco here played 28. NxN and after 28...dxc4 would have been mated had he played the seemingly obvious 29. Bxc4. Brody would then have played 28...Qxh2+ and after the forced 29. KxQ Brody would have had a mate in three beginning with 29...Rh6+.
What a wonderful conception. Just one problem. Marco played not 29. Bxc4 but the brilliant 29. d5!! Brody now won the exchange with 29...RxR+ 30. QxR Bf6 31. Bxc4 BxR 32. QxB. But now after 32...Re8 (32...Qg6 was best but also loses) 33. 33. Bd4 Qg6, Brody was up the exchange for a pawn but dead lost.