Commentary: 2017 Isle of Man International, Round 9 (Tarjan - Kosteniuk)

This next commentary game contains some themes for improving players at several different levels of analysis - meta, strategic, and tactical. "Meta" in this case refers to the overall context - the fact that GM James Tarjan, one of the best US players in the 1970s, was at the time of this game in the third year of his chess career comeback and at age 65 defeated both GM Alexandra Kosteniuk (below) and super-GM Vladimir Kramnik during the 2017 Isle of Man International. This was no fluke, as he had also played for the US in the 2016 chess Olympiad. Seeing these kinds of examples helps combat the "inevitable decline" narrative associated with the aging process, or at least provides fewer excuses for not undertaking effortful study.

While we (or at least I) may not have Tarjan's level of inherent talent, his approach and the example of play here are understandable and instructive. My top observations from the game:

• Tarjan's opening choice is designed to allow White to "play chess" rather than debate opening theory. This strategy used to be frowned upon in general, with purists insisting White always play for a forced if slight advantage. However, Carlsen's repeated use of it over the years has lent it more legitimacy; one game of his is in a similar variation is included in the game notes.
• Master-level choice of candidate moves. I highlight multiple instances where White's move choice probably would not occur to an amateur. I find these to be one of the most important aspects of studying and analyzing master-level games, as they demonstrate how new ideas can be introduced into your own play.
• The interplay of tactical and strategic considerations is evident throughout, especially when Black - probably under time pressure - starts missing key tactics in the latter part of the game. Using tactics to achieve more of a strategic/positional advantage was also possible at several points in the game for both sides.

[Event "Chess.com Isle of Man Open - Masters"] [Site "Douglas (Isle of Man)"] [Date "2017.10.01"] [Round "9.32"] [White "Tarjan, James"] [Black "Kosteniuk, Alexandra"] [Result "1-0"] [ECO "A22"] [WhiteElo "2412"] [BlackElo "2552"] [Annotator "ChessAdmin/Komodo Dragon"] [PlyCount "99"] [EventDate "2017.09.23"] 1. c4 e5 2. g3 Nf6 3. Bg2 d5 4. cxd5 Nxd5 5. Nc3 Nb6 6. b3 {this move takes the game out of reversed Sicilian territory. Nf3 is the most played in the database.} Be7 {Black commits the bishop early. The following top-level game shows an alternate plan of development.} (6... Nc6 7. Bb2 Be6 8. Bxc6+ bxc6 9. Nf3 Bd6 10. d4 exd4 11. Qxd4 f6 12. Qe4 Kf7 13. Qxc6 Qe8 14. Qe4 h5 15. Nd4 Bd7 16. Qxe8+ Rhxe8 17. f3 c5 18. Nc2 Be5 19. Nd1 Bxb2 20. Nxb2 Bb5 21. e4 f5 22. Ne3 fxe4 23. f4 Red8 24. Rd1 Rd4 25. Nf5 Rad8 26. Nxd4 cxd4 27. a4 Ba6 28. a5 Nc8 29. Nc4 Ne7 30. Kf2 Nf5 31. Rhe1 e3+ 32. Kf3 Bb7+ 33. Ke2 Ke6 34. Rg1 Be4 35. Ne5 Rb8 36. Nf3 Bxf3+ 37. Kxf3 Rxb3 38. Ke4 e2 39. Rd3 Rb5 40. Re1 Rxa5 41. Rxe2 g6 42. Kf3+ Kd6 43. h3 Ra4 44. g4 hxg4+ 45. hxg4 Ne7 46. Ke4 Nc6 47. Rh2 Ra1 48. Rh6 Re1+ 49. Kf3 Rf1+ 50. Kg3 Rg1+ 51. Kh4 Rh1+ 52. Kg5 Rxh6 53. Kxh6 Kd5 54. Kxg6 Ke4 55. Rd1 Kxf4 56. g5 Ne5+ 57. Kh5 Ke4 58. Re1+ Kf5 59. Rf1+ Ke4 60. Re1+ Kf5 61. Kh6 Nf7+ 62. Kh5 Ne5 {1/2-1/2 (62) Carlsen,M (2863) -Nepomniachtchi,I (2784) Lichess.org INT 2020}) 7. Bb2 {the bishop must develop here and there is no reason to postpone it. It also gives White a look at Black's next developing move before making any other decisions about piece placement.} Nc6 (7... O-O {seems more consistent with the early Bishop move, also waiting to see where to go with the Nb8.}) 8. Bxc6+ $5 {Tarjan is the only one in the database to play this, sending the game on to a different strategic path. It is a classic trade-off decision, eliminating the strong White bishop but inflicting permanent structural damage on Black and providing White with relatively easy targets to work against. The engines evaluate the position as equal.} (8. Nf3 {is of course the conventional move.}) 8... bxc6 9. Nf3 {developing and attacking e5.} Qd6 (9... f6 {would be the more standard way to protect the pawn in similar Black structures.}) 10. O-O {tucking the king away before Black can play ...Bh3. Interestingly, the engines do not consider this such a problem.} (10. Rc1 $5 {immediately moving to increase pressure on the c-file.} Bh3 11. Qc2 $11 {White's king is in no danger and the rook can go to g1 and still be useful.}) 10... Bf5 $6 {this fights for the e4 square, but not very effectively, considering White's next move.} (10... Bh3) ( 10... O-O) 11. d3 $14 {now White is threatening Ne4, with a discovered attack on the e5 pawn.} Qe6 {avoiding a double attack on the queen after White's next. } 12. Ne4 {Black now has an important strategic decision to make. Her choice to castle queenside is very imbalancing and appears to go against the position's needs. Kosteniuk must have felt that either the king was safe enough there, despite the weak pawns, or that she would be able to attack White first.} O-O-O $6 {Black tactically protects e5, as the Ne4 is now hanging due to the pinned d-pawn. However, White gets out of the pin by moving his queen to a better square, lining up on the c-file, so it is a net minus for Black.} (12... f6 {is the engine recommendation, protecting the pawn.}) ( 12... Bxe4 $6 {looks like a reasonable idea, inflicting some pawn structure damage on White in return. However, Black's c-pawns are still worse than White's e-pawns and Black would no longer have the two bishops.}) 13. Qc2 $16 f6 14. Rfc1 {putting a rook on c1 is clearly a good idea, creating a battery with additional pressure. An argument could be made for moving the other rook, since the a-pawn will not need its support to advance, but White evidently wanted to put both heavy pieces on the queenside.} Kb7 {a reasonable-looking defensive move, but Black may be better off looking for some counterplay.} ( 14... Bg4 $5 {is the engines' preference, threatening to exchange on f3 and undermine White's d-pawn.}) 15. b4 $5 {a move that no amateur would be likely to consider, as it just appears to lose a pawn.} (15. a4 {is the more obvious way of launching operations on the queenside.}) 15... Bxb4 16. Rab1 {now we see the idea of the pawn sacrifice, to attack down the b-file. It appears to be slower than alternatives, though, as Black looks to equalize with ...Rd5 or exchange off an attacking piece with ...Bxe4. In practical terms, it is still difficult to meet, however.} (16. a4 Rd5 17. a5 Nd7 18. Bxe5 $5 fxe5) 16... Be7 {withdraws the bishop from its current vulnerable square, eliminating tactics involving a discovered attack after the Bb2 moves (for example Bd4 or Bxe5). This does little to impede White's attacking ideas, however.} (16... Rd5 { this defends along the 5th rank, although White still has pressure and tactical ideas in the center against the rook and bishop. For example} 17. Bd4 Ba3 18. Nc5+ Bxc5 19. Bxc5 e4 20. Nd4 Rxd4 {the defensive exchange sacrifice is necessary} (20... Qd7 $2 21. dxe4) 21. Bxd4 {at this point the engine shows complete equality, despite the material difference, meaning that with accurate defensive play Black will be all right. If Kosteniuk calculated this far, however, it would hardly look appetizing.}) (16... Bxe4 $5 {may be the best practical choice here, eliminating an attacking piece and reducing complexity.} 17. dxe4 a5 {physically impeding White's future a-pawn advance} 18. a3 Be7 $14) 17. a4 {with the idea of pushing away the knight. Bringing another piece into the attack might boost its effectiveness.} (17. Nfd2 {appears to be a useful preparatory move, as the Nf3 now is not doing much otherwise.}) (17. Bc3) 17... Ka8 {getting off the open file.} 18. a5 Nd7 {although Black has been pushed around, White does not have anything decisive and now looks to recover the pawn. However, there is no rush to do so and he ends up in a slightly worse position with the move played, so alternatives are worth looking at.} 19. Qa4 ( 19. Ra1 $5 {with the idea of Ba3 and trading off White's worse bishop, helping fight for the dark squares onthe queenside.}) (19. Qxc6+ Qxc6 20. Rxc6 { regains the pawn more forthrightly and seems to give Black less leeway than the game continuation.}) 19... Rb8 {challenging on the open file and giving back the pawn for equality.} (19... Nb8 $6 {is a move amateurs might play, defending the c6 pawn at all cost. This would renew White's attacking chances, however. For example} 20. Nfd2 Rd5 21. Ba3 Bxa3 22. Qxa3 Na6 23. Rc4 {followed by Rbc1.}) 20. Rxc6 {this forces Black's next move, otherwise the c-pawn is lost.} Qb3 21. Qxb3 Rxb3 22. Nfd2 {finally the knight gets into the action.} ( 22. Rxc7 $2 {greed is punished by} Rhb8 23. Rc2 Ba3 $19) 22... Rb7 {now Black is doing fine and the open b-file is more of a benefit for her than it is for her opponent.} 23. Rc2 {this overprotects the Bb2, freeing the Rb1 to move, but seems a bit passive.} (23. a6 $5 Rb6 24. Rxc7 Rhb8 25. Rc2 Rxa6 {with a slight advantage to Black, with the passed a-pawn.}) 23... Rhb8 24. Nc4 { overprotecting the Bb2 again and getting the knight further into play.} a6 { physically blocking the further advance of the White a-pawn.} 25. Rbc1 { moving the rook out of the pin and doubling pressure on the c-file. White has managed to rearrange his pieces to be less awkward and can look to exchange off his worse dark-square bishop with Ba3.} Rb3 {this temporarily stops the bishop exchange idea, but White could still insist on it.} 26. Bc3 (26. Ned2 R3b5 27. Ba3 Bxa3 28. Nxa3 Rxa5 29. Nac4 Rc5 30. Ra2 Rc6 31. Rca1 $11 {with play against the a-pawn.}) (26. f4 $5 {could also be played immediately.}) 26... Be6 {Black again passes up the chance to exchange on e4. This seems to help validate White's previous move, though, as the bishop is now more useful on the a5-e1 diagonal protecting the a-pawn.} (26... Bxe4 27. dxe4 Nc5) 27. f4 $6 {one of the common cases where playing the right idea at the wrong time - one tempo later, in this case - could result in a setback.} exf4 {Black makes the obvious move, to avoid losing the e-pawn (attacked three times, defended twice).} (27... Bxc4 {this possibility is the difference.} 28. dxc4 exf4 29. gxf4 Nc5 30. Ng3 Bd6 31. e3 $17 {now White has three weak pawns (a5, c4, e3) to defend and Black's pieces are much better placed.}) 28. gxf4 {this allows Black to play ...Bxc4 again, but she overlooks this.} (28. Ncd2 {is the engines' recommendation, sacrificing the pawn for vigorous piece play.}) 28... R3b5 29. Bd4 $11 {now White is completely equal.} Bb4 {attempting to pick up the a-pawn. Komodo recommends to simply ignore it, as White's rooks could then make threats on the c-file.} (29... Bxc4 30. Rxc4 $11) 30. Bc3 {an equal defensive move. Given the repetition of moves that occurs here, it appears both players were likely low on time.} (30. Ne3 $5 {this unleashes the rooks.} Bxa5 31. Rc6 Bb6 32. Bxb6 R8xb6 33. Rxc7 $16) 30... Be7 31. Bd4 R8b7 32. Kg2 { White appears to be just marking time here. Perhaps this succeeded in provoking Black's next, which is a non-obvious blunder.} Kb8 $2 {this appears to be a reasonable move, getting the king out of the corner and reinforcing the pawn on c7. It has a tactical problem, though, which Tarjan finds.} (32... Bd5 {or moving to f7 or g8 would avoid the problem, which is created by the king making itself vulnerable to a back-rank check, physically blocking the rook from coming back to b8.}) 33. Ncd6 $1 {an example of a reloader tactic on d6, made possible by the threat of Rc8+} Bxd6 (33... Rb1 {the engines suggest leaving the bishop on the board rather than exchange it, but it's still a win for White, who is up the exchange with no compensation for Black.} 34. Nxb7 Rxb7 (34... Rxc1 35. Rxc1 Kxb7 36. Nc5+ $18) 35. Kf2 $18) 34. Nxd6 {and now Black loses material.} Rd5 35. Nxb7 Rxd4 36. Nd8 Bd5+ 37. e4 Kc8 38. Nc6 Bxc6 39. Rxc6 Rxd3 40. Rxc7+ $18 {at the end of the sequence, White is simple an exchange up with a dominating rook pair.} Kd8 41. Ra7 {both getting behind the a-pawn and leaving the c7 square potentially available for the other rook.} Rd6 42. Kf3 g5 {attempting to get any counterplay possible, by breaking up the pawn shield in front of White's king, or getting a kingside pawn majority.} 43. Rg1 h6 {this looks like it is reinforcing the g-pawn, but ends up giving another pawn to White.} 44. h4 {another interesting master idea, although an amateur might have more of a chance of spotting this tactic. The h-pawn cannot be taken due to the mate, and Black's g-pawn cannot be further reinforced, so it is lost.} Ke7 45. hxg5 hxg5 46. fxg5 fxg5 47. Rxg5 Kf6 48. Kf4 Ke6 49. Rg6+ Nf6 {allowing one final tactic.} 50. Rxf6+ {and now the pawn forks on e5 after the rook is recaptured.} 1-0