The very strong tournament finished Thursday. A wrap-up by World Chess’s columnist.

The Reykjavik Open in Iceland ended Thursday and Anish Giri of the Netherlands, the top seed, finished in clear first with 8½ points out of 10. He raced out of the gate with victories in his first four games, including what turned out to be a crucial fourth-round victory over his young countryman, Jorden Van Foreest. Van Foreest finished tied for second, half a point behind Giri, so the tournament might have had a very different outcome had he managed to draw.

Giri, Anish vs. Van Foreest, Jorden
Reykjavik Open 2017 | Reykjavik ISL | Round 4.1 | 21 Apr 2017 | 1-0
1. d4 d5 2. c4 c6 3. Nf3 Nf6 4. e3 g6 The Schlechter Slav is solid to the point of passivity. Against tactically impatient players it's not a bad choice, but it's questionable strategy against a player who loves to play for 6-7 hours hoping to grind out the full point - a player like Giri, in other words.
5. Nc3 Bg7 6. Be2 O-O 7. O-O e6
7... dxc4 8. Bxc4 Bg4 is a livelier option, while  )
7... Bg4 8. cxd5 cxd5 9. Qb3 b6 10. h3 probably counts as the absolute main line at the moment. Here Black has done reasonably well with both 10...Bxf3 and 10...Bc8.  )
8. b3 Re8 9. Bb2 Nbd7 10. Qc2 b6 11. Rad1 Bb7 12. Ne5 Qc7 13. f4 Rad8
13... Red8!? 14. g4 Ne8  )
14. Bf3
14. g4!  )
14... c5
14... b5!?  )
15. g4! This direct approach is surprisingly awkward for Black.
15... cxd4 16. exd4 a6 17. Qg2 Nf8 18. f5? Objectively bad, but it's not surprising that it worked.
18. h4  )
18... exf5 19. gxf5 Qd6?
19... gxf5 may have been hard for Black to stomach, but it was by far his best move. White has no way to exploit the g-file in the near future, and his advantage has diminished.  )
20. fxg6 Nxg6 21. Nxd5 Nh4 22. Nxf6+!?
22. Qg5 is a good alternative.  )
22... Qxf6 23. Bxb7! Nxg2 24. Rxf6 Bxf6 25. Ng4 Bg7 26. Bxg2 Re2 27. Ba1 Rxa2 Two minor pieces are generally more than a match for a rook and a pawn, and here Black doesn't even have the pawn for compensation. It will take time for White to win, but Giri is up to the challenge.
28. Ne3 Rd6 29. Bd5 Rg6+ 30. Kh1 Rh6 31. Ng2 Rh3 32. Rf1 Kh8 33. Re1 Kg8 34. Rf1 Kh8 35. c5 bxc5 36. b4! Rd2 37. bxc5 Bxd4 38. Bxd4+ Rxd4 39. Bxf7 Rc3 40. Nf4 Rd8
40... Rxc5? 41. Ne6  )
41. Ne6 Rc8 42. Rg1! h6 43. Rg6 Kh7 44. Rg7+ Kh8 45. Rg6 Kh7 46. Rg7+ Kh8 47. Bg6! Rb8 48. Rh7+ Kg8 49. Ra7 Rc1+ Black would like to use his rooks to create a mating net, e.g. with
49... Rb2 , but White's mate comes first.
50. Ra8+ Rb8 51. Rxb8#  )
50. Kg2 Rc3! But now Black does threaten a lethal check on b2.
51. Kg1 Rc1+ 52. Kf2 a5 53. Bh7+! Kh8 54. Nf4 The threat of Ng6# wins the exchange.
54... Rf8 55. Ke3 Rxf4 56. Kxf4 Rxc5 57. Bf5 Kg8 58. Ra6 a4 59. Rxa4 Kg7 60. Ra7+ Kg8 61. Ra6 Rc7 62. Rxh6 If White were missing the h-pawn it would be drawn, and if both sides' rooks were gone it would also be drawn. This, however, is not a draw.
62... Kg7 63. Kg5 Rf7 64. Ra6 Rc7 65. h3 Re7 66. Rg6+ Kh8 67. Rh6+ Kg8 68. Rh4 Rg7+ 69. Bg6 Ra7 70. Kh6 Kh8 71. Rd4 Rh7+ 72. Kg5 Ra7 Not all rook and bishop vs. rook positions are drawn. Black can reach that ending by taking on h3, but his king is in a mating net.
72... Rxh3 73. Kf6 Rf3+ 74. Bf5 Rg3 75. Rh4+ Kg8 76. Be6+ Kf8 77. Rh8+ Rg8 78. Rxg8#  )
73. Rd8+ Kg7 74. Bf5 Rf7 75. Be6 Re7 76. Kf5 Ra7 77. Rg8+ Kh7 78. Kf6 Rb7 79. Ra8 Threatening 80.Bf5+ Kh6 81.Rh8+ and mate next move.
79... Rb6 80. Ra4 White can mate in at most two moves unless Black plays 80...Rxe6+, which is just as hopeless.

After that game, Giri drew his next three and fell slightly off the pace. He won in Round 8 to tie for first, then he beat Baadur Jobava of Georgia in Round 9, giving him a half a point lead heading into the last round.

Jobava, Baadur vs. Giri, Anish
Reykjavik Open 2017 | Reykjavik ISL | Round 9.1 | 26 Apr 2017 | 0-1
1. e4 c5 2. Nf3 d6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 a6 6. Bg5 e6 7. f4 Qb6 The biggest test for the 6.Bg5 system, either here or after a preliminary 7... h6 8.Bh4.
8. Nb3 The safe continuation, which is ruled out by the preliminary 7...h6 as Black could play ...Qe3+, winning either the e- or the f-pawn. Jobava is not exactly known as a safe player, so one might wonder why he's choosing 8.Nb3. A likely answer is that the sharp variations have been analyzed incredibly deeply, and Giri is the sort of player who knows all the analysis and then some. Unless Jobava has something new to add to existing theory that favors White, it's a better practical choice, if he's trying to win (and he is) to head for a position where both sides must play on their own.
8... Nbd7 9. Be2
9. Qf3 is the main move, and  )
9. Qe2 is another reasonably popular move.  )
9... Be7 10. Qd3 Qc7 11. O-O
11. Bf3 is the main move, but not a particularly successful one. Here's a recent high-level email game won by Black:
11... Rb8 12. O-O-O b5 13. Kb1 O-O 14. Ne2 Bb7 15. g4 b4 16. h4 a5 17. Bxf6 Nxf6 18. g5 Nd7 19. Nbd4 a4 20. Rhg1 Ne5 21. fxe5 dxe5 22. Nxe6 fxe6 23. Bg4 Qb6 24. Qd7 Bc5 25. Rgf1 Bxe4 26. Rxf8+ Rxf8 27. Qxa4 Kh8 28. Qb3 Bd5 29. c4 Be4+ 30. Ka1 Bf5 31. Bh3 Qc7 32. Rf1 g6 33. Bxf5 exf5 34. Kb1 Qc6 35. Qf3 Qxf3 36. Rxf3 Kg7 37. a4 bxa3 38. bxa3 Kf7 39. Rb3 f4 40. Nc3 f3 41. Ne4 f2 0-1 (41) Brookes,J (2511)-Robson,N (2602) ICCF email 2014  )
11... b5 12. a3 Bb7 13. Kh1 Rc8 14. Rae1 O-O 15. Qh3 Rfe8 16. Bd3 With the obvious intention of playing e4-e5. Black's reply is just as obvious.
16... e5
16... h6!? is worth considering as well. Just because 16...e5 is "natural' and obvious doesn't guarantee that it is best.  )
17. fxe5 The immediate
17. Re3 was worth considering, while if White wanted to get the Bd3 into action he could try  )
17. Nd5 . Black is nevertheless fine after
17... Bxd5 18. exd5 e4 19. Bxe4 Nxe4 20. Rxe4 Qxc2! 21. Qe3 Nb6! Black's position relies on tricks - which is fine as long as they work. And they do!
22. Nd4 Nxd5! 23. Nxc2 Nxe3 24. Rxe3 f6! 25. Nb4 fxg5 26. Nd5 Bf6 27. Nxf6+ gxf6 28. Rg3 Rc5 29. h4 White will escape with a draw.  )
17... dxe5 18. Re3 Nf8 19. Ne2?! White understandably tries to migrate his pieces to the kingside for an attack, but Black's defensive resources suffice and then some.
19. Bxf6 Bxf6 20. Rxf6 This sacrifice "suggests itself", as the cliche puts it, but it's only enough for a (likely) draw.
20... gxf6 21. Qh6 Qd6 22. Nd2 Ng6 23. Rh3 Rxc3 24. bxc3 Qxa3 25. Qxh7+ Kf8 26. Nf1 Qc1 27. Rf3 Ke7  )
19... Qd7! 20. Rf5 The queen wants to stay on h3, and of course White wants to avoid exchanging queens. Unfortunately, the rook now sits where White hoped his knight would go in a couple of moves.
20... Qe6 Black is in great shape, but for a while Jobava handles this middlegame better than his elite opponent.
21. Nd2 N6d7
21... g6  )
22. Bxe7 Rxe7 23. Nb3 g6 24. Qg3 Nf6 25. Rf1 N8d7 Black is still better, though he hasn't made the most of his opportunities.
26. Qe1 Kg7 27. Ng3
27. Qb4  )
27. Nc3  )
27... h5 28. Ne2 Rh8 29. h3 Qb6 30. Ng1 h4 31. Nf3
31. Ref3  )
31... Ree8 32. Kh2 Rh5 33. Re2
33. Na5  )
33... Reh8 34. Qf2? A mistake. With queens off the board Black can throw everything at White's king without taking any risk.
34... Qxf2 35. Rfxf2 g5 36. Rf1 g4 37. Nfd2 Rg5 38. c4 Nh5! 39. cxb5 axb5 40. Bxb5 Ndf6 41. Nc5 Bc8 Even without a queen Black's attack is very dangerous, and maybe even winning.
42. Ree1
42. hxg4 Nxg4+ 43. Kg1 Nf4 44. Rf3 may be White's best, though he's still probably losing after
44... Rd8 45. Nf1 Nxe2+ 46. Bxe2 Nf6  )
42... Nf4
42... gxh3  )
43. hxg4 Nxg4+ 44. Kh1 Nf6! 45. Nd3 Rg3
45... Nxd3  )
46. Rf3
46. Rxf4 was forced, though it may not be good enough after
46... h3! 47. Rf3 hxg2+ 48. Kg1 Rh1+ 49. Kf2 g1=Q+ 50. Rxg1 Rgxg1 51. Nxe5 Rg5  )
46... Bg4 47. Rfe3 Rxg2
47... Bd7! 48. Bxd7 Nxd3!  )
48. Nc4 Rc2 49. Ndxe5?? Now White is not just the likely loser; it's guaranteed after Black's next move.
49... N6h5! White will be mated or lose the house.
50. Kg1
50. Nxg4 Ng3+ 51. Rxg3 hxg3+ 52. Kg1 Rg2+ 53. Kf1 Rh1#  )
50... Ng3 51. Rxg3 hxg3 52. Ne3 Nh3+ It's mate next move.

Had Giri drawn in the last round he’d have finished in a five-way tie for first. But, facing another countryman, Erwin L’Ami, Giri closed the tournament with an exclamation point, winning a short and brutal game.

Giri, Anish vs. L'Ami, Erwin
Reykjavik Open 2017 | Reykjavik ISL | Round 10.1 | 27 Apr 2017 | 1-0
1. Nf3 Nf6 2. d4 e6 3. g3 b5 4. Bg5 Bb7 5. Nbd2 d5
5... c5 and  )
5... Be7 are more usual.  )
6. e3! Combining g3 and e3 is often a poor idea because of the light squared weaknesses it creates. It works here though, because Black must make positional concessions to guard the b-pawn.
6... a6 7. a4 b4 8. Nb3 Nbd7
8... a5 was probably better, despite the weakening of the b5 square.  )
9. Na5 Qc8 10. Bxf6!?
10. Bg2 Ne4 11. Bf4 is a little better for White, thanks to his better coordination.  )
10... Nxf6 11. c4 dxc4?!
11... bxc3 12. bxc3 c5 13. Rb1 Bc6 14. Nxc6 Qxc6 15. a5 Bd6 was a better choice, probably sufficient for equality.  )
12. Nxb7 Qxb7 13. Bg2 c3 14. bxc3 bxc3 15. O-O Bb4 16. Ne5 Nd5 17. a5! Black's position doesn't look that bad. He has an extra, passed pawn, and the only way to kick the knight, with e3-e4, simultaneously blocks the long diagonal. Maybe L'Ami saw this position back when he was thinking about 11...dxc4, and thought everything was fine. As it turns out, however, he's just about lost.
17... Rb8?
17... Rc8  )
17... c5  )
18. Qa4+ Qb5 19. Bxd5?! The immediate
19. Nc6 was even better - there's no rush to take on d5 immediately.  )
19... exd5 20. Nc6 O-O 21. Rfb1 White's best is the odd-looking
21. Rfc1! , which is testament to how tied down Black is along the b-file.
21... Qc4 22. Rab1 Ba3 23. Qxa3 Qxc6 24. Rb3! Rxb3 25. Qxb3 Qb5 26. Qxb5 axb5 27. Rxc3 Rc8 28. a6 b4 29. Rb3 Kf8 30. Rxb4 Ke7 31. a7 Ra8 32. Ra4 Kd7 33. Kg2 Kc6 34. Kf3 Kb6 35. Rb4+ Kc6 36. Rb8 This variation, though excessively long, nicely illustrates Black's many long-term problems.  )
21... Qc4?
21... Qxa4 22. Rxa4 Rb5 23. Raxb4 Rxb4 24. Nxb4 Rb8 25. Rb3 c2 26. Rc3 Rxb4 27. Rxc2 Rb5 28. Rxc7 Kf8 29. g4 Rxa5 I would not claim that Black should draw, but he'll still have chances to survive with half a point.  )
22. Rxb4 Rxb4 23. Nxb4 The rest is simple. White has no back rank problem to worry about, so there aren't any tricks to watch out for.
23... Rb8 24. Nc2 Qe2 25. Ne1 h6 26. Qc2 Qc4 27. Nd3 Rb3 28. Kg2 g6 29. Nc5 Rb2 30. Qd3

Van Foreest had the best tiebreaks of the four players tied for second, and his victories in the last two rounds are worth examining. First was a miniature against Awonder Liang of the United States, one of the country’s top prodigies. Liang, who turned 14 earlier this month, is already an international master and is close to achieving the grandmaster title.

Van Foreest, Jorden vs. Liang, Awonder
Reykjavik Open 2017 | Reykjavik ISL | Round 9.10 | 26 Apr 2017 | 1-0
1. e4 c6 2. d4 d5 3. e5 Currently the bane of the Caro-Kann player's existence.
3... Bf5 4. Nf3 e6 5. Be2 c5
5... Nd7 and  )
5... Ne7 are major alternatives.  )
6. O-O
6. Be3 cxd4 7. Nxd4 Ne7 8. O-O Nbc6 9. Bb5 a6 10. Bxc6+ is another main line, and here both recaptures are possible.  )
6... Nc6 7. c3 Bg6 8. Be3
8. a3 is another move, intending to continue b2-b4 to close lines on the queenside.  )
8... Qb6
8... cxd4 9. cxd4 Nge7 10. Nh4 Nf5 11. Nxf5 Bxf5 12. Nc3 Be7 is an important alternative.  )
9. Na3 Time will tell whether this is an important novelty or not. Practically, it worked like a dream, as Black replied with an outright blunder.
9... c4?? 10. Nxc4! dxc4 11. d5 Bc5 12. Bxc5 Qxc5 13. dxc6 Qxc6?!
13... bxc6 leaves Black with a terrible structure and a poor position after
14. Qd4 /+- , yet this was still his best option.  )
14. Nd4 Qa6 15. a4! Qb6 16. a5! Qxb2? 17. Qa4+ Kf8 18. Qxc4 Threatening Ra2.
18... Qd2 19. Qb4+ Ne7 20. Rfd1 Qh6 21. Qxb7 Re8 22. Bb5 White is winning serious material, while Black is left with a wretched position.

To finish in second (and to have a chance to tie for first) Van Foreest needed to defeat Vidit Santosh Gujrathi, a strong Indian grandmaster. Van Foreest, who had Black, prevailed after Vidit got carried away with dreams of a kingside attack.

Vidit, Santosh Gujrathi vs. Van Foreest, Jorden
Reykjavik Open 2017 | Reykjavik ISL | Round 10.4 | 27 Apr 2017 | 0-1
1. Nf3 d5 2. g3 g6 3. Bg2 Bg7 4. d4 Nd7 5. a4 a5 6. O-O Ngf6 7. Na3 O-O 8. c4 c6 9. Bf4 dxc4 10. Nxc4 Nb6 11. Nce5 Ne4 12. Qc1 Nd6 13. Re1 Be6 14. e4 Ndc4 15. Qc3
15. Nxc4  )
15... Nxe5 16. Bxe5 Bxe5 17. dxe5 The opening went extremely well for White, but it looks like he's too hungry for a kingside attack. Chasing this end, he surrenders too many other goods.
17. Nxe5  )
17... Qc7 18. Ng5 Nd7 19. f4 Qb6+ 20. Kh1 Qb4
20... Nc5  )
21. Qf3?
21. Rec1  )
21... Rad8 22. f5 Nxe5 23. Qf4 Nd3 24. Qh4 h5 25. Rf1 Bc4 26. g4 Qxb2! 27. gxh5 gxf5 28. Rab1 Qd4 Black's pieces dominate the center. White has a good number of pieces near Black's king, but his pawn on h5 prevents them from achieving anything.
29. Rxf5 f6 30. Nh3
30. Nf3 Qxe4! 31. Qxe4 Nf2+ 32. Kg1 Nxe4  )
30... b5 31. axb5 cxb5 32. Nf2 Nxf2+ 33. Rxf2 Kh8 34. Qf4 Rf7 White has no threats, and no way to create any threats.
35. Bh3 Rg7
35... Qd3!  )
36. Bf5 Qd1+! 37. Rf1 Qd2 38. Qxd2 Rxd2 39. Rfc1? Rgg2 Avoiding the mate (...Rxh2+ Kg1 Rdg2#) will cost White at least an exchange, and he'll still have fatal difficulties coping with Black's queenside passers.

Two others in the tie for second, Abhijeet Gupta, another Indian grandmaster, and Sergei Movsesian, a grandmaster from Armenia, faced off in Round 7. Had Movsesian drawn, he might have been in the race for first at the end.

Gupta, Abhijeet vs. Movsesian, Sergei
Reykjavik Open 2017 | Reykjavik ISL | Round 7.4 | 24 Apr 2017 | 1-0
1. d4 Nf6 2. c4 e6 3. Nf3 d5 4. g3 Be7 5. Bg2 O-O 6. O-O dxc4 7. Qc2 a6 8. a4 Bd7 9. Qxc4 Bc6 10. Bf4 Bd6 11. Qc1 Nbd7 12. Nc3 Qe7 13. a5 e5 14. dxe5 Nxe5 15. Nd4 Bxg2 16. Kxg2 Ng6 17. Nf5 Qe6 18. Nxd6 cxd6 19. Rd1 Ne4 20. Nd5 Rac8 21. Qe3 Rfe8 22. f3 Nf6 23. Qxe6 Rxe6 24. Nc3 Nxf4+ 25. gxf4 d5 26. Ra4 Kf8 27. Rb4 Re7 28. e4 dxe4 29. fxe4 Ne8 30. Kf3 f6 31. Rb6 g5 32. f5 h5 33. Rd4 Rf7 34. Ke3 Kg7 35. Rdb4 Rcc7 36. Re6 Rc8 37. Reb6 Rcc7 38. Re6 Rc8 39. h3 Rd8 40. Reb6 Nd6 41. Kf3 Rdd7 42. Ke2 Nc8 43. Re6 Ne7 44. Ra4 Nc6 45. Nd5 Rf8 46. Ke3 h4 47. Ra1 Nd8 48. Rc1 Nxe6 49. fxe6 Rd6 50. Rc7+ Kh6 51. e7 An exciting up-and-down battle has reached this critical position. If Black plays correctly, the game should end in a draw; if not, White's more active army will prevail.
51... Rh8?
51... Re8! 52. Rxb7 g4 53. Rb6 Rxd5! 54. exd5 Rxe7+ This is the critical difference with 51...Rh8. Black buys some time and gets rid of White's most dangerous pawn.
55. Kf2 Re5 56. hxg4 Rxd5 57. Rxf6+ Kg5 58. Rxa6 Rb5  )
52. Rxb7 Kg6
52... g4 53. Rb6 Now
53... Rxd5 makes little sense, as White is completely winning after the recapture.
...  54. exd5  )
53. Rb6 Rxb6 54. axb6 Kf7 55. Nc7! Kxe7 56. Nxa6! Guaranteeing that the b-pawn will cost Black his rook.
56... g4! 57. b7!
57. hxg4?? h3  )
57. Kf2? g3+ 58. Kg2 Rc8! 59. b7 Rc2+ Black will keep checking, as White's king cannot go to the third rank or come close to the rook, as ...g2 will promote.  )
57... gxh3 58. Kf3 Kd6 59. b8=Q+ Rxb8 60. Nxb8 h2 61. Kg2 Ke5 At a glance Black's drawing chances look very good. He's taking the e-pawn next, and White's knight seems too far away from the b-pawn. Further, if it defends the pawn from a4 or c4, Black's king on b3 will win either it or the pawn - drawing in either case.
62. Nd7+ Kxe4 63. Nxf6+ Kd4 Draw?
64. Ng4 Kc4 65. Ne3+ No: White's knight reaches d1, and there it is safe from Black's king - if it ever goes to c2, White plays b4 and the pawn promotes. Otherwise, Black's king must wait, and White wins easily by taking the h-pawns with his king and then bringing it back to the queenside. Black therefore resigns.

The last member of the quartet of runners-up was Gata Kamsky of the United States, a former challenger for the World Championship. At his peak, Kamsky defeated many strong players, even other world championship contenders, by creating small problems for hour after hour until they snapped. That’s just what happened to Emre Can, a Turkish grandmaster, in Round 10. He came within a move of drawing by the 50-move rule on his way to a 121-move loss.

Kamsky, Gata vs. Can, Emre
Reykjavik Open 2017 | Reykjavik ISL | Round 10.5 | 27 Apr 2017 | 1-0
1. e4 My apologies to anyone who wanted in depth analysis of this 121-move monster. This beast will stand as its own monument of the player Gata Kamsky is, and especially the player he was in his heyday. Decades before Magnus Carlsen came on the scene, Kamsky was grinding out tiny advantages against even elite opposition, just as he does here against a strong grandmaster.
1... c5 2. Nf3 d6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 a6 6. a4 e5 7. Nf3 Be7 8. Bc4 Be6 9. Ba2 Nc6 10. O-O O-O 11. Bg5 Nb4 12. Bb3 Rc8 13. Qe2 Qc7 14. Rfd1 Rfd8 15. Rd2 h6 16. Bxf6 Bxf6 17. Rad1 Be7 18. h3 Qb6 19. h4 Kh8 20. g3 Bg4 21. Re1 Rf8 22. Qe3 Qxe3 23. Rxe3 Be6 24. Kg2 g5 25. Bxe6 fxe6 26. hxg5 hxg5 27. Nh2 Rc5 28. Ng4 b5 29. axb5 axb5 30. Re1 Kg7 31. Ra1 Nc6 32. Ne2 b4 33. Ra6 Rc8 34. Ne3 Na5 35. b3 R8c6 36. Ra7 Rc7 37. Rxc7 Rxc7 38. Rd1 Rc5 39. Nc1 Nc6 40. Nd3 Ra5 41. Rc1 d5 42. c3 bxc3 43. Rxc3 dxe4 44. Nb2 Rc5 45. Na4 Rxc3 46. Nxc3 Nd4 47. Nxe4 Nxb3 48. Nc4 Kg6 49. Kf3 Kf5 50. g4+ Kg6 51. Nxe5+ The engine assures us that the position is equal. This is true in the sense that the ending must be a draw. There's no doubt that both Kamsky and Can realized this, and it's also a certainty that they both knew that the other player knew that it was a draw. To say that it is a drawn position does not mean that it's an easy draw, and White has several practical advantages in his favor. The first and most obvious advantage is that Black's pawns are split, and therefore cannot offer each other any support. Second, White has two knights against a bishop and knight. This material balance often favors the side with the bishop, but with all the pawns on the same side of the board the knight's ability to move to squares of both color is more important than the bishop's long-range potentiality, as all the action will occur in a small space.
51... Kg7 52. Nc6 Bf6 53. Ke3 Kg6 54. f3 Start counting for the 50-move rule.
54... Kg7 55. Kd3 Nc1+ 56. Kd2 Nb3+ 57. Kc2 Na1+ 58. Kd3 Nb3 59. Nb8 Kg6 60. Nd7 Bg7 61. Nec5 Nd4 62. Ke3 Nc2+ 63. Ke4 Nd4 64. Nf8+ Kf7 65. Nfd7 Ke7 66. Nb6 Kd6 67. Nd3 Nb5 68. Nc4+ Ke7 69. Nf2 Kf7 70. Kd3 Nd4 71. Ke4 Nb5 72. Nh3 Kg6 73. Ne5+ Kf6 74. Nd7+ Kg6 75. Ne5+ Kf6 76. Nc6 Bf8 77. Kd3 Bg7 78. Nf2 Kg6 79. Ne4 Nc7 80. Kc4 Nd5 81. Kc5 Bf8+ 82. Kd4 Bg7+ 83. Ne5+ Kh6 84. Nc5 Nf4 85. Ke4 Bf8 86. Ncd7 Kg7 87. Nd3 Ng6
87... Nxd3 88. Kxd3 Bd6 was drawing, and possibly a more practical option. Maybe Can felt that he was close enough to drawing by means of the 50-move rule that he didn't want to start all over.  )
88. N7e5 Nh4 89. Nc6 Bd6 90. Nd4 Kf6 91. Nb5 Bf8 92. Ne5 Bb4 93. Nc7 Ba5 94. Na6 Bc3 95. Nc6 Ng6 96. Kd3 Be1 97. Nc5 Nf4+ 98. Kc4 Bh4 99. Ne4+ Kg6 100. Kc5 Kh6
100... Kf7 101. Kd6 Ke8 102. Nc5 Kf7 103. Nd8+ Kf6 104. Ndxe6 also collects the pawn at the last possible moment, though here Black can swap a pair of knights. This makes his defense easier, though he may not have realized it at the time.  )
101. Kd6 Be1 102. Nd8 Bb4+ 103. Kd7 What a heartbreaker! After Black's next move - assuming it's with a piece rather than the e-pawn - both sides will have made 49 moves since White played 54.f3. On White's 50th move he will take Black's e-pawn, and all Black's work will have been in vain: he'll have to start all over again, down a pawn. The position is still objectively drawn, but between the exhaustion, lack of time, and the discouragement of coming up a move short of achieving an immediate draw it would have taken a nearly inhuman toughness to pull out the draw from here.
103... e5 104. Nf7+ Kg7 105. Nxe5 Nh3?? That's right: a waiting move with the bishop. Black can even lose the second pawn and draw - not in just any situation, but here it's okay.
105... Be1 106. Nxg5 Kf6 107. Ngf7 Bc3 108. g5+ Kg7 109. Ke7 Bxe5 110. Nxe5 Nh3 111. g6 Nf4 112. Kd6 Kf6  )
105... Kh6 is the most natural move, and it's good enough.  )
106. Ke6 Ba3 107. Kf5 Bc1 108. Nd3?
108. Nc5 Ng1 109. Ne6+ Kg8 110. Nxg5  )
108... Be3?
108... Ng1 is again objectively drawn.  )
109. Ndc5 Kh6 110. Ne6 Bc1 111. Nd6 Ng1 112. Nf7+ Kh7 113. Nfxg5+ Kh6 114. f4 Ne2 115. Ne4 Kh7?
115... Bb2 116. g5+ Kh5 117. Nf6+ Kh4 118. Kg6 Bc1 119. Nd5 Ng3 may yet be drawn - though I wouldn't bet on it. The salient point is that if Black can give up both pieces for both pawns it is a draw, as White cannot force mate with a king and two knights against a bare king.  )
116. g5 Now it's unquestionably lost for Black.
116... Bb2 117. Nf6+ Kh8 118. g6 Ng3+ 119. Kg5 Bxf6+ 120. Kxf6 Nh5+ 121. Ke5 White will shove the f-pawn up the board until Black surrenders his knight, whereupon he'll win with his g-pawn.
121. Kf7 is fine but not an immediate knockout, thanks to
121... Ng7  )

Finally, an honorable mention to one of the players who finished in the group with 7.5 points. John Pigott is only a FIDE master, and he was born in 1957, but no matter: In the last two rounds he crushed a pair of grandmasters. His first victim was Magesh Chandran Panchanathan of India, and if that wasn’t enough he disposed of Alexei Shirov of Latvia, once among the top three in the world, and still No. 51, in the final round.

Pigott, John C vs. Shirov, Alexei
Reykjavik Open 2017 | Reykjavik ISL | Round 10.7 | 27 Apr 2017 | 1-0
1. e4 c5 2. Nf3 Nc6 3. Bb5 d6 4. O-O Bd7 5. Re1 a6 6. Bf1 Bg4 7. d3 Nf6 8. Nbd2 e6 9. h3 Bh5 10. c3 Be7 11. g4 Bg6 12. Nh4 Nd7 This is a well-known position that has arisen in hundreds of games. About half the time White plays 13.Ng2, the other half 13.Nxg6.
13. Ng2 h5
13... e5  )
13... h6 and  )
13... O-O are all significant alternatives.  )
14. Nf4
14. f4 has also been tried, but without much success. Here's an example from the highest level:
14... hxg4 15. hxg4 Qc7 16. Nf3 O-O-O 17. Ne3 Nb6 18. Nc4 Nxc4 19. dxc4 f5 20. exf5 exf5 21. g5 Bf7 22. Qc2 g6 23. Qf2 d5 24. cxd5 Bxd5 25. Be3 Bxg5 26. Qg3 Be7 27. Bg2 g5 28. Nxg5 Bxg5 29. Bxd5 Rxd5 30. Qxg5 Qf7 31. Kf2 Rh2+ 32. Kf1 Rd8 33. Qg3 Qc4+ 34. Kg1 Rxb2 0-1 (34) Anand,V (2791)-Carlsen,M (2776) Nice 2009 (blindfold)  )
14... Nde5
14... Bh4 may be best.  )
15. gxh5! Ugly but strong.
15... Bh7
15... Bg5 was played in an earlier game.
16. hxg6 Bxf4 17. gxf7+ Kxf7 18. Nf3 Qf6 19. Bxf4 Qxf4 20. Re3 Ke7 21. Nxe5 Nxe5 22. Qe2 Raf8 23. Bg2 c4 24. dxc4 g5 1/2-1/2 (24) Levine,Y (2062)-Horvat,D (2192) Iasi 2011  )
16. Nb3 Bh4?
16... Qb6!  )
17. Ng2 Qf6 18. Nxh4 Qxh4 19. Re3 f5 20. d4 f4 21. Re1 Black's position isn't very good, and if he retreats the knight White plays Qg4, sending everyone home. Shirov, being Shirov, finds an imaginative way to keep his kingside initiative alive, but it's based on a miscalculation.
21... Bg6?
21... g6! 22. h6! Nf7 was Black's best.  )
22. hxg6 Rh6 23. Bxf4! Shirov probably missed this shot. Without the f-pawn and with the queen pulled off the h-file, Black's attack loses its bite.
23. dxe5?? Rxg6+ 24. Bg2 Qxh3 25. Kf1 Rxg2  )
23... Qxf4
23... Rxg6+ 24. Bg3 Rxg3+ 25. fxg3 Qxg3+ 26. Bg2 is even worse.  )
24. dxe5 Rxg6+ 25. Bg2 Qg5 26. Qg4 Black's attack is finished, and since he's down a piece and a pawn there's no material compensation either.

—————————————————————-

Dennis Monokroussos is a FIDE master who has written about chess on his blog “The Chess Mind,” since 2005. He has been teaching chess for almost 20 years and for the last 10 years has been making instructional chess videos, which can be found at ChessLecture.com. Between 1995 and 2006, he taught philosophy, including a four-year stint at the University of Notre Dame.