They are a half point ahead of Ray Robson and a full point ahead of Hikaru Nakamura, Alexander Onischuk and the surprising 15-year-old, Jeffrey Xiong.

After five rounds, two of the three pre-tournament favorites, Fabiano Caruana and Wesley So, are dominating the field at the United States Championship, but there have been some unexpected turns in the last couple of rounds.

In Round 4, Caruana took a huge step towards the title by winning a smooth game against the other favorite, Hikaru Nakamura. But in Round 5, Caruana couldn’t make it past a surprisingly solid Alexander Shabalov, which let So catch up with another inspired tactical win against Varuzhan Akobian.

So’s crushing tactical knockout was probably extremely satisfying considering their history — in last year’s Championship, So had been forfeited against Akobian after six moves because So had scribbled words on his scoresheet. Meanwhile, Nakamura, the defending champion, showed that he has no intention of quietly giving up his title quietly by overpowering Samuel Shankland. 

In the Women’s Championship, one co-leader after Round 4, Tatev Abrahamyan, played a relatively short draw in Round 5 against Katya Nemcova. That gave the other leader, Nazi Paikidze a chance to grab the sole lead with a tough win against teenager Jennifer Yu. Meanwhile, the rating favorites, Irina Krush, the defending champion, and Anna Zantonskih, had a long game that ended in a draw. Krush is tied with Abrahamyan, a half point begind Paikidze, while Zatonskih is another half point back.

Before the tournament, I had wondered if the games between the three favorites might be somewhat anticlimactic as it appeared to me that their best strategy would be to play solidly against each other, and aggressively against the rest. But Nakamura had other plans in Round 4.

He went for a very provocative Najdorf Sicilian against Caruana’s 1.e4. In the past, Caruana has had losses in the Najdorf English Attack, which might have inspired Nakamura. But Nakamura’s interpretation of the line was unusual at best:

Caruana, Fabiano vs. Nakamura, Hikaru
U.S. Championship 2016 | | Round 4.1 | 17 Apr 2016 | *
e6 e5 is much more popular these days at the highest levels, but this is perhaps an even more theoretically intense line.
7. Be3 h5!? The idea of preventing g4 is common enough, but it is very unusual in this situation. It is much more common in the lines with 6...e5 for instance. What is the big difference? One important problem was shown by Caruana in this game:
8. a4 A switch of strategy! Now White does not want to castle queenside anymore. It might seem that f3 was pointless, but it did force black's weakness on h5. And castling kingside for Black looks extremely iffy.
8... Nc6 9. Bc4 drastically altering White's usual plans. The bishop might seem useless staring at e6, but it is better placed on c4 than on most other squares available to it. And in some variations, it is possible to sacrifice material on e6!
9... Qc7 10. Qe2 Be7 11. O-O Ne5 12. Bb3 Bd7 13. f4 Neg4 14. Kh1 Nxe3 15. Qxe3 At first glance, Black seems to be doing ok, and maybe he is. But practically, White's position seems slightly easier to play because Blacks king will definitely be uncomfortable wherever it goes. While the computer might not care too much about this, for human players this feels like an annoying thing to deal with.
15... Qc5 16. Rad1 g6
16... O-O 17. Qe2  )
17. Qe2

Caruana’s plan of simply switching to kingside castling was really smart and unexpected. Nakamura could also have castled kingside, but then the pawn on h5 is a severe weakness. However, castling queenside was extremely risky as well, as Caruana demonstrated very smoothly:

Caruana, Fabiano vs. Nakamura, Hikaru
U.S. Championship 2016 | | Round 4.1 | 17 Apr 2016 | ECO: B80 | 1-0
O-O-O?! I do not like castling kingside, but the king looks even more vulnerable on the queenside.
18. f5
18. e5 /\ Ne4 and Rf3 might have been even stronger but Caruana had an interesting pawn sacrifice planned:  )
18... e5 19. Nf3 gxf5 20. Ng5! the Nxf7 threats are too hard to deal with.
20... f4 fxe4 does not look appealing because after Nxe4 White has all sorts of rook lifts on the third rank to the queenside.
21. Rd3! Avoids the skewer by Bg4 and threatens Nxf7. White is also ready to lift his rook to the queenside after Nd5 which makes life very scary for Black. He should have perhaps tried to save the exchange, as after
21... Kb8 22. Nxf7 White has a material advantage and the initiative:
22... h4 23. Nxh8 Rxh8 24. Qf2! Exchanging queens would be very nice, so Black continued:
24... Qb4 25. Nd5 Nxd5 26. Bxd5 Black's queen is now misplaced:
26... Bxa4 27. Ra3 h3 28. c3 Qb5 29. b3 Bh4 30. bxa4 Qd3 31. g3

In Round 5, Caruana outrated his opponent by almost 300 points, but nobody would consider an experienced grandmaster like Shabalov to be a pushover. The game highlighted the hard choices the top three players have to make when playing Black against most other opponents: Go for something risky or just play solid chess? Not having the advantage of playing the first move is a huge factor in determining the way the game develops.

Shabalov has a reputation for playing extremely flashy, attacking chess. Understandably, Caruana chose to play something solid, perhaps hoping for a bit of unnecessary aggression from Shabalov. But Shabalov played an extremely solid game — opting for a symmetric Slav structure, and never giving Caruana enough room to create serious chances.

When playing White, the top three players are a level above the rest. So and Nakamura demonstrated that again in Round 5 by bulldozing their opponents, both of whom were rated higher than 2600.

In So vs. Akobian, the players quickly castled on opposite sides. Akobian’s decision to take the pawn on g2 seemed really brave, but the speed with which it was punished was phenomenal:

So, Wesley vs. Akobian, Varuzhan
U.S. Championship 2016 | | Round 5.2 | 18 Apr 2016 | 1-0
Bd5 A strange move. After
17. c4 Bxg2 is more or less the only justification, but why would you want to take such a pawn?
18. Bc3 Qb6 19. Rg1 Bc6 20. Nxf7! White probably had a great position even without this, but this is nicer:
20... Kxf7 21. Rxg7+! Kxg7 22. Qxe6 Qxf2
22... Re8 23. Qf5! creates beautiful mating threats.
23... Kf8 24. Bd2! and the king has no escape.
24... Red8 25. Qg6! Rxd3 26. Bxh6#  )
23. Qxe7+! Kg8 24. Bh7+ and after Nxh7, Qg7#.

Nakamura went for a slightly rare continuation against the Caro-Kann Defense though he ended up in a fairly typical middlegame position. Shankland seemed to be doing fine, but Nakamura’s blunt attack on the kingside was both surprising and incredibly powerful:

Nakamura, Hikaru vs. Shankland, Samuel L
U.S. Championship 2016 | | Round 5.4 | 18 Apr 2016 | 1-0
Bf8?! The start of a slightly dubious plan. The computer actually thinks it is a good plan, but I think that moving the bishop away from g7 is inviting trouble. Black had a super solid kingside, and suddenly he has real weaknesses. Nakamura was quick to pounce on it:
20. g4! Bc5 This makes matters worse, but I guess Shankland wanted to justify the Bf8 choice.
21. Nd4 f5 doesn't look scary immediately because e5 is hanging but it can be easily prepared.
21... Qb6 22. Bg1 Preparing f5.
22... b4 23. f5! and some nice tactical calculation.
23... bxc3 24. bxc3 Ncxe5 25. fxe6 fxe6 26. Nxe6! Bxg1 27. Rxg1 Kh8 28. Qxd5 all Black's pawns are gone.
28... Rab8 29. Rge1 Qb2 30. c4 Rb6 31. c5 Rbc6 32. Rcd1 h6 33. Nf4 Rf6 34. Nd3 Nxd3 35. Qxd7 Nf2+ 36. Kg1 Rcf8 37. Qd2

Jeffrey Xiong, who is only 15 years old, also won, beating Gata Kamsky in a tense game. The more experienced Kamsky played really enterprising chess throughout the game, starting with an exciting piece sacrifice:

Xiong, Jeffery vs. Kamsky, Gata
U.S. Championship 2016 | | Round 5.5 | 18 Apr 2016 | 1-0
d5! A very interesting piece sacrifice:
13. c5 Nc6! fast development.
14. Nd4
14. cxb6 d4 lets Black win it back.  )
14... Bxd4 15. Bxd4 Re8 An overly romantic move. It is hard to resist such a beautiful looking piece sacrifice, and Black has great compensation, but there was no need for it.
15... Qh4! would not have required a piece sacrifice and the position would have still been great for Black.  )
16. cxb6 Qh4 17. Ne2 axb6 18. Qd2 Bf5 Black has great compensation and it looked like Kamsky would have a clear upper hand, but White played very stubbornly:
19. Kd1 Rxe2 20. Bxe2 Nxd4 21. f3 Qf2 22. Re1 Be6 23. Rc3 Nc6 24. Bf1 Qh4 25. Bb5 Nd4 26. Bf1 Nf5 27. Rd3 d4 28. g3 Qf6 29. Ke2! Bd7 30. a4 b5 31. a5 Bc6 32. Kf2 Ne3 33. Be2 Rd8 34. Kg1 Great defensive technique!
34... Kg7 35. Bd1

A couple of decades ago, Kamsky’s play might have been enough to overpower a strong grandmaster. But Xiong’s stubborn defense shows how good the new generation of (computer-inspired) players are in defending such positions. Xiong held on, bringing the king back to safety and probably equalizing, when Kamsky made a huge blunder:

Xiong, Jeffery vs. Kamsky, Gata
U.S. Championship 2016 | | Round 5.5 | 18 Apr 2016 | ECO: B03 | 1-0
Bxf3?? Time trouble strikes again.
36. Bxf3! Qxf3 37. Rexe3! dxe3 38. Qb2+ The move missed by Kamsky!

Alexander Onischuk balanced it out for the old guys by winning a smooth game against the other teenager in the event, Akshat Chandra. 

The last game of Round 5 to finish was between Ray Robson and Aleksandr Lenderman. Lenderman was in deep trouble but he created an incredible fortress. The position probably deserves Dvoretsky-level analysis [Editor’s note — as in Mark Dvoretsky], but my first impression is that it is very hard to improve White’s play because an extra knight can be a really awful piece in the endgame. It can jump around a lot, but it is really inflexible.

Robson, Ray vs. Lenderman, Aleksandr
U.S. Championship 2016 | | Round 5.3 | 18 Apr 2016 | 1/2-1/2
38. f5 It seemed like White will win this endgame. But things become interesting as Lenderman puts up a stubborn defense:
38... Kc7 39. fxg6 hxg6 40. Kf2 Kb6 41. Ke3 Kxb5 42. Kd4 a5 43. Ne3 a4 44. g5 Kb4 45. Nc4 The White knight is awkwardly placed. It can't capture the a pawn, while it can't go to the kingside either because then it will be impossible to block the a-pawn.
45... f5 46. Kd3 Kb3 47. e3
47. Nd2+! should be winning, but it was far from easy to calculate:
47... Kb4 Kb2 is too slow because a3 isn't a threat as Nc4+ would win.
48. Kc2 Kc5 49. Nf3! Kd5 50. Nh4! Ke4 51. Nxg6 and Black has captured the g6 pawn just in time. Now h4...h5, etc., should win.  )
47... Kb4 48. Kd4 Kb3 49. Kd3 Kb4 50. Ne5? I don't know if White was still winning, but now it is too late. It would be interesting to analyze options other than Ne5.
50... Kb3 51. Kd2 a3 52. Nf3 a2 53. Nd4+ Kb2 54. Nc2 Kb3 The White knight can definitely no longer leave.
55. Kd3 Kb2 56. Kd2 Kb3 57. h4 Kb2

In the crucial game in the Women’s section, Yu had been keeping pace with Paikidze, her more experienced opponent, but Yu made a subtle mistake that dramatically altered the game:

Paikidze, Nazi vs. Yu, Jennifer R
U.S. Women's Championship 2016 | | Round 5.2 | 18 Apr 2016 | 1-0
Kg7? This move looks perfectly harmless but now Black is completely lost.
32... Rd8! 33. Rxd8+ Qxd8 and ideas like Qd1+ insure that Black has a perpetual.  )
33. Rd6!! The amazing thing is that this is like a zugzwang position!! For example, Rd8 is no longer possible because of Qd4+ while Kg8 loses to Rd7 because of Qxc6 and then Rd8+. It's hard to imagine, but Black runs out of moves soon!
33... c3
33... h5 34. h4 Kh7 35. Rd7! Qxc6 36. Rxf7+ Kg8 37. Qe7  )
34. Qxc3+ Kh6 35. g4 Now things are simple:
35... f6
35... Qxd6 36. g5+ Kh5 37. Qf3+ Kh4 38. Qg4#  )
36. Qxf6 Qa7+ 37. Kg2 Rc7 38. g5+ Kh5 39. Qe6

In Round 6, Nakamura faces a huge test when he plays Black against So. It is unfortunate for Nakamura that he has had Black against both the other top players, but if he plays aggressively once again, things could be very interesting. Meanwhile, Caruana will get White against Onischuk, so he should be raring to go.


Parimarjan Negi is an Indian grandmaster who is the second-youngest ever to earn the title (at 13 years 4 months and 22 days). Ranked No. 90 in the world, he is currently a sophomore at Stanford University.