Caruana stopped, but $100,000 richer

Fabiano Caruana finally conceded a draw but won the Sinquefield Cup with two rounds to spare

The overwhelming focus of the chess world was on the clash between the World Champion and his potential future nemesis, but let’s first take a brief look at the rest of the Round 8 action.

Hikaru Nakamura 0 – 1 Veselin Topalov

It's all going wrong for the American no. 1 | photo: official website

Nakamura’s disastrous tournament went from bad to worse, with a fourth loss in St. Louis. It was a case of Berlin blues, as the American fell victim to the super-solid Berlin Wall (replay the game with computer analysis here). Topalov’s 16…h5! was the fruit of a well-spent 18 minutes, while after 18…Bc8 nothing was working for White and Black had seized the initiative. Nakamura's position was soon pitiful to behold:

Sure enough, Topalov brought home the full point, though Nakamura showed his masochistic streak for a second day in a row, playing on until move 57.

The result puts Veselin back up on 50% and only half a point behind Carlsen – something that would have been hard to imagine when he started with 0/2 and a lost position against Nakamura in Round 3.

Levon Aronian 1/2 – 1/2 Maxime Vachier-Lagrave

Levon Aronian spoiled what looked on course to be a sparkling miniature | photo: official website

In the wake of his loss to Fabiano Caruana the French no. 1 Maxime Vachier-Lagrave had said his first aim was not to suffer a collapse. That looked very much on the cards after he rejected the bait of a g4 pawn sacrifice on move 5 (!) and was close to lost by move 10. Aronian was soon crashing through the Frenchman’s defences and felt he was on course for his best game in St. Louis, but 

...once I’d got a winning position I just started playing very mediocre moves… there’s something wrong with me in this tournament.

The game ultimately turned on the position after 22…Kc7:

Despite letting some of his edge slip the Armenian’s creative play has kept him well on top, until he blundered here with 23. Rde1?, allowing Maxime to collect the crown jewel on h7 - 23…Rexh7. The game then ended in a draw at breakneck speed.

Aronian explained afterwards:

I was planning to get this position and I thought if I don’t have Re1 then there’s nothing much here. The whole point of this was Re1, but then I just blundered that he can take on h7. That’s ridiculous, of course… I just noticed he can even take with either rook – that’s embarrassing.

The only solace Levon could find after a tournament where he’d ceded the world no. 2 spot to Fabiano Caruana was that the Italian’s feat would see everything else forgotten:

Well at least the attention will be taken from my terrible play here.

Fabiano Caruana 1/2 – 1/2 Magnus Carlsen

Fabiano Caruana means business as he takes off his jacket | photo: official website

And so we come to the game of the round, and perhaps the most anticipated of the last year. Our guide is the one and only Peter Svidler (with some quotes from the players included for good measure):

The most eagerly anticipated game in a while - what would it bring? 

Although the recent encounters between these two featured constant bloodshed, smart money was on a draw - Magnus does not lose too many important games, and he clearly was extra motivated to end the streak and return the world order to some semblance of normalcy.

As for the idea of Caruana losing - we all know it's going to happen again, some day, but in this event? Fairytales aren't supposed to work like that, and as Fabiano himself said, not losing this game was a huge priority for him. 

And draw is what we got - but not the kind most people would predict.

1. e4 c5 2. ♘f3 ♘c6 3. d4 cxd4 4. ♘xd4 g6 The time difference makes watching this event live a tricky proposition for Russian parents with school-age kids, but I was determined not to miss this game. I will admit this choice filled me with apprehension, though - the Accelerated Dragon is not the most exciting of openings, especially after

5. c4

5. ♘c3 ♗g7 6. ♗e3 ♘f6 7. ♗c4 supposedly has more potential for a sharp fight.

5... ♘f6 6. ♘c3 d6 White has a number of options here.

7. f3

7. ♘c2 is an attempt to keep more pieces on the board. I played this myself a number of times - but objectively Black should be fine in positions arising after 7... ♗g7 8. ♗e2 0-0 9. 0-0 ♘d7 10. ♗d2

7. ♗e2 is the main move - but the positions after 7... ♘xd4 8. ♕xd4 ♗g7 , and now, for instance, 9. ♗g5 (or 9. Be3) 9... 0-0 10. ♕d2 tend to end up as ever-so-slightly better for White endgames - not what most of us wanted to see.

7... ♘xd4 8. ♕xd4 ♗g7 9. ♗e3 0-0 10. ♕d2 Compared to 7. Be2, White is slightly less developed - but with the structure as solid as this, it's hard to exploit the half-a-tempo difference effectively.

10... a5⁉ As far as I could work out, if Black wants he can still get the extremely well-known endgame by playing

10... ♗e6 11. ♖c1 or (11. ♗e2 ♕a5 12. ♖c1 ♖fc8 13. b3 a6 ) 11... ♕a5 12. b3 ♖fc8 , and after 13. ♗e2 a6 14. ♘a4 ♕xd2+ 15. ♔xd2 ♘d7 16. g4 f5 (227 games in my database - I was not kidding about the popularity of this line) 17. exf5 gxf5 Black is currently doing OK. However, Magnus had other plans.

11. b3

11. ♗e2 was played by Garry Kasparov in the famous match against the Israeli national team in 1998, and after 11... a4 12. 0-0 ♕a5 13. ♖ac1 ♗d7 14. c5 ♗c6 15. cxd6 ♖fd8 16. ♘b1 ♕xd2 17. ♘xd2 exd6 18. ♘c4 d5 19. ♘b6 ♖a5 20. e5 ♘d7 21. f4 White was better: Kasparov-Alterman, Tel Aviv 1998. Not allowing a4 and Qa5 seems like a very logical idea as well, of course.

11... a4

11... ♗d7 12. ♗e2 ♗c6 13. 0-0 ♘d7 is always possible - but Black is relatively passive, and only real Accelerated afficionados, such as Rauf Mamedov, have opted for this in the past. 11...a4 is an attempt to disrupt White's serene development.

12. b4 ♗e6 13. ♖c1 ♘d7 14. ♗e2 ♘b6 Black is putting immediate pressure on the c4-pawn. White's next decision is quietly noteworthy.

15. ♘b5⁉

15. ♘d5 is what you would play in a blitz game - it forces 15... ♘xd5 (15... ♗xd5 seems extremely unattractive to me)

a) 16. cxd5 ♗d7 17. a3 (17. 0-0 a3 18. ♗d4 ♗xd4+ 19. ♕xd4 e5! is fine for Black: 20. dxe6 ♗xe6 21. ♗c4 ♕e7 ) 17... b5 18. 0-0 ♖c8 is a little bit better for White - but the most likely outcome is all the rooks getting traded along the c-file, and then Black should hold quite easily - without a knight to head towards c6 White will find it hard to get too much out of his space advantage.

b) 16. exd5 ♗f5 17. c5! dxc5 18. bxc5 White has tangible pressure with no risk: 18... ♕c7 19. 0-0 ♖fd8 20. ♖fd1+/= Fabiano has entirely different ideas, though.

15... a3!

15... ♖c8 16. ♘a3! stabilizes the queenside, leaving Black with a number of pieces not doing very much, particularly that knight on b6. After 16... f5 17. exf5 ♗xf5 18. 0-0 White is clearly better.

16. ♘d4 Preventing

16. 0-0 ♗b2 17. ♖c2 ♖c8

16... ♗d7 So far so good. Time to castle, bring the rooks towards the centre, aim for c4-c5 at an opportune moment and perhaps consider trading the dark-squared bishops via Bh6?

17. h4⁉ Nope.

Caruana: It looks very dangerous for him. His knight on b6 is stranded and the king has no defenders, so I decided just to go for it. 

17. 0-0 ♘a4 18. ♖c2 ♕c7 19. ♖fc1 is what most players would have played here - and White is, in fact, somewhat better.

17... h5 Black could ignore the kingside for now:

17... e5 18. ♘b3 ♗e6 19. h5 ♘a4 , and White would most probably have to switch back to positional play: 20. hxg6 fxg6 21. 0-0 ♖f7 22. ♘a5 with definite pressure: 22... ♕d7 23. ♗d1! ♘b2 24. ♖c3 , and the pawn on a3 is a target. 

However, 17...h5 is an extremely natural move to make. After all, White's king is not on b1, behind the three pawns - it's on e1. Surely he can't afford to break open his structure even further? Even after these moves appeared on the board, I was finding it hard to believe it was any good.

18. g4!

Fabiano did not really have such qualms. 

Carlsen: I thought when he plays h4 and g4 he needs to play for mate, which he certainly doesn’t because positionally he’s doing quite well.

Sleep was clearly no longer an option - I had to know how this ends. As I watched, I also decided to try and analyze the events on the boards 'live', so to speak. Most of the lines that follow were produced in real time, edited and expanded next morning - but I will also try to give you my original impressions.

18... hxg4 19. fxg4 e5 After

19... ♘a4 20. h5 e6 the machine suggests the remarkable 21. ♔f2! ♖c8 (21... e5 22. ♘b5 ) 22. ♔g2 , calmly improving the position of the king, and getting ready for the final assault. Black can't really do without e5.

20. ♘b3 On Twitter, Kasparov liked

20. ♘b5 better: 20... ♗xb5 21. cxb5 d5 22. h5! , and White has definite pressure. There is absolutely nothing wrong with Fabiano's choice either, though.

20... ♗c6⁈ By far the most natural move in the position - but I believe this is the root of Black's troubles.

20... f5 is bad: 21. ♗g5! ♕e8 22. gxf5 gxf5 23. h5 , and White is crashing through. However, after

20... ♗e6 21. h5 ♘a4! the evaluation is still quite unclear. I became very interested in this position as Magnus was considering his options - especially after I realized (with help from the comp, of course - a luxury Magnus did not have) just how unpleasant Black's life is after 20...Bc6. After some prodding, the machine suggests White is better here as well, but the following play is hardly obvious, to put it mildly:

a) 22. ♘a5⁉ ♕c7 , and only now 23. ♗g5! , creating a huge threat of h5-h6-h7. 23... gxh5 (23... f6 24. ♗e3! gxh5 25. gxh5 f5 26. exf5 is now bad for Black - the inclusion of Na5 and Qc7 makes a lot of difference: 26... ♗xf5 27. ♕d5+ ♔h7 28. h6 ♗f6 29. ♖f1± ) 24. ♖xh5 f6 25. ♗h6 ♖f7 26. g5 ♕b6 27. ♔d1! Bringing the king over to the safe haven on c2. White is better here - but with impending time-trouble, and about half a dozen viable alternatives at every turn, I am only giving these lines so that my futile efforts to figure out what's going on after 20...Be6 don't go entirely to waste.

b) 22. ♗g5⁉ is also interesting: 22... f6 (22... ♗f6⁉ 23. ♗xf6 ♕xf6 sees White suddenly switch tack: 24. ♘a5! aiming for the b7-pawn, and later, with Rh3 - for the one on a3. The exchange sack 24... ♖xa5 25. bxa5 ♘c5 26. ♕e3 does not promise Black an easy life either) 23. ♗h4! (23.♗e3⁈ gxh5! 24. gxh5 f5 25. exf5 ♗xf5 26. h6 ♗f6 27. 0-0 (a mildly remarkable case of late castling - but things have gone wrong for White) 27... ♕d7 is very unclear) 23... ♕d7 24. hxg6 ♕e8 25. ♗f2! (25. ♕xd6 ♖d8 26. ♕c7 ♖d7 27. ♕a5 ♗h6! allows a lot of counterplay) 25... b6 26. c5! , and the ensuing complications seem to favour White. I say 'seem' because not only are these moves hardly natural - I can't even bring myself to believe the assessments, despite running the beast for quite a while.

21. ♗f3

21. ♗d3 is also a decent move, but 21. Bf3 is much more natural, and very good.

21... f5 Objectively, Black is in trouble after this - but with the bishop on c6, he does not really have any other sensible plans.

21... ♘a4 22. h5 is no longer a viable option - White controls most of the board, and after 22... ♕f6 he can even play 23. 0-0 ♕e6 24. ♘a5 with a large edge.

22. gxf5 Starting with

22. ♘a5 would be a mistake: 22... ♗xe4⁉ 23. ♗xe4 fxe4 24. ♘xb7 ♕f6! 25. ♗xb6 ♕f3 26. ♖g1 ♖ab8 27. ♕d5+ ♔h8 , and it's a mess, completely unnecessarily for White.

22... gxf5 23. ♘a5! Keeping the central breaks under control.

23... f4 Black's choices are limited here: after

23... ♖xa5 24. bxa5 ♘a4 25. ♗g5 ♕d7 26. ♖g1 he is always a tempo short: 26... ♔h7 (26... fxe4 27. ♗h6 ♖f7 28. ♗h5 ♖e7 29. ♕g5 , and White is winning) 27. ♗h5! ♕e6 (27... ♗xe4? loses to a very nice blow: 28. ♗h6! ♗xh6 29. ♗g6+ ♔g7 30. ♗xf5+ ♔f7 31. ♗xd7 ♗xd2+ 32. ♔xd2 , and Black is lost) 28. ♖d1 , and White is in time with his attack:

a) 28... ♕xc4 29. ♗e7! ♕xe4+ 30. ♔f2 , and it turns out White's threats are far more important: 30... ♗h6 (30... ♖g8 31. ♕g5!+− ) 31. ♗g6+ ♔g7 32. ♗xf5+ ♔f7 33. ♗xe4 ♔xe7+ 34. ♔e2 ♗xd2 35. ♗xc6 bxc6 36. ♔xd2 , and White should win - he has two outside passers

b) 28... fxe4 29. ♗d8‼ is yet another hard-to-find computer move which wins on the spot 29...♗h6 (29... ♘b2 30. ♕xd6+− ) 30. ♕xd6+−

24. ♗f2 Black has managed to slow down the attack and stabilize somewhat - but now that the tension in the center is gone, he is stuck with permanent problems. His king is weak, the bishop on g7 is shut out, both d6 and a3 are potentially weak, and most of the endgames will be difficult as well.

24... ♘a4 25. ♘xc6 A very practical decision - White could postpone this for now, but then he would have to constantly ask himself if Rxa5 is any good.

25... bxc6 26. 0-0⁈ Up to this point Fabiano has played an excellent game - you could argue with some of his decisions, but his play was logical, consistent and strong. However, he was also running a bit short on time - unsurprisingly, considering how difficult this position is to evaluate and play - and from this move on, the game becomes much more haphazard.

26. ♖g1 ♔h8 27. h5 ♗f6 28. ♔e2 seemed much more natural to me as I was watching - and not only to me. The king is safely ensconced on e2, and White is ready to continue with the kingside assault.

The text move does not give away all of the advantage - but it showed that Fabiano has lost some of his earlier rhythm.

26... c5 Magnus was playing very fast, as well - trying to put as much pressure on Fabiano as possible. This move has a very sound idea behind it - it limits the scope of the f2-bishop and prepares the knight retreat to b6, where it will an eye on the c4-pawn and keep the d5-square in check. The protected passer on b5 is not to be sneezed at, however. After the normal

26... ♔h8 27. ♔h1 ♗f6 28. ♖g1 ♖g8 29. ♖xg8+ ♔xg8 30. ♖g1+ ♔f7 31. h5 Black is in trouble - he has no real counterplay, and is facing threats from every direction.

27. b5 ♗f6 It was too early for

27... ♘b6 - 28. ♖fd1! ♖f6 29. ♗xc5 wins material: 29... ♖g6+ 30. ♔f1 ♕xh4 31. ♗xb6 ♕h3+ 32. ♔e2 is just one check.

28. ♔h1 ♘b6⁈ Magnus feels he has to bring his pieces back into play - but in fact, he needed to spend one more tempo on safety.

28... ♔h8 was perhaps the best option: 29. ♖g1 ♕e7 30. ♖g6 ♖g8 31. ♖h6+ (31. ♖cg1 ♖xg6 32. ♖xg6 ♖g8 33. ♖xg8+ ♔xg8 34. h5 is a bit better for White, but he can aim for a bigger edge than that) 31... ♔g7 32. ♖h5 ♘b6 33. ♗e2 ♔f8 34. ♖d1 ♖d8 35. ♖f5 , and White controls the entire board. It will be very hard for Black to keep an eye on all his problems here.

28... ♔f7 29. ♕d5+ ♔e7 30. ♗g4 ♖h8 31. h5 is quite bad as well;

29. ♖cd1? This does spoil the game, however. It seems the pressure got to Fabiano after all. Both

29. ♖g1+⁉ ♔f7 (29... ♔h8 30. ♗g4! ♖a7 31. ♗f5 , and White has a tremendous attack 31... ♗xh4 ⁇ 32. ♖g4!+− ) 30. ♖cd1 ♖h8 31. h5 , with a huge edge; 

Carlsen: I intended to capture the pawn on h4, but the computer just told me that’s mate.

and in particular the prosaic:

29. ♖fd1! ♗e7 (29... ♗xh4 loses: 30. ♖g1+ ♔h8 31. ♖g4! ♗e7 32. ♗xc5! ♖f7 33. ♗xb6 ♕xb6 34. ♔g2 , and the black king is doomed) 30. ♗xc5! ♔h8 (30... dxc5 31. ♕g2++− ) 31. ♗f2 , winning an important pawn and keeping everything under control, were extremely strong.

29... ♔h8! Of course, Magnus does not fall for

29... ♗e7? 30. ♗xc5 But now both h4 and c4 are hanging, and things are much trickier for White. With time running quite short, Fabiano decides to make the game safe:

30. ♕xd6 It was still possible to play sharper:

30. ♗g4⁉ ♗xh4 (30... ♘xc4? is bad: 31. ♕e2! ♘b6 32. ♗f5 , and once agains White has a huge attack) 31. ♕xd6 ♕xd6 32. ♖xd6 ♗xf2 33. ♖xb6 , and White has some pressure in this endgame - but it's hard to fault Caruana for taking a safer option. He must have felt he lost the thread somewhat - and when that feeling comes, it tends to influence your next decisons as well.

30... ♕xd6 31. ♖xd6 ♘xc4 Exchanging d6 for c4, and trading queens, helped Black enormously. He no longer has to worry about the safety of his king, and his knight is suddenly taking a huge part in the action. 

With more time available on the clock White could, perhaps, pose some slight problems even here - but you could feel that Magnus was now very unlikely to slip up again. The best chance for a barely believable 8th straight win is gone by this point, and in time-trouble the game fizzles out.

32. ♖d5

32. ♖d7⁉ is a bit trickier, but after 32... ♖fd8! 33. ♖fd1 ♘e3 34. ♗xe3 fxe3 Black holds as well.

32... ♘e3 33. ♗xe3 fxe3 34. ♗e2 ♗xh4 The machine suggests

34... ♖a4 , planning to meet 35. ♖d6 with 35... ♖b4! 36. ♖fxf6 ♖xf6 37. ♖xf6 ♖b2 , and Black holds: 38. ♖a6 ♖xa2 39. ♗c4 e2 40. ♗xe2 ♖xe2 41. ♖xa3 ♖b2 , with an easy draw. One more example for the 'who would ever play like that' folder.

35. ♖f5

35. ♖xf8+ ♖xf8 36. ♖d3 is a bit more challenging, but after 36... c4! 37. ♖xe3 (37. ♖xa3 ♖f2 38. ♖xe3 c3 is a draw as well) 37... ♖c8 38. ♖c3 ♗e1 39. ♖xc4 ♖xc4 40. ♗xc4 ♗a5 Black is fine.

35... ♖xf5 36. exf5 ♖d8 We're almost done here.

37. ♖xd8+ ♗xd8 38. ♔g2 ♔g7 39. ♔f3 ♔f6 40. ♔e4 c4 41. ♗xc4 ♗b6 42. ♗d3 And a draw was agreed. 

A fantastic fight, despite the mistakes - and yes, for the 1st time in this event, Fabiano made actual mistakes, not inaccuracies in a completely winning position. If this is what the future holds, we're in for a fantastic ride - not every game between these two will be as exciting as this one (or have that much statistical relevance), but I think it's clear that there is now a proper rivalry at the very top, with two players unwilling to give a single inch - which is great news for chess.

There is also a small matter of technical results. With this draw, Fabiano Caruana has won the strongest ever tournament with two rounds to spare. Read that again - with TWO rounds to spare! With a score, so far, of 7.5/8 - and a legitimate reason to be somewhat disappointed it's not all eight. This performance is already legendary, it will stay legendary for as long as people follow chess - and it's been a privilege to watch it, and describe parts of it to the best of my ability.

1/2-1/2

Fabiano again won the day's understatement contest when he was asked how he'd approach the final two rounds:

I won the tournament but it’s still not over and it’s important to end on a high note. I’ll take them very seriously and try to do my best. I’ll try to play the way I’ve been playing, which has so far worked out.

Magnus Carlsen didn't get where he is by a lack of confidence, though. Asked about the prospect of a match against Caruana he commented:

Let him play a couple more tournaments like this and then we’ll talk (smiles). What he’s done here is absolutely incredible but I don’t think we should completely forget about what’s happened the last four years or so either. Naturally with this result he’s the clear no. 2 in ratings and obviously with such a result he deserves to be there.

2014 Sinquefield Cup standings after Round 8

So even Magnus Carlsen will be unable to make up a 3-point gap in two games, but it’s still all to play for in the race for second place, especially as Topalov has White against Carlsen in the final round. 

Round 9 pairings

Round 9 is none too shabby either – the official no. 2 plays the no. 1, while Caruana can remove any shadow of doubt about how impressive his tournament has been if he manages to inflict more misery on bottom-dweller Hikaru Nakamura.

Don’t miss live commentary with GMs Maurice Ashley, Yasser Seirawan and WGM Jennifer Shahade here on chess24! You can also rewatch the full Round 8 show here.

See also:

• Round 7: The Fabnificent Seven
• Round 6: Fabiano remains flawless
• Vachier-Lagrave on breaking into the Top 10
• Round 5: Just call me Fabi
• Round 4: Fab fourth
• Round 3: Who is the top Car now?
• Round 2: Caruana hot, Topalov not
• Round 1: Caruana takes first round by storm
• Record breaking Sinquefield Cup starts in St. Louis