If I’m totally honest, I don’t think I learned to fully appreciate rapid tournaments until this year. It took three tournaments to change my mind: the 2015 Chess World Cup, the Ultimate Blitz competition featuring Garry Kasparov, and today, the first leg of the Grand Chess Tour in Paris. Unlike longer time control games, rapid chess emphasizes strong, practical play, and takes the spotlight off of brilliant opening preparation. At this level of competition, winning implicitly requires two elements: accurate calculation and the ability to convert better endgames. In the first day of competition alone, I found five endgames worth sharing and wanted to break down each of their critical moments in today’s critical endgame posts. Remember, as we move through each game, take a minute to assess the various defining features of the position: activity, solidarity, king safety, and ability to improve.
For our first endgame, we start with the protagonist of the story thus far, Magnus Carlsen. While his Grand Chess Tour started with an eerily similar first round, it’s important to not overlook the accuracy he brought to this endgame against Wesley So’s particularly stingy defense.
Carlsen–So, Paris 2016
White to Move
On face value, the position seems fairly equal. After trading rooks on e8, the position provides us with a symmetrical pawn structure and equal material. However, two elements stand in the way of the American achieving full equality. First, the bishop on a7 is dormant, pushed away from the action thanks to the bishop on g3 and the pawn on d4. Furthermore, his pawn on b7 is backward, and can easily become a target should White move his knight to c5 in the future. Black’s plan here is to march his king to c8 to cover b7 and prepare …Ba7-b8, and with only one real structural weakness in the position, should have enough to hold a draw. Magnus can’t really do too much to stop this idea, so he makes the most of his turn with his next move, 27. a4!
The easiest way to improve the position! Here Magnus plans a2-a4-a5 with the idea of fixing the queenside pawn structure, particularly the b7 pawn. While Wesley will be able to trade dark-squared bishops, the downside will be that the dark squares in his structure will be weak, and White will gain time to put further pressure on b7. 27…Qe7 28. a5 Kd8 29. Qd1 Qe4 30. Kh2 Ne7 31. Qb3
Neither side is really in a rush to convert or prove anything, so each side marked time by improving their respective positions. Magnus by making his king safer and fixing the b7 pawn, Wesley by centralizing his queen and bringing his king closer to c8. Here Carlsen offers his knight since 31…Qxd3? 32. Qxb7 is close to lost for Black. The bishop on a7 is still trapped, and the queenside pawns are falling. Here Black correctly chose to continue his plan. 31…Kc8 32. Qb4 Qe6 33. Nf4 Qf7
Wesley may be moving backward, but he still boasts a solid defense. As long as he has only one weakness, it will be very difficult for Magnus to make progress. In the next “phase” So executes the dark-squared bishop, and the f4 knight finds the c5 square. 34. Kg1 Bb8 35. Nd3 Bxg3 36. fxg3 Nf5 37. g4 Ng3 38. Nc5 Again, the game is relatively equal, and Wesley has put up the toughest defense we’ve seen in this series thus far.
White counterintuitively doubled his pawns, giving the Black knight targets from f5. While I appreciate the idea of compactness, I think this structural decision made life for Magnus a little more complicated. Instead of 34. Kg1, perhaps he could have considered other prophylactic resources, but in this position, he’s still doing fine. White now has the pressure he wants on b7, but the problem now is that his pawn structure closes his army off from the kingside, giving Wesley the break 38…h5 39. gxh5 and the natural 39…Qe7. But as it turns out, this gives Magnus a tactical opportunity in 40. Ne6!. These moves are hard to find in rapid play, so I can’t really blame Carlsen for the miss.
Anyways, this move would have been an amazing find. By revealing a discovered attack on the queen, Black’s options are limited. Already we can see that 40…Qxe6 41. Qf8+ Kd7 42. Qxg7+ will win back the knight back and retain a healthy pawn advantage. More critical was 40… Ne2+ 41. Kf2 Qxe6 41. Qf8+ Kd7 42. Qxg7+ where White doesn’t pick up the knight, but the h-pawn is simply unstoppable (see diagram).
Black can consider 40… Qxb4, but the knight and pawn endgame is worse for Black after 41. cxb4 Nxh5 42. g4 Ng3 43. Kf2! stopping the fork on e2, and once the g7 pawn falls, White’s h-pawn becomes a headache. That being said, these moves are really unnatural but I like how it highlights flaws in Black’s position. Black has two concrete weaknesses, b7 and g7, and the task of covering both of them is extremely difficult if White plays the best moves.
Instead, Carlsen chose 40. Kf2 and the game continued. 40…Nf5 41. g4 Qe3+ and equality was temporarily reached.
One of the problems with Magnus’ position in this game was that his focus on b7 dragged his pieces away from protecting his king, thus allowing Black to infiltrate through the center. Surprisingly, Black can’t coordinate his knight and queen to deliver mate, but he has many perpetual options. Given the nature of rapid chess, Wesley naturally tried for a win by improving his position with 42. Kf1 Qxh3+ 43. Ke1 Qg3+ 44. Kd2 Nd6
The retreat not only protects b7, but it intends to reroute the knight to either e4 or c4 in the future. For those trying to find better for Black, it’s quite difficult since Qb4xb7 is a constant threat, and defending the b7 requires a passive retreat. I was really surprised with how quickly Carlsen made his next move, but it makes a lot of sense. After 45. Nxb7! Carlsen gives himself a lot of chances. If 45…Nxb7 46. Qf8+ wins the g7 pawn, and again we see the danger of the passed h-pawn. With best play, Black should be able to find a perpetual, but it’s in these complications Wesley finally errs and his position goes south. 45. Qg2+ 46. Kc1 Qf1+ 47. Kc2 Qe2+ 48. Kc1 Qe1+ 49. Kc2 Qe4+ 50. Kb3 Nxb7 51. Qf8+ Kc7 52. Qxg7+ Kb8
53. h6 Qd3? +-
I was watching the live commentary from St. Louis at this moment, and was surprised they didn’t scrutinize this moment, because once this move is made, Wesley can never hope to recover. Black should have been able to find 53…Nxa5+ 54. Ka2 Qd1, the idea being that White cannot stop all the checks on a4, b3, and d1, so perpetual is forced. The problem with Wesley’s move is that it does nothing to improve his position. His next move, 54…Qb1 shows he wasted a tempo, and unfortunately, it’s enough to ensure Magnus a second queen. 54. Ka3 Qb1 55. h7 Qa1+ 56. Kb3 Qd1+ 57. Kb4 Ka7
With no more checks in the position, Wesley moves his king away from a future check. Both players were in severe time trouble, but it was still a surprise when the game suddenly concluded after 58. h8=Q Qa1 0-1 and it was Black who had won, not White (see diagram)!
With about twenty seconds left (not to mention a ten-second increment), Carlsen found himself stuck between 59. Qxb7+ and 59. Qh2, both of which were completely winning. In a moment of curiosity, Carlsen decided to look at Qh8-h2 into more depth, and completely forgot about the clock, letting his time reach zero!
Despite the drama, the reigning World Champion played a great game, pushing Wesley each move to find the best moves. So, of course, played solidly as well, but as we’ve seen so many times this series, one mistake in the endgame can quite often be unforgivable. Accuracy counts, and at the end of the day, it’s what goes on the scoresheet.
Our next three examples all occurred in the third round, and each provided instructive moments.
Fressinet–Caruana, Paris 2016
White to Move
After what had already been a complicated rook and pawn endgame, we see that the Black king’s inability to get into the game is causing Caruana great difficulties. The live commentary team in St. Louis found some nice ideas to potentially reach equality earlier in the game, but already it’s too late. The French wild card needed to get his king off of b8, and played 51. Rc1 to prepare Kb8-c8 and promote his pawn. Once again, Fabiano tried the interference idea of 51…Rc3, but now with the rook to the right side of the pawn, White won with 52. Rxc3 h1=Q 53. Kc8 Qh8+ 54. Kc7 Qh2 55. Rc5 Qxf2
56. Rc6 Qa7 57. Kc8 Qa4 58. Rc7+ Kg6 59. b8=Q
And Fressinet went on to convert the material and win the game. So what was the difference between taking on c3 and a3 you may ask? Well, winning or not to put it simply. If Fressinet had played 51. Rxa3? his rook doesn’t have a check on c7, and after 51…h1=Q 52. Ka7 Qc1,
White cannot hope to promote the pawn and keep his material advantage. Again, accuracy is the critical difference between winning and drawing.
Having proven himself to be a very capable escape artist, Wesley So once again found himself in trouble against Maxime Vachier-Lagrave. Unlike his lucky break against Magnus, he failed to find any miracles and lost this pawn down queen ending.
So–Vachier-Lagrave, Paris 2016
White to Move
I decided to insert this game since Black still has to be careful. Between pushing the c-pawn and avoiding perpetual checks, Maxime must also cover the f7 pawn, which makes his task a little more difficult. On the bright side, all queen trades are winning for Black, so it will be very difficult for White to create serious pressure. Wesley start his defense by playing 39. Qa1+ to maneuver the queen to f6 and directly attack f7. 39…Kh7 40. Qf6 Qd5 41. g3 c5 42. Kf1 c4
Black is making progress, but his position is also easier to play now. With the c- and f-pawns both protected by the queen, MVL can take a few moves to improve his position. 43. Ke2 Kg8 44. Qc3 h5 45. h4 Kf8 46. Qe3 Qd6
Here Maxime has made a little bit of progress, but now he must figure out how to make his king more active. After, 47. Qc3 Qc5 48. Qe3 Qd5 49. Qd2 Qe5+ 50. Kd1, it turns out that Wesley can do little to stop the advancing Black monarch.
50…Ke7 with the threat of …Qe5-d6! 51. Kc2 Ke6 52. f4 Qd5 0-1
Perhaps at the expert level, White can hope to play on, but this endgame is lost. Black’s king will waltz to g4 and pick up all of White’s kingside pawns, and White can’t stop all of Black’s pawns. Wesley resigned, leaving us one more great endgame from the round.
Carlsen–Aronian, Paris 2016
Black to Move
With a little help from the computer, GM Eric Hansen had a nice find here in 29…Qa1!, which should draw after 30. Qc5 Qa7 31. Qb5 Qa1 with repetition. The real idea is that 30. Qxb7 Qxc3 31. Qxc7 at least offers Black a lot of activity and decent drawing chances. But of course, Stockfish doesn’t play for us in tournaments, and the natural 29…Qa8 was played, giving White a nice edge since his pieces can be activated faster than Black’s. Skip ahead a few moves, and Black found himself completely paralyzed.
I really liked this moment of the game, as Carlsen realized that his would be much safer on the kingside, not to mention, an incredible for the b-pawn. 50. Ke2! Kg7 51. Kd3 Ng8 52. Ne8+ Kh8 53. Kc4 h5
As Black begins to open the kingside, it’s Magnus’ king that has found refuge, and the entirety of Aronian’s position submits itself to passivity. The next part of Magnus’ plan is to capture the c6 pawn and use his passed b-pawn to limit Black’s queen. 54. gxh5 Qh6 55. Qxc6 Qd2 56. hxg6
In trying to create activity, Black has to give up his g-pawn. While Black may have some checks now, he has the constant issue that ideas like Qg7 and Qh7 are checkmate! Just like our first Endgame Essentials post, king safety proves to be Aronian’s undoing. 56…Qe2+ 57. Kc5 Qxf2+ 58. Kb5 Qxg3 59. Qd7 Qxg6 60. Ka5
Black may have regained his pawn by force, but the threat on g7 is constant, and the Black knight can’t help Aronian salvage the position. 60…Qg3 61. b5 Qc3+ 62. Ka6 Qa3+ 63. Kb7 Qg3
Once again highlighting Black’s problems. Whenever Aronian runs out of checks, he must return to the defense of g7, giving White a tempo to push his b-pawn further down the board. 64. b6 Qg6 65. Ka7 f5 66. exf5
I was really impressed watching Magnus here. Basically everything wins here, but after Aronian’s f-pawn push, he stopped, calculated and found the move that allowed the least amount of counterplay. A great micro-moment from Magnus here that showed his master class despite the rapid time controls. 66…Qg3 67. f6 Qa3+ 68. Kb8 1-0
With no complications to offer, Aronian threw in the towel here, as both the b- and f-pawns are preparing to promote and sink the ship that is Black’s position. With a win here, Magnus won a second straight, proving he was completely unfazed by his surprising first round “defeat”.
For our fifth and final endgame, I wanted to share a nice idea found by the commentary team that shows a benefit of the opposite-colored bishop ending. In this fifth round encounter, an early slip from Magnus gave Hikaru Nakamura an opportunity to press before cashing in on a draw. While the engines do agree that the position has relative equality, from a more human point of view, Black had a nice geometrical idea to press even further.
Carlsen–Nakamura, Paris 2016
Black to Move
Here Black settled for a perpetual with 33…Qg5+ 34. Kf1 Qc1+ and so forth. Here, Black could have tried 33…Qh2+ 34. Kf1 Qxh3+ 35. Ke2 Qh2
In this position, White has an extra pawn but the queen and bishop battery actually stop each of White’s pawns from making progress (b8, d6, and f4 are all covered, so promotion is not a threat from White. Black would put his queen on f4 to overprotect f7, followed by pushing the h-pawn. Nakamura would still have a lot to prove, but it’s clear he has nothing to lose.
Wow, what a day! I suspect tomorrow has even greater games in store, featuring a Carlsen–Kramnik clash, as well as Caruana–Nakamura. With the way he’s been playing, I suspect Magnus to hold his lead after four rounds tomorrow, and it will be interesting to see if Nakamura can keep up!