For today’s post, I wanted to discuss a phase of the game in which we are all shamelessly guilty for not studying enough of – the endgame! When I first started playing in 2003, my dad bought me Yasser Seirawan’s Winning Chess series to help me break 1000. The one book I didn’t read? Endgames, of course! What scholastic player needs to know more than rook and king against just a king? No 600 rated player should be bothered with the Lucena position, right…?
Perhaps. But everyone has to start somewhere. I took endgames a little more “seriously” once I broke 1300, trying to work through Van Perlo’s Endgame Tactics, but aside from some cool drawing mechanisms like kamikaze and so forth, I must admit this read was a bit over my head at the time, and endgames once again took the back seat.
While I may have gotten away with a lack of it at the sub-1800 level, having a strong endgame knowledge is extremely important. Let’s take an example from the recent US Women’s Chess Championship:
Nemcova – Bykovtsev
Here Black is faced with a critical decision. Should she defend the a6 pawn, or trade it for White’s f2 pawn? Unfortunately for the youngster hastily played 33… Rc6? (just 34 seconds spent!), making White’s ability to defend the draw much easier! The game continued for 20 more moves, but White traded the queenside pawn and was successfully able to hold the theoretically drawn position.
As it turns out 33… Rxf2! would have made the conversion process a lot simpler. White will have a queenside passed pawn, but the 3 v 1 on the kingside should be enough to win the game and give Black the point.
This game is a perfect example of the importance of endgame knowledge. Black chose not to simplify the position to a more decisive edge, and ultimately paid the price for not being able to recognize the potentiality of the position!
While theoretical endgames are interesting, and in some cases can be challenging to find over the board, its practical endgames that often catch masters off guard. Knowing how to analyze and develop heuristics in practical endgames is crucial to convert advantages into theoretically won endgames. While often times our opponent’s will resign before this happens, we still have to have ideas as to how to reach such positions.
For the next few chess^summit posts, I wanted to discuss the endgame play of one of the strongest player’s in chess history, Magnus Carlsen. Each of today’s endgames are from 2007 when Carlsen was only 16 years old. Don’t worry – he was just a measly 2700! Surely nothing we can’t handle… just ask Aronian!
Our first position comes from the 2007 Candidates tournament, where as a 16-seed, Carlsen managed to take the 1-seed Aronian to a tiebreak match before being eliminated. Even though he lost, Carlsen gave us some nice endgame entertainment!
Carlsen – Aronian
White may visually be better, but the task isn’t easy as it seems. As Gelfand points out in his book Positional Decision Making in Chess, doubled f-pawns can make for stubborn defenders, and in some cases are better than a more fluid pawn structure! Furthermore, it’s not completely clear who will win the battle for the c-file in the near future. After Carlsen played 21. d5 Na5, a simple move like 22. Rac1 might be expected, contesting Black’s hold on the file, but here Carlsen decided to test Black’s hold on the position with 22. h5!.
Carlsen realized that the c-file was not crucial in light of …Na5-c4-d6, blockading the pawn and pressuring the e4 pawn, after which, White would have to move the hypothetical rook away from c1, thus conceding the file anyways. Carlsen follows a more principled approach. Each of his pieces could use some improvement, but there isn’t an obvious square for either of the rooks. With the h-pawn push, Carlsen intends to open a maneuver for his knight, Nf3-h4-f5 to improve his position.
22… Nc4 23.Nh4 Nd6 24.h6!
And this move highlights today’s theme, restricting activity. Carlsen’s intentions are to limit the Black king’s mobility and create potential back rank problems before simplifying the position. With an eventual trade on the c-file seeming to be inevitable, White makes his position as strong as possible before simplifying. While this idea may seem simple enough, Aronian, one of the world’s elite players, manages to find himself lost in just a few moves.
24…Rc3 25.Rac1 Rfc8?
Tactically, an exchange on c3 means that in each line, Black must resolve his back rank issues. Meanwhile, the simplification on c3 makes the passed d-pawn stronger. While the position may have still been tenable, Black is already close to reaching paralysis. For example, the pawn-grabbing 25…Rxc1 26. Rxc1 Nxe4? is punished by 27. Rc7! += as Black is exposed on the 7th rank. White will play Nh4-f5 at the right moment to push the d-pawn, and Black’s rook is confined to the 8th rank as to avoid mating ideas. Starting with 25… Nxe4 isn’t much different. 26.Rxc3 Nxc3 27. d6 is enough to guarantee that White keeps a slight edge.
26.Rxc3 Rxc3 27.Nf5 Nxf5 28.exf5 Kg8
Rooks belong behind passed pawns, but here has Aronian tried 28…Rd3? 29. Rc1! is catastrophic. While this move is a step in the right direction, Black still needs one more move before his back rank issues are solved. With this tempo, Carlsen takes affirmative action by creating another passed pawn.
29.Re4 Kf8 30.Rg4 Rc7 31.Rg7 +-
We’ve already talked about how the h- and a- pawn can be extremely useful, and this is another great example. Carlsen realized that the d-pawn alone would simply not be enough, and in turn created another weakness. Black’s king once again is relegated to passivity as it must stay in the corner of the board.
31…b5 32.Rxh7 Kg8 33.Rg7+ Kh8 34.d6
Extremely precise. White intends to bring in the king (34. Kf3 is fine as well), but this move forces Black’s rook to take a more passive route first, giving the White king the time to reach e4. As we will see this move will eventually force Black off his own 2nd rank, giving White the opportunity to take on f7 and have more luft for his rook.
34…Rd7 35.Kf3 b4 36.Ke4 Rxd6 37.Rxf7 Ra6 38.g4!
The final straw. Carlsen’s intention is to march the g-pawn and force Black to take it on g5. Once this happens, White will have a route for his king via e4-f5-g6 and will win the game.
Not so fast mister… that king will stay on h8! Black is powerless to stop any of White’s ideas.
39…Kh8 40.g5 fxg5 41.f6 1-0
Aronian throws in the towel here as Black will get checkmated once the White king reaches g6. Arguably, the theoretical endgame you needed to know was rook and king against king, but the hardest part of the game was over after White was able to limit Black’s king. Even though we don’t think of the king to be an important attacking piece, here we can just see a visual comparison between the two sides’ monarchs.
It turns out that Carlsen felt like he needed to give Aronian two doses of the same lesson, and prescribed the same shot in the same match!
Already we have a good knight against bad bishop endgame, as Aronian’s c8 bishop can’t pressure any of Carlsen’s weak pawns. We can also already see that White has already limited Black’s king, but it’s not really enough right now to win the game.
28. Rb6 Ra3 29. Rc1 Be6 30. Nf3
Just like the last game, White improves his position before taking drastic measures. From f3, the knight has a lot more scope, but that’s only the first half of his idea…
30…Rfa8 31. h4!
This is a moment that separates Carlsen from the rest. Once again we see the h-pawn push, this time with the intention of opening the g-file. It’s not enough to win, but Carlsen continues to press for small advantages.
31…h6 32. Ne5 Ra1?
Once again, we see a bad rook trade from Aronian. While his passive king problems may not be as pronounced, it’s another small concession that pushes the scale in Carlsen’s favor. The win of the c3 pawn is forced, but as Black will discover, it comes at great cost.
33. Rxa1 Rxa1+ 34. Kh2 Ra3 35. Rb8+ Kh7 36. f4!
Now is not a time to play with fear! With this move, Carlsen now has two options to prod Black’s kingside. Watch how quickly Carlsen is able to take advantage of the weak f7 pawn. Given how quickly Black collapses, I’m guessing that even Aronian didn’t sense the true danger to his king.
36…Rxc3 37. h5!
Once again, Carlsen refuses to stand down. The insertion of this move is necessary to make a future f4-f5 push work. As Grandmaster Alexander Fishbein told me shortly after beating me at last spring’s Pittsburgh Open, it’s not about the number of pawns, it’s about how active your pieces are. At the conclusion of this line, Carlsen will have given three pawns, just to win the pawn on f7. As it turns out, this same pawn was Aronian’s Achilles’ heel.
37…gxh5 38. Rf8 Ra3 39. f5! Bxf5 40. Rxf7+
Carlsen achieves his dream position. Now with a rook on the 7th and a knight in the center, White’s pieces are much more active than Black’s, but more importantly, the f6 pawn is now the most dangerous passed pawn on the board.
40…Kg8 41. Rg7+ Kf8 42. Rb7
I feel like here, a pressured amateur may play 42. Rc7?! in an attempt to follow the dogmatic approach of putting the rook behind the passed pawn. But here, it’s much more important that White has the ability to threaten mate on the 8th rank, thus giving him more mobility. Because White controls the c7 square, and Black has no way of easily inserting his own rook behind the passed pawn, the fact that the c4 pawn is passed is not of concern to White just yet.
42…Ra8 43. Kg3 Rd8
On a surface level, 43… Rc8 may seem to produce more counterplay, but in light of 44. Kf4 Be4 45. Rf7+ Ke8 46. Re7+ Kf8 47. Nd7+ (with mate looming), Aronian had to play the text move to be able to take the knight on d7 should Magnus opt for this line. This concession alone shows the importance of an active king.
44. Kf4 Be4 45. g3!
Not rushing into things! White has “all” the time in the world to convert, so here he spends one tempo to avoid 45. Rf7+ Ke8 46. Re7+ Kf8 47. Nd7+ Rxd7 48. Rxd7 Bxg2 and while still losing for Black, White still has a game to win. If it’s a won endgame either way, the safer road is the road best taken!
45…c3 46. Rf7+ Kg8 47. Rg7+ Kf8 48. Nd7+
And now Black must give up the rook in light of 48… Ke8 49. Re7#.
48…Rxd7 49. Rxd7 1-0
Another great game by Carlsen! In just one tournament, we see two different demonstrations of how limiting the king’s mobility can lead to big problems in the endgame. For Aronian, the results were drastic, as in each loss he quickly found himself in a mating net thanks to the superb coordination of White’s army.
For our last endgame of today, I’d like to discuss a game where Carlsen combined a the positioning of a weak king with the opponent’s temporary limitations of a bad rook to win a nice World Cup game against Cheparinov. Winning this game gave Carlsen a berth to the semifinals, where he eventually lost to the eventual winner, Gata Kamsky.
Carlsen – Cheparinov
Already, this position poses a different argument than the last two. With a material imbalance, White has two minor pieces to Black’s rook, but given the simplified nature of the position, that should not be enough to win it. Upon deeper observation, it should be noted that White does have a passed pawn (Black’s pawn on e7 is so far back that it’s not relevant in our assessment of passed pawns) on g6, and the promotion square does match our bishop. Of course, the biggest note is that Black’s rook, while seemingly attacking our pawns, is actually offsides and out of play. Already we can see that Black will need to spend some time to get this rook to a more appropriate square. In a position that the computer claims is only slightly better for White, Carlsen shows that the position is quite rich with the move 36. Bg2!
This move comes with the short term idea, 36… Rxa2?? 37. Bd5+ +-, but of course there’s more to it. White puts his bishop on a much more active diagonal to assist his g6 pawn. While Black may eventually win White’s queenside pawns, again it will be White’s activity that will determine the outcome of this contest.
36…c4 37. a3 Rb1 38. Be4 Rxb2+ 39. Nc2
Almost mocking Black’s rook for its passivity. Black may have won a pawn, but here we already get the sense that something has gone quite wrong. White’s created a cage for Black’s only piece, giving him even more time to improve his position.
39…Kg7 40. Ke3 Rb3 41. Kd2
A backward looking move, but now that Black’s rook is on b3 and not b2, it makes more sense to come back to d2 since there is no pin along the second rank. By not going to d4, White leaves a rite of passage for his knight to aid the g-pawn.
41… Kf6 42. Nd4 Rxa3 43. Nxb5 Ra5 44. Nc7
Even outside of the box, Black’s options for his rook aren’t exactly impressive. Meanwhile, White’s knight will enter the fray via d5 or e6 (or even e8!) and begin to limit Black’s king.
44…Kg7 45. Ne6+ Kh8
Carlsen must have been extremely satisfied to have pushed his opponent this far. Already the king is forced to the h8 square, many thanks to White’s light-square domination. Should Black have tried 45… Kg8?!, he may have found it uncomfortable to defend 46. Bc6!
The idea is that White can play Bc6-e8-f7 and improve the position with check, or simply wait for Black to move the rook and occupy the d5 square. Cheparinov chose to avoid such lines, but without a bishop, he can’t ignore these problems forever…
46. Ke3 Ra1 47. Kd4 a5 48. Bc6
Even with the a-pawn marching, Black is still powerless to stop White from creating a mating net. After all, a checkmate is much more valuable than having an extra queen!
48…a4 49. Be8 Rg1 50. g5!
The critical move Carlsen must have seen moves ago! The idea is that should Black take the pawn with …h6xg5, the file closes and White will promote the g-pawn.
50…a3 51. Bf7 Rxg5
Naturally this is losing, but Black, as it’s been for the duration of the endgame, has not had the luxury of choosing. With no other way to stop the g-pawn from promoting, Black throws in the rook.
52. Nxg5 hxg5 53. Bxc4
Given the nature of the World Cup, Cheparinov played for a few more moves to avoid elimination, but a lost position is lost. White simply weaved the king over to pick up the a3 pawn, then brought it back to stop Black’s pawns.
In just three positions, we’ve come to learn the importance of restricting enemy pieces (especially the king!) in the endgame, and how such technique can give us the time we need to build our positions to full strength. As Carlsen showed, the king is an important attacker, and its a critical part of the fight too!
For my post later this week, I’ll discuss more endgames of Carlsen on his road to becoming World Champion. Don’t miss it!