Kramnik's challenge: Anand provides the solution

by Sagar Shah
2/28/2019 – On Monday we told you about a very tough problem which was given to Vishy Anand by Vladimir Kramnik — and challenged you to solve it. A number of GMs and IMs had not succeeded, when Anand gave the problem to the players of Pune. In the end he discussed the moves that needed to be found, and Sagar Shah filmed it all for us to see. Don't miss Vishy showing you the solution.

My Career Vol. 1 My Career Vol. 1

The first DVD with videos from Anand's chess career reflects the very beginning of that career and goes as far as 1999. It starts with his memories of how he first learned chess and shows his first great games (including those from the 1984 WCh for juniors). The high point of his early developmental phase was the winning of the 1987 WCh for juniors. After that, things continue in quick succession: the first victories over Kasparov, WCh candidate in both the FIDE and PCA cycles and the high point of the WCh match against Kasparov in 1995.
Running time: 3:48 hours


"I have an idea"

In our previous article we told you about how Anand gave the students — and the GMs and IMs present — a mate in three problem he had received from his World Champion predecessor Vladimir Kramnik. First a few impressions of the ChampCoach, for which Anand conducted a two-day workshop for young players from all over India that was held in Pune. 

The intro video of Vishy Anand's ChampCoach in Pune

Anand with GMs Abhijit Kunte and Vidit Gujrathi and some participants of the ChampCoach. You can see 92 more group photos like the above on the ChampCoach Facebook page...

... or simply smile at this fantastic overall group shot [click to enlarge]

ChampCoach is a concept where the five-time World Champion and the chess legend of India imparts his knowledge and experience to the young talents of the country. The first edition happened in Pune and was sponsored by Persistent Systems. Players from 22 cities attended the camp. Anand was assisted by Vidit Gujarathi and it was a unique learning opportunity for the budding talents of the country from a world-class player like Anand.

The brochure of the camp

At a press conference Anand spoke on various subjects, like Kramnik's retirement or on school versus chess (what is more important?). These interviews are well worth watching.

Kramnik's chess challenge

First of all we want to thank our readers for the many messages, and for posting interesting comments in the feedback section below the original article without revealing the solution. This was the problem that the readers of our news page were asked to solve:


As you probably know it is possible to move pieces on the diagram above

If you missed the previous article you may want to spend a little time trying to find the solution here. The most important thing to note is that the bishop on h8 is hanging. The rook on a8 attacks it, and we have to find a way to ensure that it is not captured. At the same time it seems as if the rook, bishops on b1 and h8 and the knight on b5 are enough to weave a mating net. Why is there a queen on h1? Think about that — you are competing with three GMs: Vidit Gujrathi, Abhimanyu Puranik and Abhijit Kunte, who are standing around the board in the video below.

And now for the solution. Watch the following video of Anand inaugurating the PYC Hindu Gymkhana training center, and mingling with the students and guests.

At some stage (one minute into the video) Anand pulls out his mobile phone from a back pocket and searches for Kramnik's chess puzzle. At 1:34 he starts to dictate the position to the kids. At 2:50 he tells them this problem is "by courtesy of the newly retired Mr Kramnik." Do not miss the picture on the wall behind Anand: that's Kramnik watching it all of this transpire. You can sit back and watch these pre-teen chess talents work on the position — with Anand guiding them through the different proposed lines. 

At 6:00 minutes one of the students regrets the fact that Kramnik is not there to solve the position, upon which GM Abhijit Kunte offers the service of someone who recently defeated the 14th World Champion: Vidit Gujarathi (who incidentally was also not able to solve the mate-in-three).

An interesting bit arises at 7:47 min, when the students suggest 1.d5. Anand requires almost half a minute to find the refutation: 1...e1, and after 2.a2 (threatening 3.b1#) Black has, as Anand points out, 2...b3, preventing mate on the next move.

At 9:50 Anand shows the students an important try: 1.a3 xa3 2.xb3#, but this is refuted by 1...bxa3 2.b7 b3!

At 10:30 min he finally reveals the solution, which is very difficult to find. This for two reasons: on the one hand the natural variations and their refutations will help you to look for a move which sidesteps the defences mentioned above. On the other you will see that the solution is quite anti-intuitive and cannot really be found at first sight. However, experienced solvers will disagree with this. They are in fact always looking for anti-intuitive moves from the beginning!

MokalIf you watch the video carefully, at 9:20 minutes you will hear a faint voice saying "I have an idea" — and then goes on to give the entire solution. That was IM Prathamesh Mokal, who is a full-time trainer in Pune. After he solved this, some of the people asked him, "Why aren't you playing chess anymore!" Well, if Prathamesh does decide to come back to competitive chess, some of the credit will surely go to Anand, Kramnik and Valentin Marin y Llovet!

And now for the solution: you have to find a drastic queen sacrifice, which is the only way to mate in three. 1.c6!! We discover that after 1...bxc6 the move 2.a3 suddenly works, because the b-file is free: 2...bxa3 3.b8#. Anand also discusses other defences, like 1...e1Q, after which White plays 2.Qc1! and mate on move three.

It is all fascinating to watch, and to make sure you have not missed any of the lines of this remarkable problem we describe them in detail. In the end you can go through them on our game viewer:

[Event "1p, Norwich Mercury"] [Site "?"] [Date "1904.??.??"] [Round "?"] [White "Marin y Llovet, Valentin"] [Black "White to play and mate in 3"] [Result "*"] [Annotator "Sagar Shah"] [SetUp "1"] [FEN "r6B/1p5p/3p2pK/1N5P/1p3P2/3p4/1R2p3/kB1b3Q w - - 0 1"] [PlyCount "5"] [EventDate "1972.??.??"] 1. Qc6 $3 {Such moves do not come all of a sudden! You have to first work with the natural moves in the position, to arrive at a move like Qc6.} (1. Rxb4+ $2 Rxh8 2. Bxd3 Rc8 {and there is no way to mate on the next move.}) (1. Rxe2+ $2 { This was another attempt which looked very logical.} Rxh8 (1... Kxb1 2. Qxd1#) 2. Bxd3 {And now it seems that Qxd1 mate is unstoppable. However, Black has the brilliant resource under his sleeve to delay the checkmate by just one more move!} Rc8 $1 3. Qxd1+ Rc1 {and it will be a mate, but not in three moves! }) (1. Bg7 $2 Rg8 {The rook pursues the bishop and we have simply wasted a move.}) (1. Bf6 $2 Rf8 {Again the rook just follows.}) (1. Bd4 $2 {Looks like a very serious attempt to mate in three.} Ra2 $1 {The only move to spoil White's plans.} 2. Bxa2 (2. Rxa2+ Kxb1) 2... b3 $1 3. Rb1+ Kxa2 4. Ra1# { Yes, it is mate, but in four moves!}) (1. Qd5 $2 {is the most common response from a lot of readers. The idea is to move the rook to b4 with a check, and when Black takes on h8, then to give a mate on a2 with the queen. How to stop such an idea?} e1=Q $1 2. Rxb4+ Qe5 $1 {This move is easy to miss. Now you can no longer mate in three.}) (1. Na3 {also looks very promising.} bxa3 (1... Rxa3 2. Rb3#) 2. Qxb7 Bb3 $1 {A move that is easy to miss.} 3. Rxb3+ Rxh8 4. Rxa3# { Once again, a move too late!}) 1... bxc6 (1... d2 2. Rxb4+ Rxh8 3. Qc3#) (1... b3 2. Rxb3+ Rxh8 3. Qc3#) (1... Rxh8 2. Qc1 $1 {The only move to win.} d2 3. Ra2#) (1... e1=Q 2. Qc1 $1 {This is not an easy move to spot.} Qc3 3. Bc2#) 2. Na3 $1 {The deep point of the study. Now taking with the rook mates with Rb3. But what if the pawn takes?} bxa3 (2... Rxa3 3. Rb3#) 3. Rb8# $1 {And now you see the point of Qc6!! It was to clear the b-file so that the white rook can interfere between the rook on a8 and the bishop on h8. A very pretty solution, don't you agree?} *

We hope you have enjoyed this excursion in the world of artistic chess problems. Want more — like further remarkable problems by Marin? Tell us what you think.

Source: ChessBase India

Sagar is an International Master from India with two GM norms. He is also a chartered accountant. He loves to cover chess tournaments, as that helps him understand and improve at the game he loves so much. He is the co-founder and CEO of ChessBase India website, the biggest chess news outlet in the country.


Rules for reader comments


Not registered yet? Register

Timothy Chow Timothy Chow 3/7/2019 02:01
Very nice problem! Minor quibble: This is strictly speaking not a "study" since the word "study" refers to a position where the stipulation is simply to win or draw, with no limit on the number of moves. Also, it is generally considered an aesthetic plus if the set position in a study looks like it could have arisen in a real game, whereas mate-in-three problems such as this one are only required to be legally reachable from the starting position (though promoted pieces in the set position are usually frowned upon, unless it takes some subtle retrograde analysis to prove that they are promoted).
JimNvegas JimNvegas 3/2/2019 09:33
I'm ashamed to say I blew it. I was certain 1Qxb7 Rxh8 2.Qg7 (renewing the discover check threat on the next move) was the answer. What I overlooked was 2...ed1(Q) and then it can interpose at e5 after 3. Rb3 or Rxb4. Close but no cigar. Teach me not to be so sure of myself. Great puzzle though and thanks for submitting it.
chessisamazing chessisamazing 3/1/2019 06:25
@Amirov 2. Nc3+ - b4xc3
Amirov Amirov 3/1/2019 04:36
1.Ra2+ Kxb1 2.Nc3+ Kc1 3.Qxd1#
KevinC KevinC 3/1/2019 01:05
Let me explain how I got this. In the first article about this, I mentioned that I had just the day before done some mate-in-two problems so I had a little timely practice, and that virtually all mate-in-x problems start with a crazy move or a quiet move, and virtually never with a capture or check.

That immediately led my eye to the crazy move Kg7, knowing that problem creators like to force kings to move, but Rh8 and that try was quickly discounted since next move Rg8+ will take too much time. I then was drawn to the quiet move Qg1 with the idea Qd4 replacing the power of the Bh8, and thought that was it, but Rh8 Qd4 Rc8 cooked that since Rc3 next.

So that left only one more crazy move: Qc6, and one other thing about problems that is a big hint is that when you find one mate-in-x, you know you have found it, so you just have to figure out the rest replies. I quickly saw Qc6 Rh8; Qc1 and either Rc2 or Ra2 was mate in three, so I just needed to look at the rest of the replies, and figure out why Qc6 bc worked.
lajosarpad lajosarpad 3/1/2019 11:03
I thought Qd5 is the solution, but yesterday, during work it sprang into my mind that e1=Q refutes it, since White cannot prevent that Queen from blocking the diagonal for an additional move. I had no time yesterday to look at a board, wanted to do that today, but then saw the solution on the site. Great puzzle!
guitarchess77 guitarchess77 3/1/2019 06:00
Very nice puzzle with the stamp of Kramnik, thanks a lot for share it.
marlondiaz8 marlondiaz8 3/1/2019 01:52
@frederic, thanks :)
durchbruch durchbruch 3/1/2019 01:05
Damn, I actually got it. First I though Na3, but since there was that Bb3 defense I had to figure out how to make it work. And only way to open b file is Qc6. Beautiful :)
Frederic Frederic 3/1/2019 12:36
@marlondiaz8: 2...Bb3 prevents mate on the next move.
marlondiaz8 marlondiaz8 3/1/2019 12:12
Beginner here, why Rxb4, Rxh8 2. Qd5 is not mate, its a diuble treath..
Frits Fritschy Frits Fritschy 2/28/2019 01:30
My first try was 1 Na3, threatening Ra2#, 1... bxa3 (1... Rxa3 2 Rb3#) 2 Qxb7 (2... Rxh8 3 Ra2#/2... axb2 3 Qxb2#), but then black has 2... Bb3. I was already doubting it, because a move threatening a mate in 2 in a three-mover is generally considered too coarse.
Then I looked at 1 Bd4, which seemed too obvious to be the solution, and yes: black had the wonderful defence 1... Ra2! (2 Rxb4+ Rb2! and there is no immediate mate). However, 2 Qd5! Rxb2 3 Qa2# was a nice counter. Too bad 2... Bb3 (or b3) spoils it again. But because it was so beautiful, I was convinced the solution should be connected. Blinded by beauty.
So I tried 1 Qg1, threatening 2 Qa7+ Rxa7 3 Rxb4#. But after simply 1... Rxh8 or even 1... Ra2 2 Qa7 (my initial idea) Ba4! there is no mate in three.
I did look at 1 Qc6, but just couldn’t manage to connect it with my first variation. So I gave up, being reminded that there are things outside chess problems, like work and sleep...