The gloves have now come off in the World Championship Match at the Fulton Building in midtown Manhattan, as game 3 between Magnus Carlsen and Sergey Karjakin turned into an epic marathon of a Berlin Defence in the Ruy Lopez - and one that looked as if Carlsen was going to strike the first decisive blow of the 12-game match, only for his ever-resilient Russian challenger managing to somehow hang on for a draw with a set of survival skills that would have made Bear Grylls envious.
This was the first Berlin of the match - and there could be more to come. But in this one, after a few inaccuracies at the critical moment from Karjakin, Carlsen found a crack and all but looked ready to break down his challenger’s Berlin Wall…but each time it looked on the verge of collapse, the world champion couldn’t quite find the winning blow (if there was one), and Karjakin survived time and again.
And after nearly seven hours of play and over 70 moves, the game ended in a draw. "You have to fight for every last thing there is, even if it takes energy," Carlsen said after the game. "I didn't see a forced win. "Certainly it's disappointing.” It could well have been a disappointment for Carlsen with a lot of missed opportunities, but it could also signal the start of a change in the tempo for the match with more long games coming from the world champion in an effort to try and wear down his challenger.
Everybody says they know where they were the night the Berlin Wall toppled 27-years-ago on November 11, 1989 - but I can go one better than that, because 16-years-ago to the month in London, I witnessed from the pressroom at first-hand the shock re-building at elite-level of the chessic version of a Berlin Wall that all but bamboozled a world champion into losing his crown.
The Berlin Defence in the Ruy Lopez was popular during the late 19th century and early 20th century - but it fell out of fashion for the best part of a century in elite praxis, only for Vladimir Kramnik to adopt it as his big surprise drawing weapon (8 games!) against Garry Kasparov in their World Championship Match in London 2000. Kasparov kept banging his head against what became known as the 'Berlin Wall' endgame (after 4 0-0 Nxe4 5 d4 Nd6 5 Bxc6 dxc6 6 dxe5 Nf5 7 Qxd8+), believing he could break it down; couldn’t, and memorably lost the title to his protégé.
Every chess journalist thereafter dubbed it the ‘Berlin Wall’, and everyone claimed dibs on being the first to refer to it as such - but in fact I remember as a teenager reading a very entertaining article by the first chess writer to do so, namely Jimmy Adams in his entertaining article ‘Breaking down the Berlin Wall’ that appeared in the August 1979 issue of British Chess Magazine.
Match score (best-of-12-games)
Carlsen 1.5-1.5 Karjakin
Magnus Carlsen - Sergey Karjakin
World Championship, (3)
Ruy Lopez, Berlin Defence
1.e4 e5 2.Nf3 Nc6 3.Bb5 Nf6 4.0-0 Nxe4 5.Re1 Nd6 6.Nxe5 Be7 The continuation Jimmy Adams showed in his aforementioned BCM article on the Berlin was built around (if I remember correctly) an old Steinitz favourite attacking line that went 6...Nxe5 7.Rxe5+ Be7 8.Bd3 and following up with adventurous ideas of Nc3, b3, Bb2 and launching a fierce kingside assault. All wonderful stuff - but not in a critical World Championship Match! 7.Bf1 Nxe5 8.Rxe5 0-0 9.d4 Bf6 10.Re2!? Not a novelty, but a rarity that set Karjakin thinking. The standard move here is 10.Re1 played about 650+ times now - but 10.Re2!? has been played only five times before, and along with the idea of 11.Re1, this is a very subtle (I believe) suggestion by Anish Giri published a while back in New in Chess magazine. 10...b6 And it does its job of setting Karjakin thinking what was the difference between 10.Re1 and 10.Re2? - and this psychological side in a match of playing with an opponent's mind can often be effective. And here, at least, it nothing else it had Karjakin pondering for almost half an hour before he played 10...b6 - that's 30 minutes that contributed to the Russian's time pressure throughout the rest of the game. I suppose after 10...Re8 (the reply to 10.Rel), Carlsen would have exchanged on e8 as normal, and would have gained a little psychological advantage of having made Karjakin think for that 30 mins. 11.Re1 Re8 12.Bf4 Rxe1 13.Qxe1 Qe7 Carlsen is taking a minimalist approach here. His simple idea is that, in certain lines where he exchanges queens, he could follow-up with Bxd6 and Nb5 attacking the bishop on d6 and perhaps the pawn on c7. Not an easy position, but with careful play Karjakin avoids the major pitfalls. 14.Nc3 Bb7 15.Qxe7 Bxe7 Ever the grinder, Magnus finds another little annoying move, threatening Nb5, and if ...Nxb5 then axb5 and Black has a weak a7-pawn that Black's rook will have to permanently defend. It's not big moves that Magnus beats you with - it's all his little niggling moves that soon build-up to a major problem. 16.a4 a6 17.g3 But Karjakin is more than up to the challenge, as he's one of the toughest and most accurate defenders in the game today. And here, I think Magnus overlooked that Karjakin could play 17...g5 that forces him to exchange off his bishop for the knight before he wanted to. Now the position goes a bit sterile again. 17...g5! 18.Bxd6 Bxd6 19.Bg2 Bxg2 20.Kxg2 f5 21.Nd5 Mildly annoying for Karjakin, as the knight can re-route to e3 and perhaps later Nf5 or Nc4 with a good outpost. 21...Kf7 Not 21...f4? 22.gxf4 gxf4 23.Kf3 Rf8 (the simple threat is a5! bxa5 followed by c5 and the bishop will be lost) 24.c4! a5 25.Rg1+ Kf7 26.Rg4 and suddenly White is winning a pawn - and with it, the game! 22.Ne3 Kf6 23.Nc4 Bf8 24.Re1 Rd8 25.f4 gxf4 26.gxf4 b5 27.axb5 axb5 28.Ne3 c6 29.Kf3 Ra8 30.Rg1 Ra2?! The first real crack in the match comes from Karjakin with an inaccurate move. Instead, Karjakin missed a finesse that would have given him equality after 30...Bh6! (to stop Rg5 hitting the f5-pawn) 31.Rg3 Ra4 32.c3 Ra2! and a better version of what Karjakin gets in the game, as now b3 would be weakening, and Black would have a very active rook on the seventh. Now the only logical continuation would be 33.Rg2 Ra1 34.Rg8 Rh1 and White can't make any progress with Black's rook being so actively placed behind all the pawns. 31.b3 Now it all becomes logical for Carlsen: b3 stops the annoying ...Ra4 and his knight multi-tasks by defending c2 and hitting f5. 31...c5 32.Rg8! Vintage Carlsen. A whole lot of (apparently) nothing, and then boom! And this cuts across Karjakin's cunning drawing plan, as quickly spotted by GM Peter Svidler on Chess24.com, of 32.d5? c4! 33.bxc4 Bc5! 34.cxb5 Ra3! 35.Re1 d6! and, despite being two pawns ahead here, White can never get out of the eternal pin on his knight on e3, as even after Kf2 it's still a pin, this time with the bishop on c5(!). This means that White's rook, knight and king are forever tied together and he can never win, despite the big material advantage. So kudos to the 7-time Russian champion for quickly spotting this without the aid of a playing engine! 32...Kf7 33.Rg2 cxd4 34.Nxf5 d3 35.cxd3 Ra1 36.Nd4 b4 37.Rg5 Rb1 38.Rf5+ Ke8 39.Rb5 Rf1+ 40.Ke4 Re1+ 41.Kf5 Carlsen has emerged at the other end of the first time control with a wonderfully centralised position, and 9 times out of 10, you would see him squeeze his opponents to death - but Karjakin is more resilient than most players. 41...Rd1 42.Re5+ Kf7 43.Rd5 Rxd3 44.Rxd7+ The human factor here instinctively takes the pawn with check, not that it's all bad. But with a little forethought from Carlsen, with Karjakin short of time going into the second session, arguably more awkward is the try 44.Ke4!? with the idea being that if 44...Rh3 45.Nf3! and White is defending h2 and has now made it all a tad awkward for Black with his rook cut off from the game now. It doesn't necessarily win, but it continually would have kept Karjakin thinking and worrying. 44...Ke8 45.Rd5 Rh3 46.Re5+ Kf7 47.Re2 Bg7 48.Nc6 Rh5+ 49.Kg4 Rc5 50.Nd8+ Kg6 51.Ne6 h5+ 52.Kf3 Rc3+ 53.Ke4 Bf6 54.Re3 h4 55.h3 Rc1 56.Nf8+ Kf7 57.Nd7 Ke6 58.Nb6 Rd1 59.f5+ Kf7 60.Nc4 Rd4+ 61.Kf3 Bg5 62.Re4 Rd3+ 63.Kg4 Rg3+ 64.Kh5 Be7 65.Ne5+ Kf6 66.Ng4+ Kf7 67.Re6 Rxh3 68.Ne5+ Unbelievably, Karjakin manages to hang on for dear life to salvage the draw - and escapes such as this is one of the reasons why he is one of the toughest players to beat and a worthy challenger for Carlsen. 68...Kg7 69.Rxe7+ Kf6 70.Nc6 Kxf5? The correct way to save this was with 70...Rc3! (Not 70...Rxb3?? 71.Re6+ Kf7 (71...Kxf5 72.Nd4+! wins the rook.) 72.Ne5+ is a forced mate in 6!) 71.Re6+ Kxf5 72.Nd4+ Kf4 73.Kxh4 Rd3 and it's a draw, as White can never defend the b3-pawn successfully. 71.Na5? Mutual fatigue perhaps in the marathon? After nearly seven hours of play and 70 moves of an epic struggle, both players missed their moments. Here, correctly pointed out by Svidler on Chess24.com, was for Magnus not to get greedy by going for another pawn, but the more testing 71.Re1! and his rook is ideally placed here to cut Karjakin's king off and defend against the h-pawn running. So now, if 71...Kf4 72.Rf1+ Ke4 73.Na5 suddenly it all awkward again for Karjakin, who was still very short of time. Perhaps someone will tell us exactly what's happening, if it is indeed a win or just a draw - but facing Magnus with this over the board and with little time left, it could well have been fatal for Karjakin. 71...Rh1! The rook not just allows the h-pawn to run, but also shunts the Na5 that protects the b3-pawn. Carlsen can't simultaneously stop the h-pawn running and the knight protecting the all-important b3-pawn. 72.Rb7? Again, allowing Karjakin to easily salvage the draw. Carlsen could have kept the powerplay and the pressure on with 72.Rf7+ Ke6 (If 72...Ke4 and White is making progress with 73.Kg4) 73.Rf2, and in both cases, White's pieces are all working together here - certainly still a lot of work left to save this. 72...Ra1 73.Rb5+ Kf4! The h-pawn now is the important pawn for Karjakin. 74.Rxb4+ Kg3 75.Rg4+ Kf2 76.Nc4 If 76.b4 then there's no way to stop the h-pawn without avoiding a repetition: 76...h3 77.Rh4 Kg3 78.Rg4+ (If 78.Kg5? h2 79.Rxh2 Kxh2 80.Kf4 is also a draw - but it would have given Black the psychological bragging rights to ending on a high with the better position!) 78...Kf2 79.Rh4 Kg3 etc. 76...h3 77.Rh4 Kg3 78.Rg4+ Kf2 ½-½