Georg Meier was a move or two away from a sensational win against World Champion Magnus Carlsen in Round 5 of the GRENKE Chess Classic, but with his clock ticking down to under 20 seconds he missed the key move and instead bailed out with a draw. Arkadij Naiditsch had no such time trouble excuse for letting Matthias Bluebaum escape from his clutches, as a rush of blood to the head saw him sacrifice a rook for one of those positions that “had to be mate”, but wasn’t. In the end all five games were drawn, though there were fascinating struggles everywhere.
Watch a brief video of the day’s action:
You can replay all the games from the GRENKE Chess Classic using the selector below – click a result to open the game with computer analysis, or hover over a player’s name to see all his or her results and pairings:
Jan and Peter’s live commentary on Round 5 of the GRENKE Chess Classic also featured an entertaining cameo from Fabiano Caruana’s coach Rustam Kasimdzhanov:
World no. 133 Georg Meier came incredibly close to beating world no. 1 Magnus Carlsen on Thursday in Baden-Baden. He began with his more familiar 1.d4 after experimenting with 1.e4 against Aronian and Caruana, and got a slightly better position out of the opening. Magnus admitted of what he played, “the line isn’t very good”, and concluded of his attempts to play for a win, “I guess I just wanted too much out of a position that doesn’t really warrant it”.
The moment things really sparked into life was when the World Champion went for 26…g5?!, after which he still had 29 minutes to his opponent’s 18:
Obviously 26…g5 is kind of a complete bluff, this whole thing, but it’s very, very hard to evaluate, also with little time. As long as you can’t calculate a forced sequence that gives White a huge advantage it’s not so easy.
There actually was a big move here, 27.f4!!, since the apparent refutation 27…Qh3? fails to 28.Rf2!, while 27…gxh4 28.g4! is close to strategically lost for Black. There was little wrong with Meier’s 27.hxg5, though, and throughout the whole game he played moves of a very high level. That was nowhere more striking than after 31…Ba6!?
This could be a shocking move to face with White in time trouble (Georg was now down to 10 minutes), since after 32.Bxa6?? Qf3+ Black gives mate, while 32.b5 cxb5 33.axb5? Bxb5! loses a pawn. At first glance it seems White should grovel for a draw with a move like 32.Qd2, and it’s easy to see why Magnus went for such a tempting option. He confessed, however, “I actually saw that 31…Ba6 didn’t work, then I played it anyway!”
Meier demonstrated the refutation, 32.b5 cxb5 33.Rh5! and suddenly, with the white rook and queen threatening to coordinate on the 5th rank, it was Magnus who invested 10 minutes trying to figure out what to do next. After 33…Qf7 34.Qe5! he realised he should probably have bailed out with 34…bxc5 Qxd5 and a worse but holdable ending, but as he put it, “it’s very, very sad… I have to suffer”.
In the game Magnus defied his instincts and went for a line where he was winning a pawn, even though, “I knew that this was pretty suicidal”. The move he’d missed was 38.Bg4!...
…and then the critical moment of the whole encounter came after 38…Kh8:
This screenshot is from the powerful Norwegian computer Sesse and, as you can see, it can actually calculate all the way to mate here after 39.Rh5, but at least four more moves are winning, including 39.Rb1, which was the move Magnus was drawn to in the post-game press conference.
Georg Meier was left with 1 minute and 28 seconds, though, and, like Magnus, the main move he looked at was another win, 39.Rh1. Carlsen uttered a shocked, “That was winning as well?” when it was mentioned, with both players having seen the sequence, 39…Qe7 40.Rxh7+ Kxh7 41.Rh5+ Kg6 and decided it didn’t work. In fact Black would be completely winning, if not for the move they’d missed:
42.Rh6+!!, with the main point being simply that 42…Kxh6 43.Qh5# is mate. Georg Meier was the first to spot that, with a wry smile, in the press conference. He summed up what went wrong:
It was Rh6, basically. If I find it, I will win, and if not, I made some move then because I didn’t calculate anything else.
When Peter Leko talked about what was perhaps a once-in-a-lifetime opportunity to beat the World Champion, Georg was more down-to-earth:
If I had 10 minutes and I would miss it, then ok, I would be really disappointed, but these things happen.
He elaborated later in his interview with Fiona Steil-Antoni:
To be honest, Peter was trying to get me more disappointed than I actually am, because it’s not my strongest side to calculate well with seconds. I was actually very close to spotting a line to win, but I missed a key detail, Magnus missed the same idea, then when I missed that I had like 20 seconds left and basically decided to make a draw with Ra1, because if you don’t see how you win then you don’t play without a pawn.
39.Ra1 was the move Georg played with 5 seconds remaining on
his clock, and it was an instant leveller. 39…Qe7 40.Qxg7+ saw the position
liquidated into a drawn ending, with Meier’s draw offer accepted on move 43.
Magnus had perhaps pushed the bounds of acceptable risk, although if Meier had been a little less precise we might have found ourselves talking about another gamble paying off. He was asked by Leko if he was relieved to escape:
Yeah. I’m still in the tournament, for sure. It’s not been brilliant so far, but it’s never too late to start playing well, so I hope to start tomorrow!
Don’t miss Peter Leko explaining the final stages of Meier-Carlsen and also recapping the day’s other action:
Before getting to the day’s other big miss let’s look at the three quieter draws.
Hou Yifan-Vitiugov saw Nikita Vitiugov remain in the leading pack after easily equalising in the Zaitsev variation of the Ruy Lopez. Vitiugov has often worked with Peter Svidler, and up to move 16 he still had a position Peter had played against Maxime Vachier-Lagrave in the 2016 Sinquefield Cup. Hou Yifan sank into a 30-minute think, but her 17.Ng3 didn't look like an improvement on MVL’s 17.Qf3. Soon the women’s no. 1 took the wise decision to liquidate to a draw, not repeating her mistake of drifting to defeat as she did against Maxime in Round 3:
The last four classical games had been decisive between Vishy Anand and Fabiano Caruana, and that sequence could easily have been prolonged in Baden-Baden. Fabiano went for an interesting idea in the Caro-Kann, delaying the recapture of a pawn. It wasn’t clear whether it was home preparation or not, since both players were burning up a lot of time early on. The moment of truth then came after 20…e5:
The critical try here would be for Vishy to hold on to his central passed pawn with a move like 21.Bb5, when 21…Rxb7 22.Bc6 looks promising for White. He may have disliked the counterplay after 21…e3 22.fxe3 Qb6!, though, and instead went for 21.d6 Nxd6 22.Bd5 Nxb7 23.Rxe4, when all the tension had left the position and a draw soon followed.
MVL-Aronian was an Anti-Marshall that flared into life on move 19:
19.Nxe6! Nxe6 20.Qd5! Qxd5 21.exd5 was a nice little tactical shot to win a pawn due to the two unprotected black pieces on the e-file. Had Levon missed it?
I underestimated it, let’s say it like that! I thought I can meet it with 21…Ne5 22.Nxe5 Nd4. It seemed like a great position for me, but then I missed the idea of 23.Nd3 and then 23…Nc2 24.Rxe7 Nxa1 25. Bf4 and White has very, very dangerous play there. Well, but the fact is I have this idea, and I think it should be enough to hold the position – the g5-idea.
It’s easy to see why Levon was worried – White is about to create dangerous passed pawns – but it seems objectively Black is doing fine. No harm was done in any case, since Aronian was right and Plan B also got the job done. Watch him talking about the game, the switch of venue and Vincent Keymer:
Life has been easier for some others who played in the Candidates:
Arkadij Naiditsch’s intentions for this game were clear when he launched an attack with 9.g4, and although our commentators doubted the wisdom of Matthias Bluebaum relying on the well-known rule of thumb and meeting that flank attack with the central thrust 9…e5, that was the computer’s first choice and, up to a point, it seems Black was holding on. Things only began to fall apart when Matthias decided to win a pawn…
19…Bxd3? 20.Bxd3 Rxf3 won that pawn, but at the huge cost of letting White reroute his queen with tempo: 21.Qe2! Rf8 22.Qh5! Ne7:
Naiditsch had 51 minutes to his opponent’s 7, but took only 4 to decide to win “in style” with 23.Rxg7?! Kxg7 24.0-0-0!?. Only after 24…Rf7! did he sink into a 28-minute think as the realisation dawned that there was no killer blow.
Instead 23.f4! in the position above is straightforward and crushing. 23…exf4 is of course impossible due to 24.Qxh6+ and a quick mate, while simply taking on e5 with the pawn is a serious threat. White would retain a huge attack without having to invest any more material. In the game, however, the pain continued for Arkadij. After gathering his strength he managed to adjust to the new situation and play accurately to reach what looked like an easily won ending:
Chess is tough, though, and despite Naiditsch not making any moves that you could class as glaring inaccuracies his advantage soon dwindled to nothing, with Bluebaum doing his part by playing flawlessly from the moment he found himself in a dead lost position. In the end it was actually Matthias who was nominally playing for a win, though there were zero serious winning chances in the rook vs. knight ending. That didn’t stop Spanish GM Pepe Cuenca rising to the challenge of trying to make even that ending exciting with some “fake news” – you can enjoy this quick clip without knowing a word of Spanish!
So despite all the action we got a first day of all draws, meaning the standings were unchanged except for the addition of half a point all round:
Magnus said he wanted to start playing well in the next round, and it would be a good time, since he has White against Levon Aronian. Vitiugov-MVL and Caruana-Hou Yifan might have a lot to say about who’s going to win the 2018 GRENKE Chess Classic.
We respect your privacy and data protection guidelines. Some components of our site require cookies or local storage that handles personal information.