The Extraordinariness of the Ordinary
It's amazing what you can find when you dig deeply.
The European Individual Championships were contested last week, and Aleksandar Indjic was the surprise winner. The 29-year-old Serbian is a fine GM, but with a number of current and past 2700s participating Indjic, only the 33rd seed, was not a favorite. It didn’t matter: he raced out to a lead with 6.5/7 and then 8/9, and was able to coast home with draws in the last two rounds to finish in clear first with 9/11, half a point ahead of the even more surprising Daniel Dardha and a point ahead of a large group (11 players) with 8 points.
I didn’t watch a lot of the action, but I did get to take a bit of a deep dive into Indjic’s round 10 draw with Daniil Yuffa (one of the players who finished with 8/11). I started watching here, after White - Indjic - played 23.Bf1-c4.
White threatens to take on e6, so Black played 23…Rd6, and after 24.Qf4 swapped on c4 and retreated the rook from d6 to d8, producing this position.
This is where it got fun, at least for me. Indjic spent 15 minutes on this move, and so I dug in as well, first without the engine and then checking with it. You should dig in yourself, and then keep reading. I’ll put in an ad break to help you avoid spoilers:
(And please do become a yearly subscriber, as that will automatically get you my extra coverage of the Ding vs. Gukesh World Championship match which starts tomorrow (Monday). Subscribers will directly receive a fuller analysis than I’ll include for the blog, plus video recaps of each game as well.)
Ad over; back to the game:
Alright, the first thing you should notice is that the otherwise desirable 26.bxc5 (or 26.b5, for that matter) is a blunder: Black wins the queen for a rook or a rook for nothing after 26…Rd1+!
That might lead you to consider a move like 26.Kg2, so that …Rd1 won’t be a check. Black has multiple good replies against this, including especially 26…cxb4. White should play 26.Qxb4, though it leads to a dead draw after Black trades queens and rooks and then attacks White’s b-pawn. You might think of 26.Qxc8 instead, but that’s a serious, possibly losing error, as 26…Rxc8 27.Rxc8+ Kg7 leaves White with a serious problem due to the pinned a-pawn, which will soon be lost. So 26.Kg2 is playable, but most likely a draw.
In the game, Indjic found an even more convincing way to make a draw: 26.Qc2. Now the …Rd1+ trick is off the table, so White genuinely threatens to take on c5. The players called it a day here, as neither side has any promising way to avoid a line like 26…Qxc2 27.Rxc2 cxb4 28.Rxc8 Rxc8 29.axb4, when 29…Rb8 30.Rxa7 Rxb4 is as dry as dust.
The draw suited both players, for different reasons, and there’s really nothing to protest here as the position is objectively equal. That said, going back to the last diagram, White had a really interesting option. I didn’t spot it and it took me a minute or two to get it, but once I did I loved it. It’s 26.Rab1!!
At first sight, it’s a strange move. White hangs the a-pawn and still doesn’t threaten to take on c5, so what’s the point? Just one point, really, but it’s a big one: it takes …cxb4 off the table. Now that the rook is no longer on a1, 26…cxb4 is met by 27.Qxc8 Rxc8 28.Rxc8+ Kg7 29.axb4, and now Black does have something to worry about. White’s basic idea (which isn’t guaranteed to succeed) is to swap the queenside pawns, double up on Black’s f-pawn and trade into a winning king and pawn ending. (Some of you might remember game 1 of the Kramnik-Leko World Championship match back in 2004. That’s the general idea.)
Instead of 26…cxb4, Black should prefer 26…a5, and after 27.Rc3 White does threaten at last to play 28.bxc5. So Black might now liquidate: 27…cxb4 28.Qxc8 Rxc8 29.Rxc8+ Kg7 30.axb4.
Now what? 30…axb4 will be met by 31.Rc4 or 31.Rb8 followed by regaining the pawn and then going for the 7th rank. So your task, if you choose to accept it, is to analyze from here. How should Black defend this ending? It can be done, and I expect Yuffa would have managed the task. Still, if I had been playing White in this game and saw (and understood) 26.Rab1, I’d have made Black prove that he could save this. When you’re ready to see my analysis, have a look here. (Note that I include Leko-Kramnik there.)
As you can see, the position was much deeper and more interesting than one might have suspected, and instructive, too, as endings with queens against a pair of rooks come up from time to time. (In fact, I had such an ending against GM Bryan Smith a decade or so ago, and had the queen vs. the two rooks. Remembering the Leko-Kramnik game, I was frankly relieved that he offered a draw rather than torturing me in that ending. Maybe I would have held…but I’m glad I didn’t have to find out.)