Nova Daily - 14 April 2026: The Keres English—Part I

Hi!

First of all: congratulations to Javokhir Sindarov for winning the Candidates with a round to spare! I know that you're highly unlikely to read this, but in case you do: best of luck against Gukesh Dommaraju!

The Keres English—Part I

Anyone who plays a kingside fianchetto variation of the English will at some point have to contemplate what would happen if black decided to blunt the bishop's scope along the long diagonal with c7-c6. I play such a line with black against the English myself: it's intuitive, and you get a lot of reps in with the same types of pawn structures. I like to believe that that gives me a better understanding of the intricacies of these structures, and for this reason alone I could advise myself to play the London System. Cyrus Lakdawala would be in favour of such an approach. However, I'm not confined to England's capital only. Chess is perhaps the only realm in which London has nothing to do with the English, and I don't see it as my life's task to take it upon myself to bridge them.

The variation 1.c4 e5 with a quick ...c6 is named after the most famous Estonian chess-player that's ever lived. I'd be able to write a lot of material on him, but I have absolutely no incentive to do so. And luckily, I don't need to. As it happens, FM Jimmy Adams has already penned together a comprehensive biography about him, totalling 4000 pages. I must say that even for Nova Stone standards, that's perhaps a little too much information to relay in a single blog. Adams agreed, and he spread out his work over four different volumes.

And, you know what: so will I. But not about Keres himself, but about the variation of the English that's named after him.

What separates the Keres variation from Caro/Slav variations is the most obvious thing: black's pawn is already on e5. In one way that's a strategic accomplishment for black: the pawn is functioning as an active central pawn and takes aim at the dark squares in the centre. On the other hand, combining it with c2-c3 has a dynamic drawback. Of course it's quite advisable for either side to take a lot of space in the centre early on. However, doing so too quickly can leave your centre somewhat overextended. I've dealt with this in some form in my pawns blog.

The literature

As indicated above, any coherent repertoire that suggests that white starts with a combination of c2-c4 and g3/Bg2 will have to account for black setups that involve e5 and c6. The start of my literature is Marin, complemented with Dvirnyy. In order to work on these lines, I don't have to consider The Iron English or anything that Ntirlis has written about the English, because they both recommend a move-order starting with 2.Nc3. Georgiev and Semkov do the same in The Modern English, as does Commings in Opening Repertoire: The English.

I'd like to mention here that in working through these lines, I've been considering that it may be much easier to go with 2.Nc3 and circumvent this entire theoretical complex. But I've been lazy for long enough.

Basic strategy

Black's setup is very reminiscent of the Alapin Sicilian, and this is one of the variations in which it's productive to borrow a few strategic concepts from the analogous Sicilian variation:

Black has one pawn in the centre. If allowed, black will push d7-d5 and obtain the ideal central formation: pawns on e5 and d5. Those pawns would constitute a solid space surplus, which yields black an advantage. In order to not let that happen, white has to act immediately to counter this central configuration.

In the Alapin Sicilian, black generally has two ways to counter this setup. They are both rooted in the same strategic drawback that the c2-c3 move involves:

In testing both methods for the analogous English variation, there is one thing that's different: white has already committed the king's bishop. This matters greatly for the move-choice: in the 2...Nf6 Alapin variations, black's kingside pawn structure is intact, and the bishop typically develops to e7 (or in some lines c5 or b4). Since the bishop is already committed here, that development is off the table, and it's up for debate whether the move spent on pushing g2-g3 constitutes an advantage to white:

That leaves white with one reasonable option to challenge the e5-pawn: d2-d4. Of course, both Marin and Dvirnyy suggest this move.

There are two things that need to be considered for white in order to account for the playablility of this move. The first is the obvious one: the recapture with the queen is fine, because thanks to the c6-move, the knight is unable to develop to its natural square and kick the queen away. Black being fine with the queen developed early is confirmed by the statistics: over 20 000 games in this variation, black scores 49%, which is as close to 50/50 as you may reasonably expect.

The second point is that this setup has to be compatible with having played g2-g3. That's actually quite easy to do if we approach the position strategically. Here's a schematic overview:

The above discussion shows that white's setup with d2-d4 is indeed compatible with the fianchetto of the king's bishop. This is confirmed by both engine evaluation (+0.32) and statistics (in exd4 lines, white scores 57% over 198 games).

Black's responses

While most theory books order their material based on the move-orders (which makes sense for looking things up quickly), I don't intend to do this. Instead, I'll divide the material by theme.

Overall, there are three ways for black to meet the d2-d4 thrust: push e5-e4, capture e5xd4, or defend e5 with d7-d6. In essence, this would give me three sibling variations to study. In each case, there are two separate details that need to be accounted for: whether the knight is on f6, and whether the bishop is out to b4.

Because of this, I've decided to separate the work over multiple sessions. Today will be the first one of these: variations with e5-e4. In each case, I'll attempt to answer a few questions:

• What is the overall plan for both sides based on the pawn structures?
• Is Nf6 advantageous to either side?
• Is Bb4 advantageous to either side?
• Is the combination of Nf6 and Bb4 advantageous to either side?

Each of the three different posts will have to deal with these questions. Hopefully, this research and analysis will clarify for me how things run, and whether (and when) which details will be favourable for either side.