What's Going On?

I HAVE played club chess since 1976, and tournament chess since 1980.

Up until 2023 I had faced the sequence 1.c4 e6 2.Nc3 d5 3.e3!? twice.

Position after 3.e3!?

The first time was in the Middlesex League in 1991; the second at the 1994 Hastings weekend congress.

So far, nothing extraordinary, which is not surprising as 3.e3!? is the type of move novices are warned against, not least because it blocks White's dark-square bishop.

The next time it was played against me was in the Central London League two years ago, but this year I have already faced it twice - at Mariánské Lázně and in the London League.

Has the move been recommended in a repertoire book, or perhaps a magazine article?

I try to keep abreast of chess literature, but I by no means see everything.

There are 2,043 examples of the diagram position in ChessBase's 2025 Mega database.

Last year saw 202 of those games, or 9.9%.

That is a high share for a single year, especially as the games in Mega do not cover December, November or most of October.

The last full year, 2023, saw the position occur 214 times, or 10.5% of the total, 172 times (8.4%) in 2022, 119 times (5.8%) in 2021 and 105 times (5.1%) in 2020.

Clearly 3.e3!? has been growing in popularity in the Twenties.

The move goes back to at least 1860, when Steinitz played in at Vienna, but in Mega25 it makes only one more appearance in the 19th century, before being played sporadically between the world wars.
Ian Nepomniachtchi (2774) is the highest-rated player to have tried it, although Hikaru Nakamura (2780) has reached the diagram position with white by transposition.

There are famous names among those who lost on the black side, including Fabiano Caruana and world champion Dommaraju Gukesh (admittedly five years ago when he was 'only' 2563).

So what is the antidote?

According to Stockfish17 and Dragon1, there is no need for an antidote - they reckon 3.e3!? hands Black equality.

In other words, blacks are more-or-less free to play whatever takes their fancy, but do not copy Gukesh's 3...f6?. which should have lost a pawn on the spot to 4.cxd5 exd5 5.Qh5+ etc (it was an online blitz game, and White, Armenian grandmaster Haik Martirosyan, won anyway).

Caruana's 3...Nf6 is much more reasonable, even though he lost with it to Nakamura, again at blitz.

It is easily the most popular continuation, accounting for 80.9% of games in Mega25, and scores an excellent 51%.

Naturally it is the move White will be expecting, and the commonest reply is 4.Nf3, when 4...Be7 is a normal continuation in such positions, although the engines' top choices are the more-active 4...Bd6!? and 4...b6.

Somewhat surprisingly the natural 4.d4 scores a miserable 41%, the engines liking 4...Be7 and, again, 4...b6.

Perhaps a more promising try for White is 4.b3!?, which objectively is surely no better, but scores 55% in Mega25, probably due to blacks being taken out of their comfort zone.

Here is a rather drastic example of what can happen if Black is not alert.

David Howell (2677) - Jonas Wyss (2272)

Chess.com Blitz 2023

1.e3!? d5 2.c4 e6 3.Nc3

Reaching the above-diagram position, and now Wyss plays the most popular response.

3...Nf6 4.b3!?

How should Black proceed?

*****

*****

*****

*****

*****

4...Nbd7

A typical QGD continuation. The engines like the popular 4...c5 and the rare 4...d4!? After the latter they recommend meeting 5.exd4 with 5...Nc6!?, claiming equality.

5.Bb2 b6

This is the engines' top choice.

6.Rc1!? Bb7 7.Nf3 Bd6?!

This may look like a move that combines activity (the bishop is more aggressively placed than on e7) with defence of the vulnerable c7 square, but the engines reckon it is a mistake.

How can White exploit the defects of Black's last move?

*****

*****

*****

*****

*****

8.cxd5 exd5

The engines prefer giving up a pawn with 8...Nxd5!? 9.Nxd5 Bxd5 10.Bxg5.

9.Nb5 Nc5 10.Nxd6+ Qxd6 11.Be5

White has won the bishop-pair, and has a continuing initiative.

11...Qd8?!

Probably better is keeping the queen developed with 11...Qe7.

12.d4 Ncd7

Or 12...Ne6 13.Bxf6! and 14.Bb5+.

13.Bxc7 Qe7 14.Bd3 0-0 15.0-0 Ne4 16.Bg3 f5?

Trying to change the narrative, but missing that White's last was not just a safety-first move.

17.Rc7 Rab8 18.Ne5 (1-0, 22 moves).