In the Kmoch Nimzo Indian (sometimes called the f3 nimzo)

[FEN ""]  
1. d4 Nf6 2. c4 e6 3. Nc3 Bb4 4. f3 d5 5. a3 Bxc3 6. bxc3 c5 (6... dxc4!? 7. e4)

The move 6...c5 is the most popular continuation at master level according to Lichess's master database by far. However, stockfish "cloud" analysis lists 6...dxc4 as the top choice, which less than 1% of masters chose (and scored poorly). Since this is a cloud evaluation it is presumed to be deep analysis done beforehand, not a shallow depth search running in a browser.

At first glance, it looks like the cloud analysis is simply wrong. dxc4 allows the immediate e4, which makes white's opening look like a strategic success! On the other hand, it seems unlikely to me that a high depth engine would make such a positional error.

Is it possible to say that dxc4 is a valid continuation that hasn't been explored? Or is this simply a case where an engine misevaluates a position?

  • 2
    I guess sometimes masters aren't comfortable in certain positions that an engine is comfortable in. After 6...dxc4 7.e4 b5 8.a4 c6 9.Ba3 I'd rather be White. Commented Feb 7 at 17:23
  • 1
    Knowing what I know about engines, I would say it sees the opportunity to create an isolated a pawn weakness as a larger advantage than the central disadvantage. Also, having pushed f3, it might evaluate your long-term king safety as lower. I don't know of any engine v engine games in this position, but I don't think you can say the engine evaluation is wrong. The best you can say is it's not very human.
    – JP Alioto
    Commented Feb 8 at 21:56


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.