Okay, put the kids to bed and don your protective eye gear before proceeding. This one’s really ugly.
Black to move. My thinking at the time:
I went into this position with the notion that black needed to restrain c4. If I can do that, white’s pawns stink, the e-pawn eats up space and soon the c-file will be the road to happiness and mirth. OTOH if white gets to push c4, my center will get wobbly and — I don’t know exactly what. Did I think his pieces were going to get really active? My rooks aren’t connected b/c of my undeveloped bishop. Should I kick the knight with …h6 and then play Bxh3 and f5? Looks like my king is getting drafty and my light squares are weakened. Or …h6 with some idea of playing Be6, f5 and Bf7 to get away from his potential Nh3-Nf4? I seem to lack the necessary tempi to do that.
So I got all contorted and played something that achieved none of my goals:
15…Qd8 16. Rab1 b6 17. c4
Well now I’ve botched it, and black is worse, for all the reasons listed above. So then a bunch of nonsense ensued, as I tried to come up with some concrete solution to white’s light-squared play, and lucked into a drawing cheapo after mutual blunders.
17…dxc4 18. Rfd1 Na5 19. d5 Re8 20. Nxe4 Kg8? 21. Qd4? Rxe4 22. Qxe4 Bf5 getting the exchange back on c2. The next day my opponent mentioned that we both overlooked 21.Nf6+, simply winning the exchange.
In position A, Stockfish says it’s around -.75, way better for black. Its recommendation is the very simple 15…h6 16.Nh3 Bxh3 17.Qxh3 and now Rc8 and f5 are both good choices.
- Tactically, c4 is prevented for the moment because d4 would hang.
- Positionally, giving up the (undeveloped) Bc8 for white’s (only active) piece doesn’t cause any concern.
But wait! There’s more!
In Position B, Stockfish says black is still marginally better (-.3 or so)! So allowing c4, while not optimal, doesn’t cause black’s position to magically collapse. The recommended series from position B is 17…f6 18.cxd5 Qxd5 19.Nxe4 Nxd4 20.Rfd1 Rd8 21.c4 Qd5. I expect the small advantage here stems from the better minor piece, as black will trade off the Be2, and/or marginally better pawn structure.
My actual choice of 17…dxc4 is panicky and terrible and gives the advantage back to white. Instead of centralizing my queen, I decentralized my knight. Which reminds me of an observation someone made about Fabiano Caruana: His pieces always seem to funnel through the middle of the board. If I take nothing else from this game, that would be a good correction to make. Given a choice, centralize! Duh!
But it feels like there’s more to be learned here. We may be seeing a confirmation of the idea that I over-value my bishops?