Lahdenmäki,L - MNb
ws H/673 ICCF, 2018
0–1
With 16.c3 my opponent chose the critical continuation. When I meet the Kalashnikov as White I'll probably play it myself. TonyR's recommendation 16...a5 did not satisfy me, because way too often Black ended up in a simplified position with a queenside majority - pawn a2 and b2 vs. pawn b4, with pawn d6 remaining backward for a very long time.
So I decided it was better to saddle White with an isolated c-pawn by 16...bxc3. That was not the end of the problems. The first one was 20.Rfb1, when I had to demonstrate it's nothing but a transposition to Cipressa-Rotella, corr. 2010. The second one was the clever novelty 27.Qa3. After several days of examining it I think I can add something to TonyR's explanation on page 58.
27.Qa3 is probably an improvement. As long as the third row remains cleared Her Majesty always can quickly turn to the kingside, so the plan f2-f4-f5 and g3-g4 is just as strong. Moreover Black feels the pressure on the a-pawn, but less clear but at least as annoying also indirect pressure on pawn d6. The line 27...Nd4 28.Kh1 f6 29.f4 a5 30.f5 Ba6 31.g4 Rd7 32.Qa4 Bc8 33.Rxd6 makes this clear. In some other lines the sac Nxf6+ immediately decides the game, as mentioned by TonyR.
So I think the plan ...a5 and ...Ba6 is not the most urgent. To avoid disaster Black must force the trade of his bishop against White's dominant knight on d5; that's first priority. As I could not find a way to do this after 27...Nd4 I decided that this obvious candidate was inferior. Of course 27...Ra5 28.Qe3 seems to win a tempo, but that's unclear after Nd4 29.f4 f6 30.R1b2 Be6.
Apparently my opponent just assumed I had played 27...Nd4 indeed. Well, a win is a win, no matter how cheap, but it's in a way a pity that I must remain inconclusive on the question if 16...bxc3 is really better than 16...a5.