As the 75th New England Open entered its final round, the top two boards in the championship section were populated by four strong Massachusetts masters, all stalwarts of local clubs. While young FM Mika Brattain would best veteran FM Chris Chase in a grind, FM Steven Winer would win his game against NM Yi Yang in an unexpected flurry to take his own share of the regional championship. 

1. d4 Nf6 2. Nf

Normally an unremarkable move, but in this case I knew before the round that I was going to play Yi Yang and some research suggested that he would play the Benko Gambit against 2. c4. I decided that although Nf3 is not objectively any better, it could lead to a position that my opponent was less familiar with.

2… c5 3. d5 b5 4. Bg5 d6?!  

Surprisingly enough this natural-looking move is already a mistake.

5. e4!  

Opening up the position makes sense with Black spending time on the queenside. I found the idea because in several variations in this line White will play e4 as an outright pawn sacrifice - I figured it would be good when it doesn’t even come at material cost.

5… Nxe4 6. Bxb5+ Bd7 7. Bd3! 

Although a retreating move it is strong because it both attacks e4 and leaves the bishop on d7 where it blocks other Black pieces.

7… f5?  

This might be good if the knight could remain on e4 but since White can force it away it only weakens Black's position – Nxg5 8. Nxg5 Qb6 was relatively better.

8. Nbd2 Nf6 9. 0-0 

I thought for quite a while on this move. On the one hand I found it hard to believe that Black could get away with Nxd5 with so little development, but I still was not totally sure as sometimes a large lead in development is not as useful as one might expect. I eventually calculated 9. 0-0 Nxd5 10. Nc4 threatening both Nxd6+ and Bxf5 Bxf5 Qxd5. Ultimately though I also was guided by the idea that it is often not possible to calculate all lines when sacrificing material, and a player must rely on experience. While chess engines can be programmed to calculate all possible options up to a certain depth, humans cannot play that way, and should not attempt to. Decisions, even risky ones, must be made with imperfect information.

9… Nxd5? 

The flip side of the attacking comment is that sometimes a sacrificial offering should be avoided based on a sense of danger rather than purely relying on calculated variations. I was fairly surprised Black decided to take the pawn. The pressure of a key last round game can influence decision making. It also influenced me later when I just wanted to win safely rather than worrying about finding the quickest possible victory.

10. Nc4 Nf6 11. Bxf6! 

At this point I figured the knight had moved so many times it made sense to trade it to emphasize Black's lack of development. It also turns out that both recaptures have tactical drawbacks.

11… gxf6

exf6 would not have been any better because of 12 Re1+ Kf7 13. Nxd6+! Bxd6 14. Bc4+ Kg6 15. Qxd6 with winning threats. I did not bother looking any deeper than that since it was clear that line was highly favorable. If I had actually reached the position then I would have looked at it more thoroughly. One needs to calculate far enough to be confident of an evaluation, but it excessive deliberation when a verdict is clear can be costly when time and energy are taken into account.

12. Nfe5 h5 

The only way to stop Qh5# 

13. Nxd7?! 

13. Be2 would have been better but the move did not occur to me. Seeing attacking retreats is often difficult unless there is a specific reason to be looking at the piece already. In this case, focus on other favorable lines left me without a great need to consider my bishop. This also illustrates the importance of identifying candidate moves since many errors are based on missing a move entirely rather than seeing it but miscalculating it.

13… Qxd7?! 

Nxd7 would at least develop some pieces although White is still winning after Bxf5. The impulse to try to keep the pawn on f5 is understandable, but in this case the pawn is lost anyway and Black's pieces remain out of play.

14. Qf3 Qc6 

14… Nc6 15. Bxf5 e6 16. Bxe6! Qxe6 17. Qxc6+ is even worse

15. Qxf5

White should have a very strong attack so long as a queen trade is avoided.

15… d5 16. Rad1!  

Rfe1 is similar. The key is to realize that dxc4 is met by 17. Be4 winning at least the rook on a8 – given that, there’s no need to move the knight from the c4 square.

16… Bg7 17. Rfe1 

Qg6+ is not necessary since chasing the king off of f8 actually may make it safer than it is on e8. Before giving a check or other forcing move it is important to figure out if forcing the piece to move is actually useful. Often it is, but chasing away a badly placed piece can actually help your opponent.

17… Qd7 18. Qf3!

Unlike move 13 here I saw the strong retreating move, though already looking at the queen – it needed to do something – made things considerably simpler.

18… Nc6 

Black has no good alternative since e6 allows Bf5 winning everything based on the numerous pins it creates.

19. Bf5 Qc7 20 Qxd5 Kf8 21. Bd7

There are faster mates according to the engines. However, finding a sure win is good enough. Bd7 wins substantial material without allowing counterplay, so I saw no need to look further.

21… Rc8 22. Bxc8 Qxc8 23. Qxc5 Rh6 24. Na5 f5 25. Nxc6 Qxc6 26. Rd8+ 

And now mate is forced, though by this point it was superfluous.

1-0