Breakthrough and Opposition

While probing the depths of the ChessBase database indexes, I came across this tragicomedy from the 2003 8^th All-Africa Games. Mark Dvoretsky employs the term tragicomedy for violations of basic endgame principles by strong players—missing technical wins or draws. In this case, Angola team member Ediberto Domingos missed a win against Kamel Sebih of Algeria.

From its beginning as a Dutch Defense (ECO A85), the game was a positional struggle that came down to a materially equal endgame. Then, Domingos, as White, gave up a pawn for a more active king. A few moves later Sebih gave the pawn back and brought his king into the struggle.

With 46…b4, Black gave up a piece to create a powerful passed pawn.

The game continued 47.Bb6+ Ka4 48.Bxc7 b3 49.Be5 Ka3. We reach the position that caught my interest this morning.

It appears to me that the white bishop must immolate itself for the passed pawn, but that White then has a won king and pawn endgame. As is my practice, I opened the position in Hiarcs and tested my ideas against the Silicon Beast. I was able to win several ways.

50.h4! +-

50.Kd5! +-

50.g4!! +-

I favor the dramatic 50.g4 because it offers the most rapid demonstration of the superior placement of White’s king. The game against Hiarcs continued 50…Ka2 51.gxf5 gxf5 52.Kd5 Kb1 53.Bd4 Kc2 54.Ke5 and the result should be clear (see diagram to right).

I did not like Ediberto Domingos’ fiftieth move: 50.Kd6, although as it turns out, that move also wins. The game continued 50…b2 51.Bxb2 Kxb2 reaching the next diagram position (below).

Here Domingos blundered (the tragicomedy) with 52.Ke7?? This move suggests that White’s idea behind 50.Kd6 was to head towards the h-pawn. However, e5 is just as close to the h-pawn as e7. Moreover, this strategy loses tactically—it becomes a pawn race that favors Black. White should aim for a breakthrough in the center.

White still could have won with 52.Ke5 or 52.g4, and perhaps even with 52.Kd5.