Sunday, 6 September 2009

First as a study, then in real life

Earlier this year I used a problem to demonstrate that computers were 'inaccurate'. It turns out my conclusions were incorrect, based on the problem being unsound. (Click here and here for the details).
I've now discovered that Peter Svidler actually reached a position with similar characteristics, and was also able to construct a win where others may have given up. In the book "Devious Chess" by Amatzia Avni he gives a position from Dvoiris - Svidler, 1997, which is shown here. The king is trapped by the bishop and pawn (as in the linked articles) so Svidler needs to rely on Zugzwang to lift the blockade. He actually does this with relative ease after
50.c3 Qe2 51.Bc4 Qd1+ 52.Kb2 Qd2+ 53.Kb3 Qc1 54.Bf7 Qb1+ 55.Kc4 Qc2 56.Kd4 Qe2! 57.c4 Qb2+ 58.Kc5 Qa3! 59.Kb6 Qxb4+ 60.Kxa6 Qc5 61.Kb7 Qd6 62.Kc8 Qe7 0-1

However Avni may not have been aware of the original study as he includes the game in a chapter headed 'Virgin Soil' and about the above positions says 'A position not to be found in textbooks'

No comments: