The breaktime chat among Chess and Backgammon players often centres around the new super computers that could ‘solve’ these games as they play 100 per cent accurately. We humans are light years behind in technical and calculation skills.
This will never be a concern for Bridge players; Bridge is a game of inevitable mistakes, and no computer can possibly calculate what those mistakes will be – what algorithm would be able to tell if my LHO is tired, dismayed or just not top-class?
Here are Danish brothers Knut and Lars Blakset proving errors can’t be predicted.
The 4-4 Spade fit was not found after North elected to use a Bergen raise*.
West led a small trump. Declarer won cheaply and ducked a Diamond – the defence winning and playing another trump. South took his Diamond ruff, but then faced the issue of how to get back to hand to draw trumps. Knut avoided the trap of leading a small spade to the Queen and instead cashed the ♠Ace, which brought the welcome news from West. Now what?
He could either try to run the ♠10 from dummy and hope West hasn’t got the last trump, or exit with the ♣K and hope West has the Ace, as he can’t give himself a Spade ruff.
An AI program would no doubt figure out that the best theoretical chance is to exit in Clubs. Knut is no supercomputer though – he’s just a very good Bridge player – and he saw another chance.
He ran the ♠10. West scooped up the trick with his last trump but then sleepily exited in Diamonds.
Knut was home again! The last two trumps strip squeeze East down to singleton ♣A and South could throw him in to get his two Spade tricks.
Comments