A number of players do not appreciate that a contract which depends on two finesses out of three succeeding has a 50 per cent chance of success. (Consider: all three right, 1/8; exactly two right, 3/8; exactly one right, 3/8; none right, 1/8. Hence, at least two right, 1/8 + 3/8 = 50 per cent.)

This was a typical example from a recent pairs event. The usual contract was 64. At my table, North opened 1NT, South launched into a Gerber enquiry with 42 and ascertained that North held three aces but no kings. With little to go on after this uninformative auction, as West I led a trump and declarer faced his problem with possible finesses in all three side suits. He also has the possibility of taking a ruffing finesse in clubs rather than a straightforward one.

I was lucky, for the declarer, after drawing trumps and ending in dummy, first tried his luck with the hearts. No joy, and after winning with the king, I switched to a low diamond. After a few minutes' deep thought, declarer finessed again and so went one off.

Now, which finesse should he have taken first, and why? I am sure that his best bet would have been to try the straight finesse in clubs first. The extra edge this gives is that if the finesse wins, the 2A and a club ruff may bring down the king. This works beautifully as the cards lie. And if the club finesse fails and a heart comes back? Ah, now guesswork comes in.

Game all; dealer North

North

4A 7

!9 5 3

#A Q 6 2

2A Q J 5

West East

4J 10 44 2

!K J 7 4 !10 8

#J 9 8 5 #K 10 7 4

2K 8 7 210 9 6 3 2

South

4K Q 9 8 6 5 3

!A Q 6 2

#3

24