There is a debate about when, if ever, artificial intelligence will play bridge as well as humans. Computers prevail in chess and Go, but bridge involves factors such as inference and deception.

In a team match, both Souths played at 3NT, and West led a heart, riding to the queen. At one table, South forced out the ace of diamonds, won East's heart return and had only eight tricks.

In the replay, South’s approach was different. After winning the first trick, he banged down the king of clubs. As it happened, West had to play low; if he won, South could run the clubs, winning four clubs, three spades and two hearts. South then switched to diamonds for nine tricks.

South gave himself the best technical chance, but even if the lie of the clubs hadn’t been friendly, the defender with the ace might have ducked the king.

I think it may be years before a computer finds a play like the second South's, with psychological implications. But I fear I may be wrong.

Daily question: You hold: ? 10 6 ? K J 9 7 4 ? 9 7 ? A 10 7 3. The dealer, at your left, opens one club. Your partner doubles, and the next player bids two clubs. What do you say?

Answer: Chances for game are bright. Partner has heart support, and his hand is short in clubs. Bid three hearts. If he holds a sound minimum double such as A Q 5 3, A 8 6 2, K 1 0 6 5 ,2, you may take 11 tricks. You would bid two hearts to compete with a hand such as Q 6, K J 9 7 4, 9 7, 1 0 9 7 3.

South dealer

N-S vulnerable

Tribune Content Agency