Inferential Reasoning

30 or 40 years ago, Terence Reese wrote a really good column about a hand in which secondary assumptions needed to be made. Since he needed some cards to be in the right spot, he was able to figure out how to guess another suit; only one guess had any chance.

Playing in a rubber bridge game against mixed opposition, I encounter something similar. LHO deals and opens 1, partner passes, and RHO bids a forcing notrump. I hold AQJ1075 75 8 KQ72. That's an obvious 2 butt-in. The opponents do no more bidding, but partner raises me to three, and I, of course, bid game. The opening lead is the A (they play ace from ace-king) and I see:

	9863 Q863 KJ9 A4

	AQJ1075 75 8 KQ72

LHO	Partner	RHO	Me
1	Pass	1NT	2
Pass	3	Pass	4
All Pass

RHO contributes the

2 to this trick and LHO goes into the tank. They play standard signals, so hearts are obviously 6-1. After about three minutes, LHO emerges with a low diamond. I was pretty sure he was going to do that; I've been thinking about the diamond suit for awhile. My table feel tells me that LHO has the ace. He's hoping to take two diamonds and two hearts; he wouldn't strand his

A unless he had a sure entry. But playing the

K is the wrong play. In a sense, I can play the

J and claim! If it loses to the

Q and LHO has the

A, then RHO has to have the trump king. LHO would never underlead an ace if he had three cashing tops and the trump king. Nor would that leave much for RHO to respond. The former inference is much better, so much so that despite my overwhelming feeling that the

A is on my left, I play the

J anyway. RHO wins with the

Q and plays a low diamond. Why not... I ditch a heart and LHO wins. He goes into another tank. I claim on the marked spade finesse.

I don't think I've ever seen this sort of inference, where one can guess a suit based on the opponents' actions relating to another suit. Probably they have been there, but I've missed them. I know this one made my head hurt.