There has to be a better way
It does not pay to double the opponents for penalties unless you are reasonably sure of defeating the contract. That is why you hear it said so often that you normally should not double unless you expect at least a two-trick set.
Let’s say you double nonvulnerable opponents at four spades, expecting a one-trick set. If you are right, you gain 50 points, scoring 100 instead of 50. But if you are wrong and they make four spades, the opponents pick up an extra 170 points. You are thus laying odds of better than 3-1 that they will go down.
The mathematics are even more prohibitive if you double a partscore of two hearts or higher — if you are wrong, the opponents score a game worth hundreds of points.
In duplicate bridge, where other factors enter the calculations, a close double is made more readily. Today’s hand occurred in a duplicate, and West doubled his vulnerable opponents in two spades, hoping for a one-trick (200-point) set.
West led a diamond, and East won dummy’s king with the ace. East correctly returned a trump, and West took the jack. West then led another diamond, taken by East with the ten.
East could now do nothing to defeat the contract. If he led a diamond, his side would wind up with three spades and two diamonds. And if he led a spade instead, the same result would accrue. Declarer would eventually be able to discard his third diamond on dummy’s ace of hearts. So South made two spades doubled for a score of 670 points.
However, the contract should have been defeated. After taking the jack of spades at trick two, West should have cashed the ace and then returned the deuce of diamonds.
East would win the J-10 of diamonds, bringing his side to five tricks, and then lead the queen. Regardless of what declarer did, West would score his queen of spades to secure the critical 200 number.