Generous Defense

Q 6 5 2
A 10 6 3
Q
J 9 8 6
J 10 7 8
7 5 4 K Q 9
J 9 5 10 8 4 2
A K 10 2 4 2
A K 9 4 3
J 8 2
A K 7 6 3 Contract: 6 spades
------ Opening lead: A of clubs

Can you make 6 spades on any defence? It looks to me that you cannot. On a 2-2 spade split, yes, it would be a cakewalk. But it looks to me as though you cannot in the face of that 3-1 split. Yet I have the play of three defending pairs all allowing the contract to be made, all starting with the same A of clubs opening lead, on three different defensive errors. Let me start with the simplest and perhaps easiest to see:
Declarer ruffs the opening lead (as they all did, of course), cashed the Q of diamonds, came to the A of spades and back to the queen, then ruffed a club, followed by ruffing a diamond. which was itself followed by a third trump lead to the A. Two clubs now went on the A K of diamonds, leading to this end-position:
------
A 10 6 3
------
------
------ ------
7 5 4 K Q 9
------ ------
K Q
------
J 8 2
7
------

Declarer cashed the fifth diamond, sluffing the 3 of hearts as West sluffed the K of clubs to save 3 small hearts, for whatever reason, and East the Q of clubs to save some bigger hearts! Declarer now led the J of hearts into the Q, and East was forced to lead from his K of hearts. Bid and made. East should certainly have seen that throw-in coming and saved the Q of clubs. Declarer has no more trump. The Q of clubs wins. This isn't the first time I've seen a defender figure that if declarer is out of a suit in a trump contract, he has no losers in that suit. Well, he can very well have losers in a trump contract when he has no more trump, no?
Here's a different way that the defense gave up that key trick. It's that third round of trump that does the trick. The first declarer cashed the third round and so should have lost a club trick on losing the lead in hearts. This declarer didn't cash the third round and so should have been set by that third trump.
Ruff the opening club lead, cash Q of diamonds, ruff a club, two top rounds of spades, two top rounds of diamonds, sluffing a club and a heart from dummy, and ruff a diamond. Now ruff a club in the closed hand, reaching this end position:

Q
A 10 6
------
------
J ------
7 5 K Q 9
------ ------
K Q
------
J 8 2
6
------

Declarer led the established diamond, both West and dummy sluffing a heart, then went to the A of hearts, cashed the Q of spades, and conceded the last trick to East. West must ruff out that diamond winner, for heaven's sake. It doesn't matter whether declarer overruffs or not. The 6 of diamonds isn't a winner, and declarer must lose two tricks if that 6 of diamonds is ruffed, whether he overruffs or not.
In the final case, ruff the opening lead, Q of diamonds, ruff a club, cash two top diamonds, sluffing two hearts, then the 6 of diamonds . . . .
Q 6 5 2
A 10
------
J 9
J 10 7 8
7 5 4 K Q 9
------ 10
K 10 Q 4 3
A K 9
J 8 2
7 6
------

Declarer led the six of diamonds, ruffed by West, whereupon declarer pitched his 10 of hearts and the rest is history. One defender couldn't bring himself to ruff that 5th diamond, a winner if not ruffed, while his counterpart ruffed the fourth round, which is not a winner. Each of these defenders nullified the third round of trump for diametrically opposite reasons. One didn't use it to ruff a winner, thereby allowed the sluff of a heart declarer should have lost, while the other in effect ruffed out his partner's 10 of diamonds, allowing declarer, who hadn't drawn any trump yet, the luxury of a 2-1 split.
After winning the trick, West continued spades, perhaps sensing that that was declarer's Achilles heel -- too little and too late. At that point, declarer has no heart losers in dummy. He has 2 clubs to take care of, which he can easily effect by ruffing one (declarer has A K 9), throwing the other on the established diamond. And so it goes.

Yes, I became aware that I've only discussed defensive errors here, while this is slotted under "Declarer Play". I have often considered a third category (after Declarer Play and Defense) bridging the two for there are categories (like this one) that might apply to either, but haven't yet gotten around to it. For the present I'll let this stand. It's not going to do any great harm.