Here is another Frank Stewart example. I am showing only the N-S hands since this is a declarer problem.
N-S vulnerable, the bidding could have been 1NT-3NT; E/W pass throughout (but the actual N-S bid 1c and 1d before going into NT).
The opening lead was the Jack of spades. Declarer had five tricks in the majors, and needed four more from the minors.
K4 AJ7 KJ873 J42 A6 KQ8 T95 AQ763
My choice of play is to win in the dummy, lead the jack of clubs and let it ride, then finesse to the AQ of clubs. This wins the four needed tricks when clubs are split 3-2, and also when the clubs are 4-1 with the king in East.
Declarer chose to win in hand, and finesse for three diamond tricks, in addition to the ace of clubs. This play works best when West has the diamond queen (it doesn't matter who has the ace). It also has "additional equity" when East has the queen and plays it on the first trick. Then South will revert to the club finesse (which wins as the cards lie).
But in the problem, one of two Easts played the ace of diamonds on the first finesse, concealing his Qx. South lost time by taking the "marked" finesse and lost the "race" for tricks.
Against a good player who will routinely false card, should I train myself to ignore potentially false signals, or at least limit them to the values from the principle of "restricted choice"?
Should I consider using a "mixed" strategy against a good player, randomly "observing" him at some times, and sometimes getting "caught," while ignoring him at other times and not reacting?
Should Stewart have pointed out that the club finesse was the superior play? (I didn't get that message from the article but believe this to be the case.)