I agree! It’s fairly clear that this was the puzzle intended; and some of the alternate protocols proposed are far-fetched.
I think the people showing alternate protocols should avoid these complaints by incanting something like “Strictly speaking, the puzzle should be written less ambiguously (as jtur88 has done above) but …”
What doubles the confusion is that some Dopers are computing the wrong probability for the model they select. Thus there are two intertwined subthreads: a vehement argument over where the problem statement is crystal clear or ambiguous (which belongs in IMHO or Mundane-Pointless I think, not GQ), and separate disagreements involving probability arithmetic. (When someone writes “1/6” it isn’t clear whether they’re using an alternate model, miscalculating for the normal model, or just making a typo.)
~ ~ ~ ~ ~ ~ ~
As I said upthread, this puzzle is in the Monty Hall category. Several weeks ago I posted another such
To avoid hijacking this thread I’ll Spoiler the problem statement.
[SPOILER]The puzzle arises in the game of contract bridge, but non-players should be able to understand the setup. You and your partner (the dummy) have nine hearts missing only the Queen, Jack, Trey and Deuce. You cash the Ace; LHO plays the Trey, RHO the Jack. You lead toward dummy’s King, LHO following with the Deuce. Should you play for the Queen to be with LHO or RHO? (Your opponents are indifferent in their plays of Trey vs Deuce, or Queen vs. Jack. However, as the play has proceeded they would not play a picture card when a small card still available.)
There are only two possibilities that remain for the original holdings:
LHO: 3 2 RHO: Q J
LHO: Q 3 2 RHO: J
The cards were shuffled randomly to begin with and, with nothing else to go on, the a priori chances for these two cases are about the same. (1st case is slightly more likely than 2nd.)
What is the percentage play now?
[/SPOILER]