gnasher, on 2011-September-23, 06:33, said:
Nor do I, but he didn't merely say "play one". His instructions to dummy were "play one, your choice". In its entirety, that is very different from "play anything".
You assert that there is a large difference, but I struggle to see any practical difference.
Clearly in "play(ing) anything", dummy is going to play precisely one, not two or more, since that would be illegal, and I think we can agree that such is not implied by declarer's wording. So in playing anything, he will play one. So any difference does not lie in "play one", it can only lie in "your choice". But to comply with "play anything", dummy must choose the one to play. I accept he could gather up the cards, shuffle them, and play the top one. But that is merely a mechanical method of making a choice. I see no requirement in the instruction "your choice" that dummy must use any specific procedure, or think about what would be better. So choosing one is precisely how dummy would comply with the instruction "play anything".
I wonder perhaps if our differences here are because you are making the same false (in my view) hidden assumption as SFI. In other words, I am suspecting you think that the phrase "your choice" implies some specific manner of making a choice, eg, choose the card you think best. But this is not the only way of making a choice, a randomisation algorith is also a way of making a choice, and "your choice" does not specify the choice method.
Perhaps more interesting is if declarer did actually give dummy an instruction to randomise the selection of the card, eg
"Shuffle them and then play the one on top". Could the defence now insist on choosing?
Edit: but I see RMB's answer at #29 covers that equally.