On Sun, Jul 28, 2013 at 12:24 AM, Veit Elser <ve10@cornell.edu> wrote:
My solution of the NYT puzzle was exactly the same as Michael's: Reduce (7,7) to (6,6) … until you get to (2,2), then you have to do a 2 vs. 2 weighing and, if they balance, a 1 vs. 1.
Actually my (2,2) solution was "Weigh 2 vs 2 until there's an imbalance", but yours is more efficient, as mine might take three weighings.
I'm still puzzled why the original puzzle asked for (7,7). Maybe Victor was as well, and asked the obvious question: For given L, how small can H be for (H,L) to be feasible? For example, (3,3) is feasible but (2,3) is not.
More natural questions: What's the first feasible (H,L) with L > H? How about with L >= H + k for k=1, 2, 3, ...? Must some (H,L) work for any k? --Michael -- Forewarned is worth an octopus in the bush.