On Mar 8, 2013, at 3:12 PM, Joe Niederberger <email@example.com> wrote:
> All that being said, I'm still impressed with the > so-called "communications channel" approach referenced earlier. I've seen the same essential idea explained differently. I've even used it myself at a much older age (grad school). No matter. I get the distinct feeling that kind of approach could crush problems that my 10-year-old poking around "strategy" sketched above could never get near. Though its beyond me at the moment to explain why.
The information theory section appears to me to be a strategy, and that is how I went from groups of 6 or 3 to groups of 4, it was fruitful considering the ternary nature of the scale. This was also my guidance towards having a group of 3 (at most) as the final step. I wasn't able to maximize the middle step. I would say (if you averaged the weighings in my cases) my result was 3.2 weighings.
I don't see the communication theory approach a strategy in that it offers no particular insight or guidance towards a solution. It seems to translate the problem into an equivalent problem that seems just as hard. I mean, when you consider the code words and the task at hand (equal number of letters in each column, no inverses, etc.) that problem seems as hard as the original, like solving a Rubik's cube. Maybe even worse since the list of code words contains both sets (inverses).
However, a computer would be more applicable to the communication theory approach because it probably has an algorithmic solution.