A question about the ettiquette of this forum: How do I share credits with Brendan McKay (for his formal answer) and Henry Cohn and Will Jagy for their pointers to other resources in the comments? Thank you to everyone who contributed. David.

To: Henry Cohn. I admit I'm geeking out here a bit on this problem. Yes, there are some common sense and "best guess" solutions. I've tried a few and then easily found other solutions that are several $100's better. I'm asking this forum to suggest scientific methods that will substantiate my proposed solution and reduce the squabbling of the siblings. I wish to say, at the end, here is a very good solution, it's the best method and its technically valid, not just my say-so. The family problems are awful. I'm trying for lemonade (geekwise) out of these lemons.

To: Joseph O'Rourke: it is the approximation algorithms that I seek. Where is one described? I've been on mathisfun and codingthewheel and don't find it there.