It seems to me that you need to pre-cool the units if you want it to be a
straightforward test of thermal retention.

Load them all with ice the night before the contest and then clean them out
and reload with ice for the measuring of the melt rate.

If you don't do that, a unit with better insulation, but a higher thermal mass
might appear to lose because of ice melted to cool the unit down.

In article 26pRf.503838$0l5.167922@dukeread06, "Glenn Ashmore"
wrote:
That is what started the whole debate. When I mentioned that test I think I
got their testosterone in an uproar. :-) The $450 Frigid Rigid was a close
second and the Icy-Tek came in a distant 6th. I got a hundred bucks riding
on the outcome and win as long as my cooler is even close to equal to
theirs.