[Ed Miller]

He's asking about the median "best" hand. For instance, say ten random hands are dealt, and out of those hands, the "best" one is KJs. Then ten more random hands are dealt, and the "best" one is KK. It you did this like 100,000 times, what would the median "best" hand be?

That's what he's asking about. Unfortunately, I am highly skeptical that this process could possibly identify any new and useful information. One major issue is the value of hands varies significantly with how many players enter the pot. There is no relaible or meaningful way to identify the "best" hand out of 10 random hands.

Even if you could reliably identify the best hand, how would knowing the median best hand help you?