At the time, at least, there was no way to index it for all 8 players involved in a hand. Each action taken would be indexed to the player that took it, and I'd need to sweep up adjacent actions for other players in each hand, but
only the players who were consistently in lots of hands with
that player. I've heard of bloom filters (now, not in 2012)... makes some sense. But the idea was to find some vector that made any
set of players unique when running through a linear table, regardless of the order they presented in.
To that extent, I submit my solution as possibly being the best one.
I'm still a bit perplexed by why you say 2^x & 2^y is tautologically sound as a unique way to map f(x,y)==f(y,x), where x and y are nonequal integers. Throwing in the bitwise & makes it seem less safe to me. Why is that provably never replicable between any two pairs of integers?
To that extent, I submit my solution as possibly being the best one.
I'm still a bit perplexed by why you say 2^x & 2^y is tautologically sound as a unique way to map f(x,y)==f(y,x), where x and y are nonequal integers. Throwing in the bitwise & makes it seem less safe to me. Why is that provably never replicable between any two pairs of integers?