When you run a poker site, you talk a lot about liquidity and about ecology. The latter encompasses a few ideas, but they’ve all proven hard to quantify. Perhaps the most important of them is how tough the games are.
Ideally, you want new and weak players to last as long as possible, while retaining a level playing field. The harder the games they find themselves in, the less likely they are to find the game fun and to redeposit.
It would be great to measure if changes you made to the site made games tougher or weaker. One way to measure it is to look at VPIP, percentage of players seeing the flop, or similar. I’ve looked at this for other sites, but I suspect the trends are in the order of years, and that they may not be occurring anymore, as players converge on the best strategies.
Average win rate doesn’t work because it should just equal the amount raked in the long term (given the site is not growing). Top n% win rate may work, but is very volatile.
I think the best ways would be to measure the distance from a fixed kind of play to the actual play in the games. The problem now becomes to find a fixed kind of play, and I think there are two: total beginners, and GTO for whatever positions it can be calculated for.
This is a project I can only guess will work – it would require hand histories to calculate (without hole cards), and I don’t have access to an unbiased sample (not just hands from one player). It would definitely be interesting to try out though.