Ok, when examining large complex systems, or large samples of results, this 80/20 ratio pops up often.

My question is, how actionable is this information? If your business (or poker strategy) is generating 80 percent if your revenue from 20 percent of your transactions, maybe this is an equilibrium of sorts. Maybe maximizing the 20 percent transactions will tilt you in ways that you can not foresee.

Also, another question is the nature of Pareto and how it is always hindsight. If this is just a common ratio, how useful is this observation?

If a player is examining a stored database of poker hands, and 80 percent of profit comes from 20 percent of hands, so what? Other than seeking out soft games and bad players, and playing as many hands as possible, I don’t see a way to “maximize” skill during those elusive 20 percent of hands.

If you have leaks, they may be tiny and sorting a database by net profit (or loss) will not find a tiny edge nor a tiny leak.

Is Pareto really a thing, or just statistical malarkey?

Thanks,

-Rob

Mathematically, it's just a Power Law with exponent 2. It tends to be a fairly common distribution pattern.

https://en.m.wikipedia.org/wiki/Power_law

These kinds of skewed distributions don't always have exponent 2 (i.e 80/20), it just tends to be common.

If you determine that a particular phenomenon does have this kind of distribution, for whatever reasons, then as long as those reasons exist it is predictable, not malarkey.

Thanks NewOldGuy. I will study up on Power Law.