> I read the section before asking the question .

So, apparently, it didn't help?

> I don't think it helps much because it assumes a very
> stable playing environment.

Why do you say that?

> What if the player plays
> several games in any one session so it is hard to get
> a constant win rate and standard deviation.

You do the best you can. You can figure the win rates separately and add them, and you can square the s.d.s, to get variances, add them, and then take the square root, to get back to the global s.d. None of this has much to do with how a team devises compensation.

> I do think it reinforces my thoughts that for smaller
> amounts of money won the amount paid to the player
> should be a smaller percentage.

Most pay schemes work on TIME played, not money won. You have no control over the latter; it is what it is. Players should be paid an agreed-upon percentage of the team win, based on time put in (pro-rated), provided a minimum amount of time is played. The formula is for determining that lesser percentage only when there is an agreement that the split can be made prematurely, with a penalty to the players for wanting to take an early cash-out.

> What if pay were given on a certain % of bank won. I
> would think a higher % would be given at 100% of bank
> won versus 10% or 25%. This method also enables there
> to be short, medium or long run agreements.

There are many ways to structure payouts. Rick Blaine goes over several of them, pointing out pros and cons, in his excellent "Blackjack Blueprint," which you should read if you haven't already.

> Related would be if a draw could be established over
> time, perhaps based on % of bank won toward doubling
> bank.

I'm against this, but that's just me. I don't think players should get paid if no money is won. The investors risk their money, if things don't go well. The players need to risk something too -- in this case, their time. I'm against using investor money to pay players, when there are no profits, or when there are losses.

Don