--- In vpFREE@yahoogroups.com, "Chandler" <chandler_re@...> wrote:
>
>
>
> --- In vpFREE@yahoogroups.com, "vp_wiz" <harry.porter@> wrote:
>
> >
> > In playing the Riviera $500 loss rebate promo for new members, Bev & I opted for $1 BDlx, balancing goals of hopefully coming away a little to the good, plus targeting an extended session that yielded residual value on back end offers. (Again, max EV strategy would be more aggressive high denom play.)
> >
> > In the event of a winning session, the loss rebate loses value as your meter climbs. I made a gut assessment that an appropriate cash out strategy was at a $600 win.
> >
> > We were each fortunate to achieve this win target ... Bev cashing out in 15 minutes, hitting two quads in short order (and ending up with weak mailers), I ran play through for close to 90 minutes before my cashout, scoring a 5 night offer with $100/night in FP.
> >
> > - H.
>
> That is interesting. I had the same $600 figure in mind while entering my play. I basically pulled it out of my butt... er, gut. Based on some experience with $DDB and what it can do to a $500 session stake. I was hoping that the better trained and more nimble minds (certainly better than my poor intrument) of vpFREE could apply some methodology. It's instinctive for me, but it is well worth positing explicitly that the rebate loses value as your win increases, and absent other advantages, the reason to continue play decreases. Taken to ridiculous extremes, a rebate offer after a royal hit is virtually worthless and after a $5 net win still has most of its' value. TY, H.
>
> Chandler
>
For a one time shot promotion such as this, with moderate value, my inclination is to go with a "visceral" estimate of the math rather than fine tune things. But I imagine that were I to look for an optimal cash out value I'd start with the fact that the EV of any particular cash out strategy = probability of a successful cashout x the average expected cashout.
From there, I'd resort to "monte carlo" simulation, and use a vp tutor such as winpoker to run through computer runs (at a practical increment of something like 50 hands at a time) to arrive at fair estimates of those values (prob / avg cashout) for a range of reasonable cashout targets.
A little cumbersome of an exercise, but hardly an unreasonable one if the stakes are sufficiently sizable.
