The part of my post that was supposed to be a table showing the probability of various results of a 5000-hand FPDW play came out a run-on mess when posted. Let me try to post that table one more time:
Here is the range of outcomes for 5000 hands of $10 FPDW: (for $1, divide win/loss by 10, etc)
RESULT PROBABILITY
lose more than 40000, 0%
lose 30000-40000, 0.1%
lose 20000-30000, 5%
lose 10000-20000, 21%
lose up to 10000, 30%
win up to 10000, 21%
win 10000-20000, 10%
win 20000-30000, 6%
win 30000-40000, 4%
win 40000-50000, 2%
win more than 50000, 2%
Hopefully that will work better!
--Dunbar
---In vpFREE@yahoogroups.com, <h_dunbar@...> wrote :
As far as being wary of an "outlier" with a high Nzero...
Consider two scenarios:
1. All your EV comes from plays with the same Nzero as FPDW;
2. The rest of your EV comes from plays with a much lower Nzero than FPDW. (that is, FPDW is an "outlier".)
If all your play has approximately the same hourly EV, you're more likely to be able to justify playing the FPDW game if it IS a high-end outlier than if it isn't. The Nzero for the year will be lower if the rest of your EV comes from lower Nzero play.
In fact, even playing 5000 hands of FPDW at every casino in Las Vegas is unlikely to get you to its Nzero of 445,000 hands. You'd have to play at 90 casinos.
Here is the range of outcomes for 5000 hands of $10 FPDW: (for $1, divide win/loss by 10, etc)
RESULT PROBABILITY lose more than 40000 0% lose 30000-40000 0.1% lose 20000-30000 5% lose 10000-20000 21% lose up to 10000 30% win up to 10000 21% win 10000-20000 10% win 20000-30000 6% win 30000-40000 4% win 40000-50000 2% win more than 50000 2%
As NOTI pointed out, the chance of losing is 55%. But there's no question in my mind that if you had this 5000-hand opportunity on Dec 31, and you were $30K ahead for the year, and your bankroll is at least the Kelly bankroll of $147,000 (for $10 FPDW), then going for that $1900 of EV is the correct decision.
Similarly, if it's the beginning of the year, and you have solid expectation of being able to play enough low Nzero games during the year to put yourself well in the clear of a $30K loss (and you have the Kelly bankroll), then it's correct to go for the 5000-hand FPDW play.
Notice that none of this violates NOTI's absolutely correct admonishment to avoid the tax consequences of a losing year. That's valuable advice that's seldom brought up and too often ignored.
--Dunbar
[Non-text portions of this message have been removed]