Poker Poker Poker!: [vpFREE] Re: Long term vs short term again

nightoftheiguana2000 wrote:

> --- In vpFREE@yahoogroups.com, "Harry Porter" <harry.porter@> wrote:
> > And consider the FPDW player w/ .2 cb: NO = a mere 252 thou hands.
> >
> > I find these numbers place the "long term" into a realm that
> > serious players should consider far from an abstract concept.
>
> Also, if you're more the visual type, run this Java calculator:
> http://www.lotspiech.com/poker/GamblersRuin.html
> Try stake $3000, retire with $8000 and game is Deuces Wild and hit
> run. Grab some java and watch the future predicted with the magic of
> mathematics. Then if you want to, go try it in the real world. The
> question is: Do you feel lucky, punk? (Remember? Dirty Harry?)

Excellent (actually, downright beautiful) suggestion.

Remember that without cashback, FPDW NO = 447,275. (Again, NO is the
point at which 84% of results can be expected to be positive.)

Allow some patience and let the calculator run to that point (for an
hour or two in the background, while you carry on with your "real"
work) to test how this percentage bears out. (You may wish to use the
stake/retire values of 4500/12000 if you decide to continue with the
second exercise proposed below.) I haven't performed this run yet --
I'll likely give it a go later today to satisfy my own curiosity.

A second exercise: Stop the calculator at that point and calculate
your expected win should you end with a positive result, and expected
loss should you end in the red. You do this by multiplying the
average value of each band by its probability and summing the result.

With an 84% chance of ending at the greater-magnitude expected win,
you'll likely consider your gaming prospects quite favorable.

(Note, if you multiply each value by its respective probability -
84%/16% - and sum the result, you should end with a value in the
ballpark of the EV of this play: (1.00762-1)*(477275*1.25)= approx
$4500. The $4500/$12000 stake/retire I suggest above will yield a
result that is closer to this value than the rougher $3000/$8000.)

------

These exercises are as strong a validation of the principals of
advantage play as I can imagine. Those who reject advantage play tend
to be blind to such results.

(This is not to say that there aren't reasonable alternative
strategies for play other than max-EV that fully recognize the
underlying math of max-EV play. The VPFree write up on Video Poker
Strategies (authored by Steve Jacobs) describes some:
http://members.cox.net/vpfree/FAQ_S.htm
Note: For several of these, I consider the consequences vs. those of
Max-EV expected results to be of such limited magnitude, given the
preferences of most players, that there is limited, if any, benefit in
them.)

Of course, give the calculator a run under a thinner edge (say, 10/7
DB with no cb, and the inherent pitfalls of near B/E play become
apparent. (Similar risks race those who might play 9/6 Jacks w/ .6%
cb/bonus incentive, "secure" that they have an edge). Concerns become
greater when error cost (due to skill deficiencies,
inattention/distraction, etc.) is factored. There's often an
impression in AP writings (casual and in print) that any edge makes
for a strong play.

------

As a side note, I give credence that there are those for whom max-EV
strategies aren't appropriate -- notably, those who don't have
patience for the math to play out before having a strong probability
of achieving a given win goal. Alternate strategies, such as that
roughly outlined by Rob Singer, can be appropriate in such a case --
provided there's a strong cognizance of the underlying loss risk
(something GREATLY downplayed in his discussions).

I come under criticism (in a rather bizarre manner) from those who
find such a suggestion objectionable. (I won't go as far as to
suggest that bizarreness takes on a "goose stepping" guise, of course
;) -- any discussion of this latter remark is best suited for the
FREEvpFREE forum.)

- Harry