I made an error on my first reply. I meant to indicate Stanford Wong VP
rather than Winpoker as "older" software. Wong's product was 100% accurate
even running on an "086" processor back in 1991.
[Original Message]
From: Steve Jacobs <jacobs@xmission.com>
To: <vpFREE@yahoogroups.com>
Date: 11/24/2005 10:21:25 AM
Subject: Re: [vpFREE] Videopoker:How reliable is the Math?> In Blackjack counting over the years what was important to your play
had
> to be re-arrainged.Despite decades of math testing by Professors and
others
> complicated systems were later "found" to yield no more return than
simple
> Hi -low.
> People (not me to scared to bet big) invested $$ in higher level
systems,
> learned to count on their toes, twisted their ring around to count aces,
> learn arcane memory word systems to keep indices in your head (this
idiot
> guilty).To find after wasting a good chunk of their time that side
counts
> got you nadda.etc.
Side counts for card counting are a lot like penalty cards in VP -- they
have the potential to improve your outcome, but they may not be worth
the additional effort.> I know this was back in the old days when a 100 million hands run
> seemed like a big number to us math impaired but Videopoker is a much
more
> complicated game what are the changes that if you rely on some of the
finer
> math pointsof the game in 10 years they going to say oophs we really
should
> have run a 100 trillion hands to be in tolerance and actually prove
> results?The commercial VP programs are not based on simulating a large number
of hands. Instead, they use combinatorial analysis to calculate the
results.
This means that the results are completely accurate, and all of the
different
programs end up with identical numbers for how often each payoff is hit,
when playing max-EV strategy.However, these numbers are "reliable" only if you understand that they
are designed for one very specific (narrow) purpose -- to maximize EV.
The VP "experts" teach that EV is the cat's meow, the one true golden
standard that should be sought above all else. In my (not so humble)
opinion, that is quite misguided.The idea that there is a single playing strategy which is always best
is a myth, but it is a myth that is likely to continue to be treated as
"fact"
for a long time to come. The brutal truth is that the best way to play
depends on what the player wants to accomplish. A recent thread on
vpFREE is a good example to illustrate my point. I've presented a
formula for computing RoRBR, which is "Risk of Ruin Before Royal",
and showed that a bankoll of 731 units gives the player a 50/50 shot
at hitting a royal, when playing 9/6 JoB. However, the thing that hasn't
been discussed recently is the fact that this number of 731 units applies
only to a player who is using the max-EV strategy, and it is possible to
do better by using a different strategy. If a player says "I don't care
about
ยทยทยท
On Thursday 24 November 2005 07:56 am, cbres77376@aol.com wrote:
EV, I care about stretching my bankroll to get the best shot at hitting
a royal flush," then the player should adopt a best-shot(royal) strategy.For 9/6 JoB, the best-shot(royal) strategy reduces the 50/50 bankroll
from 731 units to 720 units. To put that another way, if two players
start with 731 units each, the player who uses a max-EV strategy will
have a 50.02% chance of surviving to hit a royal, while the player
who uses a best-shot(royal) strategy will have a 50.54% chance of
hitting the royal. This is a lot like comparing two coin-toss games,
where one uses a coin that is almost exactly fair, while the other
uses a coin that is biased to come up heads 50.54% of the time
and tails 49.46% of the time. Which coin do you want to use?Bottom line: EV isn't everything.
vpFREE Links: http://members.cox.net/vpfree/Links.htm
Yahoo! Groups Links