vpFREE2 Forums

Bob Dancer Column - 20 DEC 2016

#1

Bob Dancer Column - 20 DEC 2016

Learning the Wrong Lesson

http://www.gamblingwithanedge.com/learning-wrong-lesson

or

<a href="http://www.gamblingwithanedge.com/learning-wrong-lesson">
http://www.gamblingwithanedge.com/learning-wrong-lesson</a>

···

*************************************************
This link is posted for informational purposes
and doesn't constitute an endorsement or approval
of the linked article's content by vpFREE. Any
discussion of the article must be done in
accordance with vpFREE's rules and policies.
*************************************************

#2

Bob wrote: “Whether or not you have made a good decision or a bad decision should be
determined at the time you make the decision — NOT sometime down the road.”

Grading (good or bad) a decision before the event is classical or a priori probability. Also grading sometime down the road or after the event is Bayesian probability. I would recommend the Bayesian approach, but hey, to each their own. Bayesian is superior when there are “unknown unknowns”, and, lets keep it real here, there are always “unknown unknowns”.

Bayesian probability - Wikipedia

Bayesian probability - Wikipedia
Bayesian probability is an interpretation of the concept of probability, in which, instead of frequency or propensity of some phenomenon, assigned probabilities represent states of knowledge[1] or belief.[2]

View on en.wikipedia.org

Preview by Yahoo

#3

The first variable you need to be concerned about is EV. In video poker, real world casino EV is not a 100% known variable. Even the legendary Bob Dancer, arguably the greatest video poker player in the free world, has admitted to making play mistakes in a casino, even though he knows strategy in his head perfectly. When your results don’t go the way you expected, as a non-compulsive gambler, you need to stop and examine the data, was it just bad luck or is something else going on? You should never have blind faith in your initial assumptions, instead expect to learn some lessons, expect to discover some previously unknown unknowns and adapt your play to them, that’s what Sun Tzu would do. Losing can be an opportunity to learn, if you so apply yourself.

#4

I’m not so sure you should limit Bob’s abilities at VP to the best in the free world. World’s best probably fits better.
Brad

#5

I agree if you are giving him the free world give him the whole thing. I don’t really see someone better coming out of Kamchatka or Nepal.

#6

Sports announcers are
notorious for crediting luck with skill.
Barry Melrose comes to mind. When
he’s asked what a certain team has to do to win, he’ll say something like
“the goalie has to play good and the defense has to play good.” John McEnroe stands out for emphasizing luck. Still, though, it’s not clear that they’re
evaluating the decision based only on the result. In activities that are less strictly
mathematical than video poker, it’s generally understood that results play a
larger part in evaluating decisions, ex ante, if only due to the lack of a cut
and dried roadmap.

My knowledge of Bayes’
Theorem isn’t very sophisticated. As far
as I know, it’s a matter of garbage in, garbage out, since the initial estimate
of the probability is always a factor in the final estimate of the probability. How would Bayes’ Theorem go about estimating
the chance that a video poker machine is gaffed? I understand that it would use the results of
play on the machine, but if it needs one’s initial estimate of the probability,
isn’t it dependent on a factor that, it’s assumed, has no value?

···

On Tue, Dec 20, 2016 at 12:17 PM, nightoftheiguana2…@…com [vpFREE] <vpF…@…com> wrote:

Bob wrote: “Whether or not you have made a good decision or a bad decision should be
determined at the time you make the decision — NOT sometime down the road.”

Grading (good or bad) a decision before the event is classical or a priori probability. Also grading sometime down the road or after the event is Bayesian probability. I would recommend the Bayesian approach, but hey, to each their own. Bayesian is superior when there are “unknown unknowns”, and, lets keep it real here, there are always “unknown unknowns”.

Bayesian probability - Wikipedia

Bayesian probability - Wikipedia
Bayesian probability is an interpretation of the concept of probability, in which, instead of frequency or propensity of some phenomenon, assigned probabilities represent states of knowledge[1] or belief.[2]

View on en.wikipedia.org

Preview by Yahoo

#7

Tom wrote: “How would Bayes’ Theorem go about estimating
the chance that a video poker machine is gaffed? I understand that it would use the results of
play on the machine, but if it needs one’s initial estimate of the probability,
isn’t it dependent on a factor that, it’s assumed, has no value?”

Almost all gambling involves an initial guess at the EV. One example I can think of that does not would be video poker with autohold (and a non-gaffed machine). Another example would be a kiosk promotion where you have equal weighted choices, also rare, most are weighted or even fixed. But the EV of most video poker (no autohold) varies widely depending on how on plays in the casino environment. That doesn’t mean the initial guess as no value, it’s either a high quality guess or a low quality guess, you use the actual results as they come in to fine tune your final estimate of what the true EV is. The point is that there is feedback, a control loop, you don’t just blindly dump all your bankroll on the assumption that the initial guess at EV is 100% correct. In the real world, results do matter. You’re either winning or losing and it’s either do to variance or an incorrect EV guess. You should figure out which it is and if it’s due to an incorrect EV guess, you should really make an adjustment, hence an adjustment based on results.

#8

An example: There’s a kiosk promotion, you might play a video poker machine depending on the EV of the kiosk promotion. In the promotion, you have a choice of three boxes, and it says the value behind each box is either $1, $5, or $1000, and after you make your choice, the other boxes are revealed so you can verify this. Your initial estimate of EV is: (1+5+1000)/3 = $335.33 . You use your watch’s second hand to randomize your pick (every gambler knows how to do this, right?) and after 10 picks you’ve received $1 every time. What’s your estimate of EV now? Still sticking with $335.33? How many picks of $1 will it take before you realize it’s a weighted or even fixed promotion? Or will you just play forever on the assumption that your initial guess at EV must be 100% correct? Your luck has to turn eventually, right?

#9

What does Bayes’ Theorem say the estimate is after 1 pick? I don’t see how the changing estimate of the probability of the truth of the original theory that the $1, the $5, and the $1000 were equally likely can be quantified.

···

On Fri, Dec 23, 2016 at 3:01 PM, nightoftheiguana2…@…com [vpFREE] <vpF…@…com> wrote:

An example: There’s a kiosk promotion, you might play a video poker machine depending on the EV of the kiosk promotion. In the promotion, you have a choice of three boxes, and it says the value behind each box is either $1, $5, or $1000, and after you make your choice, the other boxes are revealed so you can verify this. Your initial estimate of EV is: (1+5+1000)/3 = $335.33 . You use your watch’s second hand to randomize your pick (every gambler knows how to do this, right?) and after 10 picks you’ve received $1 every time. What’s your estimate of EV now? Still sticking with $335.33? How many picks of $1 will it take before you realize it’s a weighted or even fixed promotion? Or will you just play forever on the assumption that your initial guess at EV must be 100% correct? Your luck has to turn eventually, right?

#10

I have a similar question abut the Kelly Criterion. Ultimately, it’s utility that matters. The Kelly Criterion is applicable because it recognizes that money has diminishing marginal utility, but it’s anybody’s guess as to how accurately it estimates that utility.

···

On Sun, Dec 25, 2016 at 12:54 PM, Tom Robertson <007…@…com> wrote:

What does Bayes’ Theorem say the estimate is after 1 pick? I don’t see how the changing estimate of the probability of the truth of the original theory that the $1, the $5, and the $1000 were equally likely can be quantified.

On Fri, Dec 23, 2016 at 3:01 PM, nightoftheiguana2…@…com [vpFREE] <vpF…@…com> wrote:

An example: There’s a kiosk promotion, you might play a video poker machine depending on the EV of the kiosk promotion. In the promotion, you have a choice of three boxes, and it says the value behind each box is either $1, $5, or $1000, and after you make your choice, the other boxes are revealed so you can verify this. Your initial estimate of EV is: (1+5+1000)/3 = $335.33 . You use your watch’s second hand to randomize your pick (every gambler knows how to do this, right?) and after 10 picks you’ve received $1 every time. What’s your estimate of EV now? Still sticking with $335.33? How many picks of $1 will it take before you realize it’s a weighted or even fixed promotion? Or will you just play forever on the assumption that your initial guess at EV must be 100% correct? Your luck has to turn eventually, right?

#11

Tom wrote: “I have a similar question abut the Kelly Criterion. Ultimately, it’s utility that matters. The Kelly Criterion is applicable because it recognizes that money has diminishing marginal utility, but it’s anybody’s guess as to how accurately it estimates that utility.”

I would say the Kelly Criterion is useful because it tells you how much you need to hold in reserve when dealing with probabilistic events. I wouldn’t get hung up on the utility function of economics, but hey, if you want, you can go there.

#12

Tom wrote: “I don’t see how the changing estimate of the probability of the truth of
the original theory that the $1, the $5, and the $1000 were equally likely can be quantified.”

Bayesian probability is used to do the quantification of such situations.

#13

I’m not so sure you should limit Bob’s abilities at VP to the best in the free world.

When I grow up, I hope to play as well as people think Bob Dancer plays. I am arguably the best known “expert” player — but there are many intentionally not-so-well-known players with skills far superior
to mine.

And even when you’re determining “best,” it’s not a single-dimensional variable. Some people are better at figuring out the mathematics of a game. Some are better at intuiting how much slot club benefits
they are going to get for how much play. Some are better at talking casinos into giving them extra-good deals. Some are better are keeping their welcome when players with similar records are thrown out. Some are better at forming and keeping partnerships to
share in the scouting and keeing-the-machine duties. It all counts. And nobody is best at all these things.

#14

BINGO, well stated. My girlfriend hit a 20,000 royal twice in the same joint within four weeks of each other, and LOST money, net net, on the second one, not very smart.

···

Sent from Windows Mail

From: vpFREE3355 vpfree3…@…com [vpFREE]
Sent: ‎Thursday‎, ‎December‎ ‎29‎, ‎2016 ‎12‎:‎41‎ ‎AM
To: vpFREE3355 vpfree3…@…com [vpFREE]

I’m not so sure you should limit Bob’s abilities at VP to the best in the free world.

When I grow up, I hope to play as well as people think Bob Dancer plays. I am arguably the best known “expert” player — but there are many intentionally not-so-well-known players with skills far superior
to mine.

And even when you’re determining “best,” it’s not a single-dimensional variable. Some people are better at figuring out the mathematics of a game. Some are better at intuiting how much slot club benefits
they are going to get for how much play. Some are better at talking casinos into giving them extra-good deals. Some are better are keeping their welcome when players with similar records are thrown out. Some are better at forming and keeping partnerships to
share in the scouting and keeing-the-machine duties. It all counts. And nobody is best at all these things.