vpFREE2 Forums

Gaming Today Online.

--- In vpFREE@yahoogroups.com, "Harry Porter" <harry.porter@v...>
wrote:

So, you're saying that the fact that should this woman look at the
likelihood that she'll come away from her play with more than $80K,
she should simply shrug her shoulders at the immense difference is
probabilities. Hell, since she's likely to fall into the middle
ground anyways, why worry about the relative magnitude of upside and
downside potentials.

- Harry

I'm not sure I'm saying what you say, maybe I am.

Another little calculation tells us that with 99% confidence, this
lady will get back after playing her 16,000 hands in the dollar
machine, between $51,702 and $107,562 in the 9/6 JB machine; and
between $51,057 and $100,937 on the 6/5 JB machine. The lady is not
playing "long term" at all, even if her play covers her lifetime;
thus the two probability distributions largely overlap. Sure, the one
here for 9/6 JB is slightly more favorable, but are those numbers
really that important for so few hands? Let's look at other things.

Mr. Boutout has said, "Machines will hit heavy for a while & will go
cold sometimes. That is probability/volatility over the lifetime." He
has said other things, and anyone can check his messages. I don't
want to twist what he is saying, but I wanted to quote this sentence
of his because I feel it is pertinent for the lady. Should the lady
hit one of those stretches of hands when the machine is cold, say
perhaps an extreme paucity of flushes and full houses and/or an
extended draught of quads, her results on 16,000 hands can be largely
determined by that if she does not have sessions when the machine
hits heavy. Of course, the results can go the opposite way should she
hit only windfalls. This behavior of the machine is more important
for the results of her 16,000 hands than the calculations above. The
short-term behavior of the machines she plays is more important than
probability considerations for her results, and since she is not
playing too many hands, there is no time for that short-term behavior
to "smooth out".

Another consideration is royals. There are about 40,000 hands on
average between royals on the two machines. The lady is only playing
16,000 hands, and she could possibly not get any royal at all, or she
could get one or more. Of course the number of royals obtained will
also determine her outcome more than the numbers calculated above. It
is amusing to note that the cycle of royals for the 6/5 machine,
40,172, is shorter than the cycle of royals of 40,390 for the 9/6
machine.

Now, if this lady were to play a lot more hands –a different problem-
then her return would go up on each machine proportionally to the
number of hands, and the standard deviation only as the square root
of the number of hands. Thus, after many, many more hands than the
number she is playing, which is 16,000, the probability distributions
will look like two separate mountains with little overlap, and then
the advise to play 9/6 JB is very sound. This is not the case here.

Should this lady again ask me for advice on what machine to play, I
would have to tell her it does not really matter, if I'm being honest
with myself. The idea of sending her to play video Keno is not that
outrageous either, to address Glenbob's concerns. I saw in
thewizardofodds site a table of video Keno machines, and some of them
payback close to 95% which is close enough to the return of the 6/5
JB machine. Now, I would be careful before telling this lady, "go
play video Keno". I would have to find out something more about the
machine. If the volatility is too high the lower bracket above, say,
might go down to zero.

A crucial point in all of this, I repeat in a different way, is that
the lady has only one life. Should she have many lives, or should she
intend to play in her lifetime many, many more hands than 16,000, I
would tell her to worry about video poker pay tables. This would be a
different problem.

Now I ask you, am I giving this lady the wrong advice? Why? Is the
math incorrect? Is the logic faulty?

Thanks for your reply and thanks for not calling me a rodent or an
ignorant fool, though I might be both. Regards,

E.

"I have gathered a posy of other men's flowers, and nothing but the
thread that binds them is mine own." John Bartlett.

Harry Porter wrote:

> So, you're saying that the fact that should this woman look at the
> likelihood that she'll come away from her play with more than
> $80K, she should simply shrug her shoulders at the immense
> difference is probabilities. Hell, since she's likely to fall
> into the middle ground anyways, why worry about the relative
> magnitude of upside and downside potentials.

mubowar replied:

I'm not sure I'm saying what you say, maybe I am.

Another little calculation tells us that with 99% confidence, this
lady will get back after playing her 16,000 hands in the dollar
machine, between $51,702 and $107,562 in the 9/6 JB machine; and
between $51,057 and $100,937 on the 6/5 JB machine.

Ed, I'll ask your forebearance I've been running a bit tired and
haven't been quite as coherent as I would like.

My point is, using the numbers in this more recent post, that this
woman would have less than a 1% likelihood of walking away with more
than $101K from play of the 6/5 JB machine. On the otherhand, the
odds of that result playing 9/6 JB would clearly be several times
that. I think the woman would want the greater shot at a very nice
win even if, on average, her expectation in playing either machine
falls into a comparable middle ground.

But, bottom line, it should be common sense that giving up $15 with
every Full House and $5 with every flush has got to be a very costly
proposition, no matter how short your session is.

Ed, most importantly, you're looking at the sizable overlap of likely
play result distributions for these two machines and failing to see
the significant disparity in the probability of any given result. The
extent to which the 9/6 distribution lies to the right of the 6/5
distribution adds up to a lot of dough, and even in terms of a single
session involves a sizable difference.

This woman, and any player for that matter, should be very concerned
about the relative probability for a very large loss. In playing the
6/5 vs. the 9/6, the probabilities for such a loss are several times
greater.

(btw, the lower bounds of the 99% ranges cited above, both of which
are about $51K, seem suspiciously close given the appox. 3x greater
loss you'd expect in a 6/5 session vs. 9/6 when a RF isn't hit.)

- Harry

mubowar wrote:
> Another little calculation tells us that with 99% confidence, this
> lady will get back after playing her 16,000 hands in the dollar
> machine, between $51,702 and $107,562 in the 9/6 JB machine; and
> between $51,057 and $100,937 on the 6/5 JB machine.

I replied:

(btw, the lower bounds of the 99% ranges cited above, both of which
are about $51K, seem suspiciously close given the appox. 3x greater
loss you'd expect in a 6/5 session vs. 9/6 when a RF isn't hit.)

Again, I'll confess to being sufficiently "out of it" today that my
gut sense expressed in this last comment may well be misguided.

I'm always interested in the methodology used when precise statistical
estimates such as yours are stated. Care to discuss your calculations
in this case?

- Harry

Harry,

The calculation is the usual. You get the ER and the volatility
number from WinPoker. EV is that equation at the start of the dogma
people seem to love which I could spell out to you as ER * #5 coins*
# of hands. The standard deviation is the the square root of the
volatility in WinPoker, times the number of coins, 5, times the
square root of the number of hands. Now, using `Chevy Chase' theorem,
LOL, we multiply the standard deviation by ten and add it and
subtract it to the EV to get the bracket with 99% confidence. The
part that puzzles you is that in this case the lower bound is about
the same for both 9/6 and 6/5. There is a simple reason for this. EV
for 6/5 is lower than for 9/6, but the standard deviation for 9/6 is
larger than for 6/5. This comes from WinPoker. The volatility for 9/6
is 19.51 and the volatility for 6/5 is 19.04. Thus, the lower limit
of the brackets for the two machines are near each other because you
are subtracting a smaller number to the lower EV, and a larger number
for the larger EV. You might still say had you not been saturated
with caffeine, that this did not happen in the 75% case and you would
be right. Here the number one subtracts is larger. Capish? Analogy.
Subtract .6 from 4 and .4 from 3, and you get 3.7 and 2.2,
respectively, very different; subtract 3 from 4 and 2 from 3 and you
get 1 in both cases.

As far as the advise to the lady, my point about the short term does
not come through. The point is that EV does not matter much in the
short term; luck, in other words hits, does. Rather than repeat
myself, let me try to explain it differently, in an imaginary
dialogue. The characters are the lady, the chauffeur and I.

···

____________________________________________________
-I'm going to go to Las Vegas from Boston by way of Dedham and plan
to play a little video poker, could you recommend a good machine to
me?

-How long do you plan to play?

-I hope you are not implying by that question that I have become a
dissolute gambler. I plan to play a little there before I go on to
Palm Springs to visit Edna. Maybe half an hour, what I always play
once a year on my stop in Las Vegas.

-Then it does not matter what you play, frankly.

-Is that so? I have heard from Ethel that some machines, what do they
call them? Jacks or Better! Are they any good?

-Yes, I think so. The best pay table for those is 9/6, but often some
casinos only offer them in 6/5.

-What, pray, are 9/6 and 6/5?

-The coins paid per coin for a Full House and a Flush, respectively.
If you play five coins on a dollar machine, full houses and flushes
will pay you $15 and $5 more on the 9/6 machine than on the 6/5
machine.

-Is that so? I will get paid more for those hands in the 9/6 than in
the 6/5 machines. But that is only if I hit those hands. Is one of
those two machines more likely than the other to hit a flush or a
full house?

-I could not say so. I believe both are about equally likely to hit
either hand.

-So how do I decide? I don't care if one machine pays me more for a
full house and a flush if I don't have an assurance to hit a full
house or a flush.

-Well, madam, no one can give you such an assurance. I will say that
if you play long enough you will hit many flushes and full houses.

-Let me ask you the opposite question. If I play only one hand, am
not assured to hit a full house or a flush or something else on
either machine?

-No, madam, but should you hit either a full house or a flush, you
will collect more from the 9/6 machine than from the 6/5 machine.

-But my dear, can't you see? The important thing is whether one hits
a hand or not.

-Yes, I believe you are right, but a hit can't be predicted.

-Then nothing useful can be predicted from your mathematical theory.
If you can't lead me to a machine that hits, what good is your advice?

-You mustn't say that, madam, some people hang on every syllable that
comes from the mathematics. I can tell you that the ER of the 9/6
machine is higher than the ER of the 6/5 machine and after many hands
you will be more likely to get back an amount of money proportional
to that ER.

-That mathematical yabber always confuses me. I want to go for the
sure thing. Just tell me this, how long do I have to play the 9/6
machine to be sure to extract from it more money than from the 6/5
machine?

-Madam, a very long time. Not hours, or weeks but months or years.

-Are you being facetious?

-No, madam, I'm being dead serious. If you play either machine once,
your probability of getting something from either machine is the
same. If you play either machine a few thousands of hands, the hits
you make are largely due to luck, and your return could be larger on
either machine. It is only after being played for months or years or
more that the hits on both machines will be about the same, but since
the 9/6 pays more for flushes and full houses you will receive more
money from that machine than from the 6/5.

-Your mathematical gibberish confuses me again. I'm only going to
play for half an hour, and I don't intend to become a dissolute
gambler, I will only play that much every year on video poker.

-Sorry, I did not mean to be offensive, just accurate.

-Then you will not tell me what machine to play?

-I will tell you it makes no difference, madam.

-Madam, the tank is full of gas and ready to undertake the journey.

-Let's go. I will play those 9/6 Jacks or Better and let you know how
I do.
  
-Good luck, that's what you need. Fare well.

I honestly think I would have given bad advice to the imaginary lady
had I insisted on a certain machine and on her looking for it.

I know even this dialogue won't convince many. I foresee objections
along the lines that should this lady hit a lot of flushes and full
houses, she would be better off at 9/6. My point is that hitting or
not hitting is crucial in the short term. EV is more important in the
long term. If there are no hits in the short term, EV is irrelevant.
Let me repeat. The important thing in the short term is to have hits.
In the long term there will be enough hits to make the EV and all
related calculations very relevant.

Hope you enjoyed the play.

E.

--- In vpFREE@yahoogroups.com, "Harry Porter" <harry.porter@v...>
wrote:

> mubowar wrote:
> > Another little calculation tells us that with 99% confidence,

this

> > lady will get back after playing her 16,000 hands in the dollar
> > machine, between $51,702 and $107,562 in the 9/6 JB machine;

and

> > between $51,057 and $100,937 on the 6/5 JB machine.

I replied:
> (btw, the lower bounds of the 99% ranges cited above, both of

which

> are about $51K, seem suspiciously close given the appox. 3x

greater

> loss you'd expect in a 6/5 session vs. 9/6 when a RF isn't hit.)

Again, I'll confess to being sufficiently "out of it" today that my
gut sense expressed in this last comment may well be misguided.

I'm always interested in the methodology used when precise

statistical

estimates such as yours are stated. Care to discuss your

calculations

in this case?

- Harry

mubowor wrote:

The standard deviation is the the square root of the
volatility in WinPoker, times the number of coins, 5, times the
square root of the number of hands. Now, using `Chevy Chase'
theorem, LOL, we multiply the standard deviation by ten and add it
and subtract it to the EV to get the bracket with 99% confidence.

Be careful, Ed. This assumes that results adhere to a normal
distribution. That's not the case with video poker, particularly in
the short run.

- Harry

I'm mistaken, Chebyshev's Theorem is applicable to non-normal
distributions (aka non "bell-shaped").

- H.

···

mubowor wrote:
> The standard deviation is the the square root of the
> volatility in WinPoker, times the number of coins, 5, times the
> square root of the number of hands. Now, using `Chevy Chase'
> theorem, LOL, we multiply the standard deviation by ten and add it
> and subtract it to the EV to get the bracket with 99% confidence.

Be careful, Ed. This assumes that results adhere to a normal
distribution. That's not the case with video poker, particularly in
the short run.

- Harry

As far as the advise to the lady, my point about the short term

does

not come through. The point is that EV does not matter much in the
short term; luck, in other words hits, does.

I don't know whether EV as some mathematical concept "does not matter
much." I do know that in your example the choice for the rational
gambler is clear. If the lady doesn't hit a fullhouse or flush in
her session, she is neither worse or better off on either 9/6 or 6/5
machine. If she hits even a single fullhouse or flush she is much
better off on the 9/6 and much worse off on the 6/5. It seems like a
no brainer to me whether she is playing a half hour or a year or just
one hand.

> > -Then nothing useful can be predicted from your mathematical

theory.

If you can't lead me to a machine that hits, what good is your

advice?

-You mustn't say that, madam, some people hang on every syllable

that

comes from the mathematics. I can tell you that the ER of the 9/6
machine is higher than the ER of the 6/5 machine and after many

hands

you will be more likely to get back an amount of money proportional
to that ER.

-That mathematical yabber always confuses me. I want to go for the
sure thing. Just tell me this, how long do I have to play the 9/6
machine to be sure to extract from it more money than from the 6/5
machine?

You know that mathematical yabber always confuse me too, but I try to
follow as best I can;-) Next time you talk to this lady I suggest
you use my "consumer analogy" but substitute my example product of
soda with shoes. I predict she will smile, see the reasoning
instantly, and head to the first 9/6 machine. Immediately afterward
she will buy herself another pair of shoes... if she is lucky, with
the money from a couple of fullhouses and flushes;-)

···

--- In vpFREE@yahoogroups.com, "mubowor" <erchalb@c...> wrote: