On the full pay JB 99.54 the variance is 19.51.
However, when I changed the royal payoff to reflect a progressive on
the frugal software it showed the variance as much higher. Is this
right or did I make a mistake in setting up the program?
Thanks for any assistance.
al
Hi Al;
Variance is a reflection of payoffs and the frequency you get them. When you increase a payoff for a low frequency (long odds) hand such as a royal flush, the Variance will go up. If you alter your strategy to maximize your return from the game (using FVP's computer-generated strategy), the Variance will go up even higher, since you will often be playing the "high-variance play" over the lower variance play, such as holding a RF3 (3 to a royal) over a high pair.
As an example, a 7/5 bonus progressive with a triple royal has a return of 102% and a variance of 148 with a regular ("Flat-top") 7/5 strategy , but a return of 102.8% and a variance of 186 with the computer generated strategy for the progressive.
The good news is that with Frugal Video Poker, you can get either strategy or create something in between if you wish by "tweaking" the progressive strategy charts. You can choose if you wish, to sacrifice some Expected Return for a reduction in Variance.
Welcome to the wonderful world of FVP.
Skip
amarek1402 wrote:
On the full pay JB 99.54 the variance is 19.51.
However, when I changed the royal payoff to reflect a progressive on
the frugal software it showed the variance as much higher. Is this
right or did I make a mistake in setting up the program?Thanks for any assistance.
al
--
Thanks!
Skip
http://www.vpinsider.com
Skip Hughes wrote:
Variance is a reflection of payoffs and the frequency you get them.
When you increase a payoff for a low frequency (long odds) hand such
as a royal flush, the Variance will go up ...As an example, a 7/5 bonus progressive with a triple royal has a
return of 102% and a variance of 148 with a regular ("Flat-top") 7/5
strategy, but a return of 102.8% and a variance of 186 with the
computer generated strategy for the progressive.The good news is that with Frugal Video Poker, you can get either
strategy or create something in between if you wish by "tweaking" the
progressive strategy charts. You can choose if you wish, to sacrifice
some Expected Return for a reduction in Variance.
Academically, the point in your last paragraph is a fine one to make,
Skip. However, from a practical perspective, I'm inclined to
discourage anyone from attempting to moderate game variance through
strategy "tweaks".
It's my experience that tweaks to signficiantly reduce variance (as
represented by at least a 10% reduced ROR bankroll requirement) will
serve to disproportionately sacrifice ER as a consequence. It's
simply a bad move.
From what I can see, if you're uncomfortable with the game variance of
a play under consideration, the adage "if you can't stand the heat
..." pretty much applies.
- Harry
Hi Harry;
Interesting that you should bring that up, because in the July issue of Video Poker Player, I publish the results of an attempt to find a "reduced risk" strategy that has a measurable impact on Risk of Ruin, but a negligible impact on Expected Return. I used FVP and Dunbar's VP Analyzer as the tools. I will be doing more unusual strategies in the future, (hopefully some that will take a little less time investment). Without giving away the whole story I will say that I was able to achieve a small reduction in ROR with a fairly negligible loss of return. Variance, like expectations, isn't everything (sorry, Steve). It's an interesting study, I believe. To me, at least.
But as regards progressives, it's far from academic. There are a lot of people who prefer to play regular strategy on a big progressive or just make minor changes, sacrificing some return, while still getting a very big edge (like the 2% in the example). I seldom play progressives, but I would probably take a middle course. In fact, you usually don't have much choice. If you find a particular game with a particular jackpot, you can always run back to your room (or your home), fire up FVP, get the strategy, try to learn it and then return to the casino, but of course the jackpot will be gone. "Rules of thumb" come into play and I think most progressive players are happy with that solution.
Skip
Harry Porter wrote:
Skip Hughes wrote:
> Variance is a reflection of payoffs and the frequency you get them.
> When you increase a payoff for a low frequency (long odds) hand such
> as a royal flush, the Variance will go up ...
>
> As an example, a 7/5 bonus progressive with a triple royal has a
> return of 102% and a variance of 148 with a regular ("Flat-top") 7/5
> strategy, but a return of 102.8% and a variance of 186 with the
> computer generated strategy for the progressive.
>
> The good news is that with Frugal Video Poker, you can get either
> strategy or create something in between if you wish by "tweaking" the
> progressive strategy charts. You can choose if you wish, to sacrifice
> some Expected Return for a reduction in Variance.Academically, the point in your last paragraph is a fine one to make,
Skip. However, from a practical perspective, I'm inclined to
discourage anyone from attempting to moderate game variance through
strategy "tweaks".It's my experience that tweaks to signficiantly reduce variance (as
represented by at least a 10% reduced ROR bankroll requirement) will
serve to disproportionately sacrifice ER as a consequence. It's
simply a bad move.>From what I can see, if you're uncomfortable with the game variance of
a play under consideration, the adage "if you can't stand the heat
..." pretty much applies.- Harry
--
Thanks!
Skip
http://www.vpinsider.com
I thought I discovered a "winning strategy tweak" a while ago.
While playing on a 100-play machine, I would drop down to 10-hands at
one coin after a big hand, stay there for 4-5 hands, and then switch
to 100 hands at 5 coins. It seemed to work for a while.
Alas, like all the "systems" that are designed to beat the venerable
RNG, it failed after a while. And failed royally.
Don't attempt Strategy Tweaks when you are gambling. Speaking of
gambling, don't attempt Market Timing either.
A Myth
--- In vpFREE@yahoogroups.com, "Harry Porter" <harry.porter@...>
wrote:
Skip Hughes wrote:
> Variance is a reflection of payoffs and the frequency you get
them.
> When you increase a payoff for a low frequency (long odds) hand
such
> as a royal flush, the Variance will go up ...
>
> As an example, a 7/5 bonus progressive with a triple royal has a
> return of 102% and a variance of 148 with a regular ("Flat-top")
7/5
> strategy, but a return of 102.8% and a variance of 186 with the
> computer generated strategy for the progressive.
>
> The good news is that with Frugal Video Poker, you can get either
> strategy or create something in between if you wish by "tweaking"
the
> progressive strategy charts. You can choose if you wish, to
sacrifice
> some Expected Return for a reduction in Variance.
Academically, the point in your last paragraph is a fine one to
make,
Skip. However, from a practical perspective, I'm inclined to
discourage anyone from attempting to moderate game variance through
strategy "tweaks".It's my experience that tweaks to signficiantly reduce variance (as
represented by at least a 10% reduced ROR bankroll requirement) will
serve to disproportionately sacrifice ER as a consequence. It's
simply a bad move.From what I can see, if you're uncomfortable with the game variance
of
a play under consideration, the adage "if you can't stand the heat
..." pretty much applies.- Harry
Skip, Howard and Harry thanks for the replys.
Skip If I understand this right then if one plays the normal strategy for FPJB on a progressive he maintains the lower variance , lower ROR but a reduced ER. If that is the case how would one determine how much ER would be sacrificed to maintain low variance and low ROR?
Thanks
Al
----- Original Message -----
From: Skip Hughes
To: vpFREE@yahoogroups.com
Sent: Monday, July 17, 2006 3:44 PM
Subject: Re: [vpFREE] Re: "progressive newbies question"
Hi Harry;
Interesting that you should bring that up, because in the July issue of
Video Poker Player, I publish the results of an attempt to find a
"reduced risk" strategy that has a measurable impact on Risk of Ruin,
but a negligible impact on Expected Return. I used FVP and Dunbar's VP
Analyzer as the tools. I will be doing more unusual strategies in the
future, (hopefully some that will take a little less time investment).
Without giving away the whole story I will say that I was able to
achieve a small reduction in ROR with a fairly negligible loss of
return. Variance, like expectations, isn't everything (sorry, Steve).
It's an interesting study, I believe. To me, at least.
But as regards progressives, it's far from academic. There are a lot of
people who prefer to play regular strategy on a big progressive or just
make minor changes, sacrificing some return, while still getting a very
big edge (like the 2% in the example). I seldom play progressives, but I
would probably take a middle course. In fact, you usually don't have
much choice. If you find a particular game with a particular jackpot,
you can always run back to your room (or your home), fire up FVP, get
the strategy, try to learn it and then return to the casino, but of
course the jackpot will be gone. "Rules of thumb" come into play and I
think most progressive players are happy with that solution.
Skip
Harry Porter wrote:
>
> Skip Hughes wrote:
> > Variance is a reflection of payoffs and the frequency you get them.
> > When you increase a payoff for a low frequency (long odds) hand such
> > as a royal flush, the Variance will go up ...
> >
> > As an example, a 7/5 bonus progressive with a triple royal has a
> > return of 102% and a variance of 148 with a regular ("Flat-top") 7/5
> > strategy, but a return of 102.8% and a variance of 186 with the
> > computer generated strategy for the progressive.
> >
> > The good news is that with Frugal Video Poker, you can get either
> > strategy or create something in between if you wish by "tweaking" the
> > progressive strategy charts. You can choose if you wish, to sacrifice
> > some Expected Return for a reduction in Variance.
>
> Academically, the point in your last paragraph is a fine one to make,
> Skip. However, from a practical perspective, I'm inclined to
> discourage anyone from attempting to moderate game variance through
> strategy "tweaks".
>
> It's my experience that tweaks to signficiantly reduce variance (as
> represented by at least a 10% reduced ROR bankroll requirement) will
> serve to disproportionately sacrifice ER as a consequence. It's
> simply a bad move.
>
> >From what I can see, if you're uncomfortable with the game variance of
> a play under consideration, the adage "if you can't stand the heat
> ..." pretty much applies.
>
> - Harry
>
>
--
Thanks!
Skip
http://www.vpinsider.com
[Non-text portions of this message have been removed]
Al;
Yes, The variance is higher than normal JB and less that with a progressive strategy, but the Risk of Ruin should be lower than the progressive strategy in the short-term (due to the strategy changes, or rather lack thereof) and lower than normal JB in just about any term (due to the higher return). Of course, you might think that a progressive is always short-term, but if one plays them all the time, you have to take the long-term into consideration. It would be interesting to set up Dunbar to analyze this sometime.
As for the Expected Return, it's easy to calculate with FVP, you just set up the progressive game, then inherit the normal JB strategy.
Thanks,
Skip
Al Marek wrote:
Skip, Howard and Harry thanks for the replys.
Skip If I understand this right then if one plays the normal strategy for FPJB on a progressive he maintains the lower variance , lower ROR but a reduced ER. If that is the case how would one determine how much ER would be sacrificed to maintain low variance and low ROR?
Thanks
Al
----- Original Message -----
From: Skip Hughes
To: vpFREE@yahoogroups.com <mailto:vpFREE%40yahoogroups.com>
Sent: Monday, July 17, 2006 3:44 PM
Subject: Re: [vpFREE] Re: "progressive newbies question"Hi Harry;
Interesting that you should bring that up, because in the July issue of
Video Poker Player, I publish the results of an attempt to find a
"reduced risk" strategy that has a measurable impact on Risk of Ruin,
but a negligible impact on Expected Return. I used FVP and Dunbar's VP
Analyzer as the tools. I will be doing more unusual strategies in the
future, (hopefully some that will take a little less time investment).
Without giving away the whole story I will say that I was able to
achieve a small reduction in ROR with a fairly negligible loss of
return. Variance, like expectations, isn't everything (sorry, Steve).
It's an interesting study, I believe. To me, at least.But as regards progressives, it's far from academic. There are a lot of
people who prefer to play regular strategy on a big progressive or just
make minor changes, sacrificing some return, while still getting a very
big edge (like the 2% in the example). I seldom play progressives, but I
would probably take a middle course. In fact, you usually don't have
much choice. If you find a particular game with a particular jackpot,
you can always run back to your room (or your home), fire up FVP, get
the strategy, try to learn it and then return to the casino, but of
course the jackpot will be gone. "Rules of thumb" come into play and I
think most progressive players are happy with that solution.Skip
Harry Porter wrote:
>
> Skip Hughes wrote:
> > Variance is a reflection of payoffs and the frequency you get them.
> > When you increase a payoff for a low frequency (long odds) hand such
> > as a royal flush, the Variance will go up ...
> >
> > As an example, a 7/5 bonus progressive with a triple royal has a
> > return of 102% and a variance of 148 with a regular ("Flat-top") 7/5
> > strategy, but a return of 102.8% and a variance of 186 with the
> > computer generated strategy for the progressive.
> >
> > The good news is that with Frugal Video Poker, you can get either
> > strategy or create something in between if you wish by "tweaking" the
> > progressive strategy charts. You can choose if you wish, to sacrifice
> > some Expected Return for a reduction in Variance.
>
> Academically, the point in your last paragraph is a fine one to make,
> Skip. However, from a practical perspective, I'm inclined to
> discourage anyone from attempting to moderate game variance through
> strategy "tweaks".
>
> It's my experience that tweaks to signficiantly reduce variance (as
> represented by at least a 10% reduced ROR bankroll requirement) will
> serve to disproportionately sacrifice ER as a consequence. It's
> simply a bad move.
>
> >From what I can see, if you're uncomfortable with the game variance of
> a play under consideration, the adage "if you can't stand the heat
> ..." pretty much applies.
>
> - Harry
>--
Thanks!
Skip
http://www.vpinsider.com[Non-text portions of this message have been removed]
--
Thanks!
Skip
http://www.vpinsider.com
Skip Hughes wrote:
Interesting that you should bring that up, because in the July issue
of Video Poker Player, I publish the results of an attempt to find a
"reduced risk" strategy that has a measurable impact on Risk of Ruin,
but a negligible impact on Expected Return. I used FVP and Dunbar's
VP Analyzer as the tools.Without giving away the whole story I will say that I was able to
achieve a small reduction in ROR with a fairly negligible loss of
return. Variance, like expectations, isn't everything (sorry, Steve).
It's an interesting study, I believe. To me, at least.But as regards progressives, it's far from academic. There are a lot
of people who prefer to play regular strategy on a big progressive or
just make minor changes, sacrificing some return, while still getting
a very big edge (like the 2% in the example). I seldom play
progressives, but I would probably take a middle course. In fact, you
usually don't have much choice.
It'll be interesting to see what you've arrived at.
My bias against any strategy modification to manage risk arose, in
part, from earlier discussions here concerning alternate play
strategies, for which Steve Jacobs was the driving participant. The
strategies he introduced are discussed in an appendix to the Group FAQ
and can be found at http://members.cox.net/vpfree/FAQ_S.htm
He presents 10% ROR bankrolls for 9/6 Jacks with a 1300 unit RF. The
difference in bankroll between the traditional Max-EV and Min-Risk
strategies comes to a little over 2%.
Even though the ER sacrifice in moving to the Min-Risk strategy is
very modest (under .05%), I don't consider the bankroll reduction to
be sizable enough to warrant the ER cut -- or, for that matter, enough
to warrant the effort to modify strategy/practice even when setting
aside the ER consideration. I simply don't find that the bankroll
reduction is of a magnitude to make the play any more approachable
when you adjust strategy to reduce risk to the greatest extent possible.
Bottom line, if you aren't comfortable taking on the risk of a max-EV
strategy for a play, moving to a min-Risk strategy isn't going to
provide appreciably greater comfort. (Again, it's the "stand the heat
of the kitchen" thing.)
------------
My "academic" reference was specifically addressed to any attempt to
reduce risk in this manner. I recognize that I may have unduly
fixated upon that aspect of your post.
There's no question that the strategy question posed by a juicy
progressive is a key challenge of play. As you remark, the average
recreational player will pretty much stick to standard strategy
(possibly winging a few appropriate strategy changes on the fly --
potentially from the advice of more knowledgable players on the bank).
This, of course, presumes that the base paytable is reasonably
familiar and not something like 8/5 DB 
The players I most respect who are in the serious but casual category
(vs. "pro") are pre-armed with a limited number of key strategy
changes for certain breakpoints as the meter climbs -- allowing them
to glean a respectable portion of the incremental ER that can be
achieved by adjusting strategy as the RF pay advances. They focus on
the progressive games that they most likely expect to encounter (e.g.
9/6 Jacks, 9/6 DB, UD, etc - depending upon locale).
- Harry
Hi Harry;
Interesting that you should bring that up, because in the July issue of
Video Poker Player, I publish the results of an attempt to find a
"reduced risk" strategy that has a measurable impact on Risk of Ruin,
but a negligible impact on Expected Return. I used FVP and Dunbar's VP
Analyzer as the tools. I will be doing more unusual strategies in the
future, (hopefully some that will take a little less time investment).
Without giving away the whole story I will say that I was able to
achieve a small reduction in ROR with a fairly negligible loss of
return. Variance, like expectations, isn't everything (sorry, Steve).
It's an interesting study, I believe. To me, at least.
It it good to hear that someone else has caught the "alternate strategy"
bug. If my aging memory hasn't failed me completely, there should
be an article or two in the archive where I describe my method for
producing min-risk strategies. This method produces the playing
strategy with the lowest possible ROR. Now other strategy can have
lower ROR. All that is required to produce min-risk strategies is a
VP analyzer that allows fractional values in the payoff table. With
such a VP program and appropriate math, it is possible to produce
a wide variety of alternate strategies which are exactly optimal for
whatever mathematical objective is optimized.
If you contact me by email I'm sure there are some pointers I can
give you to help evaluate new strategies. I've developed a generic
framework that appears to apply to a wide variety of goals.
On Monday 17 July 2006 4:44 pm, Skip Hughes wrote:
But as regards progressives, it's far from academic. There are a lot of
people who prefer to play regular strategy on a big progressive or just
make minor changes, sacrificing some return, while still getting a very
big edge (like the 2% in the example). I seldom play progressives, but I
would probably take a middle course. In fact, you usually don't have
much choice. If you find a particular game with a particular jackpot,
you can always run back to your room (or your home), fire up FVP, get
the strategy, try to learn it and then return to the casino, but of
course the jackpot will be gone. "Rules of thumb" come into play and I
think most progressive players are happy with that solution.Skip
Harry Porter wrote:
> Skip Hughes wrote:
> > Variance is a reflection of payoffs and the frequency you get them.
> > When you increase a payoff for a low frequency (long odds) hand such
> > as a royal flush, the Variance will go up ...
> >
> > As an example, a 7/5 bonus progressive with a triple royal has a
> > return of 102% and a variance of 148 with a regular ("Flat-top") 7/5
> > strategy, but a return of 102.8% and a variance of 186 with the
> > computer generated strategy for the progressive.
> >
> > The good news is that with Frugal Video Poker, you can get either
> > strategy or create something in between if you wish by "tweaking" the
> > progressive strategy charts. You can choose if you wish, to sacrifice
> > some Expected Return for a reduction in Variance.
>
> Academically, the point in your last paragraph is a fine one to make,
> Skip. However, from a practical perspective, I'm inclined to
> discourage anyone from attempting to moderate game variance through
> strategy "tweaks".
>
> It's my experience that tweaks to signficiantly reduce variance (as
> represented by at least a 10% reduced ROR bankroll requirement) will
> serve to disproportionately sacrifice ER as a consequence. It's
> simply a bad move.
>
> >From what I can see, if you're uncomfortable with the game variance of
>
> a play under consideration, the adage "if you can't stand the heat
> ..." pretty much applies.
>
> - Harry
Al;
Yes, The variance is higher than normal JB and less that with a
progressive strategy, but the Risk of Ruin should be lower than the
progressive strategy in the short-term (due to the strategy changes, or
rather lack thereof) and lower than normal JB in just about any term
(due to the higher return). Of course, you might think that a
progressive is always short-term, but if one plays them all the time,
you have to take the long-term into consideration. It would be
interesting to set up Dunbar to analyze this sometime.
It the intent is to reduce the risk of going broke while playing to hit
a progressive, then a good strategy is to minimize "RoR before Royal".
This strategy maximizes the probability that your bankroll will survive
until you hit a royal flush. There was some discussion of this several
months ago.
On Monday 17 July 2006 9:19 pm, Skip Hughes wrote:
As for the Expected Return, it's easy to calculate with FVP, you just
set up the progressive game, then inherit the normal JB strategy.
Thanks,
SkipAl Marek wrote:
> Skip, Howard and Harry thanks for the replys.
>
> Skip If I understand this right then if one plays the normal strategy
> for FPJB on a progressive he maintains the lower variance , lower ROR
> but a reduced ER. If that is the case how would one determine how much
> ER would be sacrificed to maintain low variance and low ROR?
>
> Thanks
>
> Al
> ----- Original Message -----
> From: Skip Hughes
> To: vpFREE@yahoogroups.com <mailto:vpFREE%40yahoogroups.com>
> Sent: Monday, July 17, 2006 3:44 PM
> Subject: Re: [vpFREE] Re: "progressive newbies question"
>
> Hi Harry;
> Interesting that you should bring that up, because in the July issue of
> Video Poker Player, I publish the results of an attempt to find a
> "reduced risk" strategy that has a measurable impact on Risk of Ruin,
> but a negligible impact on Expected Return. I used FVP and Dunbar's VP
> Analyzer as the tools. I will be doing more unusual strategies in the
> future, (hopefully some that will take a little less time investment).
> Without giving away the whole story I will say that I was able to
> achieve a small reduction in ROR with a fairly negligible loss of
> return. Variance, like expectations, isn't everything (sorry, Steve).
> It's an interesting study, I believe. To me, at least.
>
> But as regards progressives, it's far from academic. There are a lot of
> people who prefer to play regular strategy on a big progressive or just
> make minor changes, sacrificing some return, while still getting a very
> big edge (like the 2% in the example). I seldom play progressives, but I
> would probably take a middle course. In fact, you usually don't have
> much choice. If you find a particular game with a particular jackpot,
> you can always run back to your room (or your home), fire up FVP, get
> the strategy, try to learn it and then return to the casino, but of
> course the jackpot will be gone. "Rules of thumb" come into play and I
> think most progressive players are happy with that solution.
>
> Skip
>
> Harry Porter wrote:
> > Skip Hughes wrote:
> > > Variance is a reflection of payoffs and the frequency you get them.
> > > When you increase a payoff for a low frequency (long odds) hand such
> > > as a royal flush, the Variance will go up ...
> > >
> > > As an example, a 7/5 bonus progressive with a triple royal has a
> > > return of 102% and a variance of 148 with a regular ("Flat-top") 7/5
> > > strategy, but a return of 102.8% and a variance of 186 with the
> > > computer generated strategy for the progressive.
> > >
> > > The good news is that with Frugal Video Poker, you can get either
> > > strategy or create something in between if you wish by "tweaking" the
> > > progressive strategy charts. You can choose if you wish, to sacrifice
> > > some Expected Return for a reduction in Variance.
> >
> > Academically, the point in your last paragraph is a fine one to make,
> > Skip. However, from a practical perspective, I'm inclined to
> > discourage anyone from attempting to moderate game variance through
> > strategy "tweaks".
> >
> > It's my experience that tweaks to signficiantly reduce variance (as
> > represented by at least a 10% reduced ROR bankroll requirement) will
> > serve to disproportionately sacrifice ER as a consequence. It's
> > simply a bad move.
> >
> > >From what I can see, if you're uncomfortable with the game variance of
> >
> > a play under consideration, the adage "if you can't stand the heat
> > ..." pretty much applies.
> >
> > - Harry
>
> --
> Thanks!
> Skip
> http://www.vpinsider.com
>
> [Non-text portions of this message have been removed]
Steve;
I've had an interest for a long time. In fact, I once wrote an article (casino Player, I think) about a reduced risk strategy for the 9/7 Jacks or Better at the Stratosphere. That tells you how long ago it was. I've been limited in working on those kinds of things because I have no real math skills and need good tools like FVP and Dunbar's analyzer. But I did follow some of your discussion here and in fact, credit you in my article with inspiring my interest in the subject. I'll be in contact with you for those pointers.
Thanks!
Skip
Steve Jacobs wrote:
It it good to hear that someone else has caught the "alternate strategy"
bug. If my aging memory hasn't failed me completely, there should
be an article or two in the archive where I describe my method for
producing min-risk strategies. This method produces the playing
strategy with the lowest possible ROR. Now other strategy can have
lower ROR. All that is required to produce min-risk strategies is a
VP analyzer that allows fractional values in the payoff table. With
such a VP program and appropriate math, it is possible to produce
a wide variety of alternate strategies which are exactly optimal for
whatever mathematical objective is optimized.If you contact me by email I'm sure there are some pointers I can
give you to help evaluate new strategies. I've developed a generic
framework that appears to apply to a wide variety of goals.
Steve,
Nice to hear from you once again! I appreciated your posts a few months back, and I'm glad the vpFREE admin posted the information at http://members.cox.net/vpfree/FAQ_S.htm but I'm still trying to fully understand it. I have a few questions:
1. How do you come up with values to use in the paytable for a given goal?
Your 9/6 min-cost-royal paytable uses a value of 4879.97 for the royal, which makes the game's ER 100.0% - that I can understand (I think). But what about the others, like min-risk (which seems like the most important to understand)? You refer to a formula for stretching the values; what is the formula you used? I understand the computations and strategy generation once you have a paytable, but arriving at the adjusted values is somewhat of a mystery to me. And you could say I'm math-inclined, so I must not be the only one! The descriptions of the process on the above page don't quite get me there. You say it's complicated, so maybe I'm opening a can of worms here. I also get max-royal (we've discussed that here in relation to tournaments). You set everything but the royal to zero (and the value for the royal payoff is irrelevant, it can be 1 or 4000 or a million - you'll still get the same resulting strategy). So perhaps you can tell us how you arrived at the min-risk
paytable, that should help.
2. Can we obtain the software you used for this analysis?
You apparently have software of your own that allows fractional paytables that you use for these purposes. Are you willing to give it away or sell it? I'm sure some others here might be interested in it. I have begun developing such a beast, but it's not finished.
It's an interesting universe you've explored, where max-EV is not always the goal but just one possible goal. I'm hoping to understand it better.
John
Steve Jacobs <jacobs@xmission.com> wrote:
It it good to hear that someone else has caught the "alternate strategy"
bug. If my aging memory hasn't failed me completely, there should
be an article or two in the archive where I describe my method for
producing min-risk strategies. This method produces the playing
strategy with the lowest possible ROR. Now other strategy can have
lower ROR. All that is required to produce min-risk strategies is a
VP analyzer that allows fractional values in the payoff table. With
such a VP program and appropriate math, it is possible to produce
a wide variety of alternate strategies which are exactly optimal for
whatever mathematical objective is optimized.
If you contact me by email I'm sure there are some pointers I can
give you to help evaluate new strategies. I've developed a generic
framework that appears to apply to a wide variety of goals.
---------------------------------
Do you Yahoo!?
Next-gen email? Have it all with the all-new Yahoo! Mail Beta.
[Non-text portions of this message have been removed]
Steve,
Nice to hear from you once again! I appreciated your posts a few months
back, and I'm glad the vpFREE admin posted the information at
http://members.cox.net/vpfree/FAQ_S.htm but I'm still trying to fully
understand it. I have a few questions:1. How do you come up with values to use in the paytable for a given
goal?
That is the hardest part of all -- finding a mathematical relationship
between the number of units in the payoff and the real "value" of that
payoff in terms of the goal.
Your 9/6 min-cost-royal paytable uses a value of 4879.97 for the royal,
which makes the game's ER 100.0% - that I can understand (I think). But
what about the others, like min-risk (which seems like the most important
to understand)? You refer to a formula for stretching the values; what is
the formula you used?
For min-risk, the formula is:
VP = (1 - R^N) / (1 - R)
where VP is the "Virtual Payoff", R is the risk parameter for the strategy
being used (same as RoR if the game if favorable to the player)
and N is the actual payoff value.
The min-risk strategy seeks to maximize the probability of playing
forever without going broke. When starting with a single unit, R is
the probability of ruin, so (1 - R) is the probability of success. Now,
if you play that one unit and hit a payoff of N units, then you've traded
your initial one-unit bankroll for an N-unit bankroll. The probability of
eventually losing the entire N unit bankroll is R^N, so the probability
of going on to play forever is (1 - R^N). So, (1 - R^N) is the probability
of success for playing an N unit bankroll.
You could just use the (1 - R^N) values for the virtual payoffs, because
they are the raw probabilities of success that go with each payoff N.
In fact, you could take these raw probabilities and scale them all up
by any factor you want, as long as all of them are scaled by the same
factor, and you could use those number to represent the "value" of
the payoffs. Scaling by dividing them all by (1 - R) does two things
that are useful. First, it gives numbers for virtual payoffs that are "close"
to the actual payoff value, and this can give you a feel for how the
goal stretches the values compared to computing EV. The other thing
this does is express the virtual payoffs as a multiple of the chance for
success that you get from a bankroll of one unit. So, if a payoff of 50
units gives a virtual payoff of 49.6 units, then this means that a payoff
of 50 increases your probability of playing forever by a factor of 49.6
compared to you chance of success for a single unit bankroll. Similarly,
a royal payoff of 1300 units might give a virtual payoff of 1009 units,
which would mean that turning one unit into 1300 units increases your
probability of playing forever by a factor of 1009. So, 1009 is the "value"
of that payoff in terms of the probability of reaching your goal of
playing forever.
These raw success probabilities satisfy the equation:
(1 - R) = p(1)*(1 - R) + p(2)*(1 - R^2) + ... + p(25)*(1 - R^25)
+ p(50)*(1 - R^50) + p(1300)*(1 - R^1300)
The real meaning of the above equation is this: The overall probability
of success for a 1 unit bankroll is (1 - R) and that value (on the left side
of the equation) is equal to the sum of all the possibilities that can
result from playing the game. For each possible payoff N, p(N) is the
probability of receiving that payoff, and (1 - R^N) is the probability
that those N units will be enough to allow you to keep playing forever.
Now if we divide both sides of the equation by (1 - R), we get an equation
and has the number one on the left, and on the right is a formula that
multiplies the p(N) values by the "virtual payoffs" [(1 - R^N) / (1 - R)].
This new equation means "the average value of the virtual payoffs is one".
In other words, when the game is viewed from the perspective of these
virtual payoffs, it appears that we are playing a breakeven game.
I understand the computations and strategy generation
once you have a paytable, but arriving at the adjusted values is somewhat
of a mystery to me. And you could say I'm math-inclined, so I must not be
the only one! The descriptions of the process on the above page don't quite
get me there. You say it's complicated, so maybe I'm opening a can of worms
here. I also get max-royal (we've discussed that here in relation to
tournaments). You set everything but the royal to zero (and the value for
the royal payoff is irrelevant, it can be 1 or 4000 or a million - you'll
still get the same resulting strategy). So perhaps you can tell us how you
arrived at the min-risk paytable, that should help.
I hope the description about helped rather than just making it more confusing,
but it was kind of rushed (I'm late for work) so please ask more questions if
it doesn't make sense.
2. Can we obtain the software you used for this analysis?
You apparently have software of your own that allows fractional paytables
that you use for these purposes. Are you willing to give it away or sell
it? I'm sure some others here might be interested in it. I have begun
developing such a beast, but it's not finished.
My program only works for games with no wild cards, and it is an unfinished
work, so I'd rather no distribute it.
It's an interesting universe you've explored, where max-EV is not always
the goal but just one possible goal. I'm hoping to understand it better.
Yes it is, and I'm always looking for new types of goals to evaluate.
Someone here recently asked about getting the best shot at reaching
a fixed coin-in requirement, as their definition of "success". This leads
to a complex strategy that has some interesting characteristics, so I'm
trying to find some time to explore that goal in more detail.
On Thursday 20 July 2006 2:04 am, John Douglass wrote:
Eureka! You have put in words something nagging me for a while.
The "real value" of a payoff, specifically applied to MultiStrike game.
Make it Five Play MultiStrike, to add a little excitement.
Bob Dancer argues that this is taken care off by adding 6, 4, 2, to the
payoffs at Levels 1, 2, 3 respectively, in these two articles.
http://www.igtproducts.com/IGTproducts/GameReview/MultiStikePoker/MultiS
trikePoker.htm
http://www.casinogaming.com/columnists/dancer/2005/0628.html
I am not quite understanding the logic there. Is that based on the goal
of maximizing the payoff for that level, assuming any sucess there
contributes potentially to that at higher levels?
What if the goal is simply to advance, at the cost of sacrificing a
higher pay off at the current level? (Am I stating the problem
correctly?)
What if you are dealt (J987),J?
Holding the four-to Straight Flush is the preferred play, but holding
the high pair guarantees advancing to the next level, on all five
hands, with all the potentialities of the next hand.
Is this adequately taken into the mix, in the 6/4/2 rule?
Inquiring minds want to know.
A Myth
PS: Unlike Harry Porter, I have no delusions of anytime soon honing my
play to the .05 level (let alone the holy grail .01). So I am content
in whiling away my time indulging in academic pursuits. But he is
right. The time is better spent in practice, practice, practice.
--- In vpFREE@yahoogroups.com, Steve Jacobs <jacobs@...> wrote:
That is the hardest part of all -- finding a mathematical relationship
between the number of units in the payoff and the real "value" of that
payoff in terms of the goal.
Okay as the progressive newbie and mathmatically challenged I'll ask the following question. What does one sacrifice if one plays a progressive with the basic 9/6 JOB stategy? Does one retain the 19.51 variance and lower volatility on the bankroll ? What happens to the expected return? What other sacrifices does one make?
Any help would be benefical in understanding this.
Thanks
Al
----- Original Message -----
From: Skip Hughes
To: vpFREE@yahoogroups.com
Sent: Monday, July 17, 2006 10:35 AM
Subject: Re: [vpFREE] "progressive newbies question"
Hi Al;
Variance is a reflection of payoffs and the frequency you get them.
When you increase a payoff for a low frequency (long odds) hand such as
a royal flush, the Variance will go up. If you alter your strategy to
maximize your return from the game (using FVP's computer-generated
strategy), the Variance will go up even higher, since you will often be
playing the "high-variance play" over the lower variance play, such as
holding a RF3 (3 to a royal) over a high pair.
As an example, a 7/5 bonus progressive with a triple royal has a return
of 102% and a variance of 148 with a regular ("Flat-top") 7/5 strategy ,
but a return of 102.8% and a variance of 186 with the computer generated
strategy for the progressive.
The good news is that with Frugal Video Poker, you can get either
strategy or create something in between if you wish by "tweaking" the
progressive strategy charts. You can choose if you wish, to sacrifice
some Expected Return for a reduction in Variance.
Welcome to the wonderful world of FVP.
Skip
amarek1402 wrote:
>
> On the full pay JB 99.54 the variance is 19.51.
> However, when I changed the royal payoff to reflect a progressive on
> the frugal software it showed the variance as much higher. Is this
> right or did I make a mistake in setting up the program?
>
> Thanks for any assistance.
>
> al
>
>
--
Thanks!
Skip
http://www.vpinsider.com
[Non-text portions of this message have been removed]
Hi Harry;
Interesting that you should bring that up, because in the July
issue of
Video Poker Player, I publish the results of an attempt to find a
"reduced risk" strategy that has a measurable impact on Risk of
Ruin,
but a negligible impact on Expected Return. I used FVP and Dunbar's
VP
Analyzer as the tools. I will be doing more unusual strategies in
the
future, (hopefully some that will take a little less time
investment).
Without giving away the whole story I will say that I was able to
achieve a small reduction in ROR with a fairly negligible loss of
return. Variance, like expectations, isn't everything (sorry,
Steve).
It's an interesting study, I believe. To me, at least.
But as regards progressives, it's far from academic. There are a
lot of
people who prefer to play regular strategy on a big progressive or
just
make minor changes, sacrificing some return, while still getting a
very
big edge (like the 2% in the example). I seldom play progressives,
but I
would probably take a middle course. In fact, you usually don't
have
much choice. If you find a particular game with a particular
jackpot,
you can always run back to your room (or your home), fire up FVP,
get
the strategy, try to learn it and then return to the casino, but of
course the jackpot will be gone. "Rules of thumb" come into play
and I
think most progressive players are happy with that solution.
Skip
Harry Porter wrote:
>
> Skip Hughes wrote:
> > Variance is a reflection of payoffs and the frequency you get
them.
> > When you increase a payoff for a low frequency (long odds) hand
such
> > as a royal flush, the Variance will go up ...
> >
> > As an example, a 7/5 bonus progressive with a triple royal has a
> > return of 102% and a variance of 148 with a regular ("Flat-
top") 7/5
> > strategy, but a return of 102.8% and a variance of 186 with the
> > computer generated strategy for the progressive.
> >
> > The good news is that with Frugal Video Poker, you can get
either
> > strategy or create something in between if you wish
by "tweaking" the
> > progressive strategy charts. You can choose if you wish, to
sacrifice
> > some Expected Return for a reduction in Variance.
>
> Academically, the point in your last paragraph is a fine one to
make,
> Skip. However, from a practical perspective, I'm inclined to
> discourage anyone from attempting to moderate game variance
through
> strategy "tweaks".
>
> It's my experience that tweaks to signficiantly reduce variance
(as
> represented by at least a 10% reduced ROR bankroll requirement)
will
> serve to disproportionately sacrifice ER as a consequence. It's
> simply a bad move.
>
> >From what I can see, if you're uncomfortable with the game
variance of
> a play under consideration, the adage "if you can't stand the heat
> ..." pretty much applies.
>
> - Harry
>
>--
Thanks!
Skip
http://www.vpinsider.com
I hope I get to see the Video Poker Player article. Not sure if the
magazine is available here in Idaho?
Anyway, my feeling is the progressive strategy should be learned
based on the ER where you would actually begin to play (say roughly
101% or higher). Maybe you'd want to tweak it a tad as the ER gets
higher. I don't think it ever makes sense to use a strategy based on
the reset ER, which for progressives is usually poor. I realize
there is a "trip" ROR factor to consider at 101%, of course.
Another angle on this is with games having multiple progressives.
Some DDBP games have 5, even 6, progressives. In those situations my
thinking is to decide what ER would convince you to play. Then,
based on the feed rate, and how it's allocated to the multiple
progressives (the allocation scheme), compute the "most likely" level
of each progressive at that EV. Then setup those levels in FVP and
learn the strategyfor that game. That's a compromise because the 5/6
progressives will never exactly agree with the "most likely" level at
once, but it's better than no method at all.
This is a lot of trouble, and the variance may be quite high, so I
wouldn't blame anyone for not wanting to take the time and trouble to
do this. On the other hand, when I come across a multiple
progressive where most of the increased return is due to the quad
5/K, that's worth playing!
--- In vpFREE@yahoogroups.com, Skip Hughes <skiphughes@...> wrote:
Al:
I think it is best to start with the basics. For that, you need two
things...the pay table and the probability that you will actually see
each entry in the table. I see from an earlier post that you use
Frugal VP, so you can get these by looking at the machine statistics.
To get the numbers you are used to, choose "View perfect play
statistics."
First...very basic (to the point of being wrong, but close).
To get the EV, which is also known as the mean or average value, you
simply take the payout for each hand in the paytable times the
probability of getting that hand for a given strategy, and add these
all together. This is usually done "per coin" so a royal is worth
800, assuming that you would actually play 5 coins. For example, if a
royal pays 800, and you see a royal .0025% of the time, it is worth
0.02 coins. This value is listed in the machine statistics under "%
contribution". A pair of Jack or better pays only 1 coin, but is seen
21% of the time, so it is worth .21 coins. Add up all these entries
to get the EV. Notice that the royal contributes very little to the
EV compared to the pair.
To get the variance, do the same thing, but square the payouts first
(this is the part that is wrong, but close).
So the royal contributes 800*800*.000025=16 to the variance, and the
pair contributes 1*1*.21=.21, so almost nothing. So if a progressive
ups the payout for a royal, it can have a huge effect on the variance,
although the effect on the EV could be small.
To make things accurate, (read: this is boring, so read at your own
risk) after you do the calculations above and add everything together,
you need to subtract off EV^2. The reason for this is that variance
is supposed to be a measure of variability, and should not be affected
by changes in the mean. This correction removes the effect of a
shifting mean.
So, now that you know how these things are computed, you can get a
feel for what happens with a progressive. It will have a small affect
on the EV, but a huge effect on the variance.
To answer the rest of your question...what happens if you play a
progressive with basic strategy? Well, we already determined that the
variance will increase. But you can't possibly lose more money than
you would have lost had you played a non-progressive...the only
possible effect is that you might win more if you hit that royal.
So, variance increases, and probably vilatility depending on how you
define it, but your bankroll does not need to be any larger (although
you might want to bring more money anyway so you can play a little
longer in pursuit of that royal).
And you really don't "sacrifice" anything by playing the basic
strategy. The progressive only makes things better for you. However,
there may be a few tweaks you might make to the strategy to increase
the chances of hitting the royal (and therefore raise the EV). These
changes will be rare, and could have an effect on your bankroll, since
you will be trying more often for the royal at the expense of taking
"safer" plays.
- John
Okay as the progressive newbie and mathmatically challenged I'll ask
the following question. What does one sacrifice if one plays a
progressive with the basic 9/6 JOB stategy? Does one retain the 19.51
variance and lower volatility on the bankroll ? What happens to the
expected return? What other sacrifices does one make?
--- In vpFREE@yahoogroups.com, "Al Marek" <amarek1402@...> wrote:
Any help would be benefical in understanding this.
Thanks
Al
Brumar;
VPP is an online publication (an "ezine") so it's available anywhere. It's formatted to be printed out so you can easily print a hard copy if you prefer. I am making a special price available to vpFREE members. I will have a post about this in a few minutes.
Skip
brumar_lv wrote:
I hope I get to see the Video Poker Player article. Not sure if the
magazine is available here in Idaho?
John:
First of all, thanks for the additional info. I believe your last paragraph has finally turned the light on for me.
Al
----- Original Message -----
From: murphyfields
To: vpFREE@yahoogroups.com
Sent: Thursday, July 20, 2006 11:46 AM
Subject: [vpFREE] Re: "progressive newbies question"
Al:
I think it is best to start with the basics. For that, you need two
things...the pay table and the probability that you will actually see
each entry in the table. I see from an earlier post that you use
Frugal VP, so you can get these by looking at the machine statistics.
To get the numbers you are used to, choose "View perfect play
statistics."
First...very basic (to the point of being wrong, but close).
To get the EV, which is also known as the mean or average value, you
simply take the payout for each hand in the paytable times the
probability of getting that hand for a given strategy, and add these
all together. This is usually done "per coin" so a royal is worth
800, assuming that you would actually play 5 coins. For example, if a
royal pays 800, and you see a royal .0025% of the time, it is worth
0.02 coins. This value is listed in the machine statistics under "%
contribution". A pair of Jack or better pays only 1 coin, but is seen
21% of the time, so it is worth .21 coins. Add up all these entries
to get the EV. Notice that the royal contributes very little to the
EV compared to the pair.
To get the variance, do the same thing, but square the payouts first
(this is the part that is wrong, but close).
So the royal contributes 800*800*.000025=16 to the variance, and the
pair contributes 1*1*.21=.21, so almost nothing. So if a progressive
ups the payout for a royal, it can have a huge effect on the variance,
although the effect on the EV could be small.
To make things accurate, (read: this is boring, so read at your own
risk) after you do the calculations above and add everything together,
you need to subtract off EV^2. The reason for this is that variance
is supposed to be a measure of variability, and should not be affected
by changes in the mean. This correction removes the effect of a
shifting mean.
So, now that you know how these things are computed, you can get a
feel for what happens with a progressive. It will have a small affect
on the EV, but a huge effect on the variance.
To answer the rest of your question...what happens if you play a
progressive with basic strategy? Well, we already determined that the
variance will increase. But you can't possibly lose more money than
you would have lost had you played a non-progressive...the only
possible effect is that you might win more if you hit that royal.
So, variance increases, and probably vilatility depending on how you
define it, but your bankroll does not need to be any larger (although
you might want to bring more money anyway so you can play a little
longer in pursuit of that royal).
And you really don't "sacrifice" anything by playing the basic
strategy. The progressive only makes things better for you. However,
there may be a few tweaks you might make to the strategy to increase
the chances of hitting the royal (and therefore raise the EV). These
changes will be rare, and could have an effect on your bankroll, since
you will be trying more often for the royal at the expense of taking
"safer" plays.
- John
--- In vpFREE@yahoogroups.com, "Al Marek" <amarek1402@...> wrote:
>
> Okay as the progressive newbie and mathmatically challenged I'll ask
the following question. What does one sacrifice if one plays a
progressive with the basic 9/6 JOB stategy? Does one retain the 19.51
variance and lower volatility on the bankroll ? What happens to the
expected return? What other sacrifices does one make?
> Any help would be benefical in understanding this.
>
> Thanks
> Al
>
[Non-text portions of this message have been removed]
It seems to me that a coin-in requirement by itself is not sufficient to determine a strategy; at least one additional parameter is needed. For example, are you trying to minimize the time required to hit a coin-in amount? (answer: avoid winning hands!?!) Minimize risk for that coin-in? (answer: regular min-risk strategy) Maximize EV? (answer: regular max-ev strategy). So it seems that coin-in becomes an irrelevant part of the equation. Please correct me if I'm wrong.
Steve Jacobs <jacobs@xmission.com> wrote: Someone here recently asked about getting the best shot at reaching
a fixed coin-in requirement, as their definition of "success". This leads
to a complex strategy that has some interesting characteristics, so I'm
trying to find some time to explore that goal in more detail.
---------------------------------
Yahoo! Music Unlimited - Access over 1 million songs.Try it free.
[Non-text portions of this message have been removed]