vpFREE2 Forums

Video Poker simulation

#1

I regularly do high speed simulations with WinPoker, usually in sets of a million hands each, different games, and I am not surprised to often see the results being more than a percent off, up or down, from what "it's supposed to be".

Today I did a million hand simulation with TDW, Triple Deuces Wild, a high variance game with an expected return of 99.92% and a variance figure of over 96. I've worked with this game before, but THIS time I was surprised. A million hands, returned 96.52%, and had lost over 160,000 coins.

A full-time player can probably play a million hands in about 7 months. How would you like to be playing TDW for quarters, and be down $40,000 in 7 months? I think many who read this board would be talking to themselves. (myself included) Some might even start believing that the machines aren't fair. But alas, it's not the machines, it's just that in video poker, stuff happens. The "long run" can be longer than you might think.

#2

It certainly is one explanation of why I had a day last weekend with no hand better than a full house.

But you can't fault those who believe machines aren't fair, either. These are intense economic times, and the temptation for casinos to do a little "tweaking" must be intense. Profit margins are scruntinized carefully by shareholders.

As in any unfair enterprise, many businesses might be avoiding unscrupulous acts out of fear of discovery, not because the acts themselves are wrong. Estimates of how many will depend not only on how cynical you are but the way you interpret your own life experiences as well.

We can never be 100% certain we're getting a fair deal on any VP machine. But, for me, repeated unusually negative results signify it's time to move on. As you say, it can get much worse. So, I'll still "chase my money" but do so elsewhere and accept I won't recoup it where it was lost, fairly OR unfairly.

···

--- In vpFREE@yahoogroups.com, "bobbartop" <bobbartop@...> wrote:

I regularly do high speed simulations with WinPoker, usually in sets of a million hands each, different games, and I am not surprised to often see the results being more than a percent off, up or down, from what "it's supposed to be".

Today I did a million hand simulation with TDW, Triple Deuces Wild, a high variance game with an expected return of 99.92% and a variance figure of over 96. I've worked with this game before, but THIS time I was surprised. A million hands, returned 96.52%, and had lost over 160,000 coins.

A full-time player can probably play a million hands in about 7 months. How would you like to be playing TDW for quarters, and be down $40,000 in 7 months? I think many who read this board would be talking to themselves. (myself included) Some might even start believing that the machines aren't fair. But alas, it's not the machines, it's just that in video poker, stuff happens. The "long run" can be longer than you might think.

#3

<<Today I did a million hand simulation with TDW, Triple Deuces Wild, a high variance game with an expected return of 99.92% and a variance figure of over 96. I've worked with this game before, but THIS time I was surprised. A million hands, returned 96.52%, and had lost over 160,000 coins.>>

This wouldn't surprise me at all - plus it's a negative game and the longer you play the worse it can get!!!!! I think too many people are doing negative plays and think they can still win - "almost positive" just doesn't get it. And even if you have just a .1 or .2 advantage, the long term is SOOOO long and the bankroll requirement is unbelievably huge.

We have a large bankroll and don't usually play unless we have at least a .5% advantage and we really look for more than that. Most wise gamblers with a smaller bankroll will need at least 1%.

···

________________
Jean $�ott, Frugal Gambler
http://queenofcomps.com/
You can read my blog at
http://jscott.lvablog.com/

#4

I had a very interesting experience on my last trip. I was playing late at night on a middle machine in a bank of slant tops, and a casino staffer started reviewing the payout percentages of all the other machines in the bank. These were IGT multi-game (slots, BJ & VP) machines, denominated 20 cents and $1. Since I was in the middle of a semi-ATM experience, there was no way she was going to get into my machine, but I could see the numbers on the machines on either side as she cycled thru the games.

The slot numbers I saw, particularly in the lower denom were in the 80%'s, i.e. customer losses in the teens. VP, except for Aces & Eights were in the high 90's (96-98+). A&8's was ~91.5%. A&8 has a pretty lousy paytable, but has something like a 50K payoff for a sequential RF. What I saw was pretty much in line with what you'd expect at Mohegan, other than on the FP PE machines.

···

________________________________

We can never be 100% certain we're getting a fair deal on any VP machine.

_,_._,___

[Non-text portions of this message have been removed]

#5

Yeah, triple deuces wild is negative, but throw in a free shrimp cocktail and its even. The appeal is the way the casino ranks the games. If they are offering .3 on TDW, and only .1 on 10/7 DB, well the TDW is the much better play. You'd be surprised at how many casino slot manager actually know via computer printout the EV of all their inventory and assign slot points accordingly.
I wrote an article a couple of years ago on TDW for VPP , and its not for the faint of heart. But... I remember that one nice day , playing nothing but quarters at Palace Station , I ran up a $200 stake to $4000 - and no tax forms or royals for playing it ! At the Vegas club a couple of years back , when they still had $1 full pay deuces I was stuck $3000, with only $100 left in my wallet. Put it in the $ TDW machine and down to only $20 hit the ducks and it was back to EVEN. WOW - what a feeling!
best wishes...Tom

···

----- Original Message ----- From: "Jean Scott" <queenofcomps@cox.net>
To: <vpFREE@yahoogroups.com>
Sent: Wednesday, February 03, 2010 2:09 PM
Subject: Re: [vpFREE] Video Poker simulation

<<Today I did a million hand simulation with TDW, Triple Deuces Wild, a high
variance game with an expected return of 99.92% and a variance figure of
over 96. I've worked with this game before, but THIS time I was surprised.
A million hands, returned 96.52%, and had lost over 160,000 coins.>>

This wouldn't surprise me at all - plus it's a negative game and the longer
you play the worse it can get!!!!! I think too many people are doing
negative plays and think they can still win - "almost positive" just doesn't
get it. And even if you have just a .1 or .2 advantage, the long term is
SOOOO long and the bankroll requirement is unbelievably huge.

#6

Only if you have money to burn.

Approximate Kelly bankroll for 10/7 DB +.1 is about 27/.25% = 10,800 bets.

Approximate Kelly bankroll for TDW +.3 is about 97/.3% = 32,333 bets.

···

--- In vpFREE@yahoogroups.com, "tomflush" <tomflush@...> wrote:

If they are offering .3 on TDW, and only .1 on 10/7 DB, well the TDW > is the much better play.

#7

A million hands commonly gets thrown about as if it's some magic number. Well, it's not. The magic number is N0, and it varies depending on the game, since its formula is variance/edge^2 hands. For your numbers, you're looking at an N0 of 96/(-.08%)^2 = 150 MILLION HANDS! (meaning about a 68% chance of getting results between even and double the average. At 4 times N0, you get about a 95% chance of results between even and double the average).

http://members.cox.net/vpfree/Bank_NO.htm

···

--- In vpFREE@yahoogroups.com, "bobbartop" <bobbartop@...> wrote:

I regularly do high speed simulations with WinPoker, usually in sets of a million hands each, different games, and I am not surprised to often see the results being more than a percent off, up or down, from what "it's supposed to be".

Today I did a million hand simulation with TDW, Triple Deuces Wild, a high variance game with an expected return of 99.92% and a variance figure of over 96. I've worked with this game before, but THIS time I was surprised. A million hands, returned 96.52%, and had lost over 160,000 coins.

A full-time player can probably play a million hands in about 7 months. How would you like to be playing TDW for quarters, and be down $40,000 in 7 months? I think many who read this board would be talking to themselves. (myself included) Some might even start believing that the machines aren't fair. But alas, it's not the machines, it's just that in video poker, stuff happens. The "long run" can be longer than you might think.

#8

Thank you, Iguana, and I get your drift. I was still surprised by this particular sim, because I have done it before, and I cannot recall a million-hand sim being "off" by 3%. That's a pretty "bad run". To be specific, there were 18 royals, and 180 quad deuces. I think any time you only have 18 royals in a million-hand sample, you're going to be "stuck", no matter what game it is or no matter what the variance number. But this game's pay table is so top-heavy, wow, it just surprised me to see such a bad run even though I should know better.

Hey, at least it was just a simulation and not a real-life experience, I'd be talking to myself. (actually, I do that already)

Bottom line, if there's a message in my ramblings, is that this is a good game to play, if one has the bankroll, and one has the stomach for it. The game is usually not penalized and usually gets full points, full comps, and is pretty easy to learn to play perfectly.

Thanks for your comments about NO, Iguana.

-BB

···

--- In vpFREE@yahoogroups.com, "nightoftheiguana2000" <nightoftheiguana2000@...> wrote:

A million hands commonly gets thrown about as if it's some magic number. Well, it's not. The magic number is N0, and it varies depending on the game, since its formula is variance/edge^2 hands. For your numbers, you're looking at an N0 of 96/(-.08%)^2 = 150 MILLION HANDS! (meaning about a 68% chance of getting results between even and double the average. At 4 times N0, you get about a 95% chance of results between even and double the average).

http://members.cox.net/vpfree/Bank_NO.htm

#9

Huh? The expectation is for about 22 royals in a million hands. One standard deviation would be sqrt(22) = about 4.7 . So, about 68% of the time you expect results of between about 17 and 27 royals, and about 95% of the time you expect results of between 13 and 31 royals. The expectation for ducks is about 224, sqrt(224) = about 15. So, about 68% of the time you expect results of between about 209 and 239, and about 95% of the time you expect results of between 194 and 254. So, your ducks result is an outlier. About 99.7% of the time you expect ducks of between 179 and 269, so your duck results are something like a 0.15% outlier (>-3SD). You notice that in TDW because the deuces win amount is 3x the amount you would win in regular deuces or 24 times the amount you would win in JOB. Put another way, a deficit of 44 ducks costs you 132,000 credits in TDW, 44,000 credits in FPDW, and 5,500 credits in JOB.

As a side note, the N0 for FPDW is about a half million hands, so 4N0 is about two million hands. But 4N0 means you expect that 95% of the time your results will be between breakeven and double the average. The notion that your results will be near the ER is largely dreaming, unless your notion of "near" is breakeven to double the average (in this case about 100% to 101.5% return). And you've still got 5% outliers. JOB is similar, the N0 is 800,000 hands, and since it's a negative expectation game, you expect results (68% of the time) of between about 100% and 99% return. Bonus Poker has an N0 of about 300,000 hands. DDB is about 420,000 .

···

--- In vpFREE@yahoogroups.com, "bobbartop" <bobbartop@...> wrote:

To be specific, there were 18 royals, and 180 quad deuces. I think any time you only have 18 royals in a million-hand sample, you're going to be "stuck", no matter what game it is or no matter what the variance number.

#10

Perhaps I wasn't clear, I do know that it was the ducks that were way off in that simulation, not the royals. But still, I stand by what I think I said, that in a million-hand sim 18 royals usually means you're going to be "stuck".

···

--- In vpFREE@yahoogroups.com, "nightoftheiguana2000" <nightoftheiguana2000@...> wrote:

Huh? The expectation is for about 22 royals in a million hands. One standard deviation would be sqrt(22) = about 4.7 . So, about 68% of the time you expect results of between about 17 and 27 royals, and about 95% of the time you expect results of between 13 and 31 royals. The expectation for ducks is about 224, sqrt(224) = about 15. So, about 68% of the time you expect results of between about 209 and 239, and about 95% of the time you expect results of between 194 and 254. So, your ducks result is an outlier.

#11

standard deviation is the square root of variance. 96 var, so stddev
is a bit under 10.

over a larger sample size, multiply stddev per hand by the square root
of the sample size for the stddev over the sample. square root of a
million is 1000, so your standard deviation on the sample is 10000
bets, or $12500.

being down $40000 on an almost-breakeven game is a little over three
standard deviations below expectation. running that bad happens about
one time in a thousand. that's pretty bad, but it happens.

http://en.wikipedia.org/wiki/Standard_deviation#Rules_for_normally_distributed_data

96 var is just huge. the variance of TDW at quarters is like the
variance of NSUD at dollars. the denomination is deceiving.

cheers,

five

···

On Wed, Feb 3, 2010 at 9:37 PM, bobbartop <bobbartop@yahoo.com> wrote:

Thank you, Iguana, and I get your drift. I was still surprised by this particular sim, because I have done it before, and I cannot recall a million-hand sim being "off" by 3%. That's a pretty "bad run".

#12

Depends on the game. A million hands of FPDW puts you ahead on average about 7500 bets or over 9 royals. So, in that game, a 4 royal deficit alone is not enough to make you "stuck".

···

--- In vpFREE@yahoogroups.com, "bobbartop" <bobbartop@...> wrote:

Perhaps I wasn't clear, I do know that it was the ducks that were way off in that simulation, not the royals. But still, I stand by what I think I said, that in a million-hand sim 18 royals usually means you're going to be "stuck".

#13

Of course, you are right, Iguana, and I don't want you to think I am arguing. For instance, I just now did a 1-mil sim of fpdw, and got 16 royals. And, it was 100.12, not 'stuck'. So maybe what I should have said is that if one only gets 18 royals, it is USUALLY well behind the expected. That's why I usually say 'usually', so I don't get backed into a corner and have to eat my words. I usually say 'usually', but sometimes I forget.

···

--- In vpFREE@yahoogroups.com, "nightoftheiguana2000" <nightoftheiguana2000@...> wrote:

Depends on the game. A million hands of FPDW puts you ahead on average about 7500 bets or over 9 royals. So, in that game, a 4 royal deficit alone is not enough to make you "stuck".

#14

Does video poker have a normal distribution? I thought not.

Cogno

···

-----Original Message-----
From: vpFREE@yahoogroups.com [mailto:vpF…@…com] On Behalf
Of fivespot
Sent: Friday, February 5, 2010 1:08 AM
To: vpFREE@yahoogroups.com
Subject: Re: [vpFREE] Re: Video Poker simulation

On Wed, Feb 3, 2010 at 9:37 PM, bobbartop <bobbartop@yahoo.com> wrote:
> Thank you, Iguana, and I get your drift. I was still surprised by
this particular sim, because I have done it before, and I cannot recall
a million-hand sim being "off" by 3%. That's a pretty "bad run".

standard deviation is the square root of variance. 96 var, so stddev
is a bit under 10.

over a larger sample size, multiply stddev per hand by the square root
of the sample size for the stddev over the sample. square root of a
million is 1000, so your standard deviation on the sample is 10000
bets, or $12500.

being down $40000 on an almost-breakeven game is a little over three
standard deviations below expectation. running that bad happens about
one time in a thousand. that's pretty bad, but it happens.

http://en.wikipedia.org/wiki/Standard_deviation#Rules_for_normally_dist
ributed_data

96 var is just huge. the variance of TDW at quarters is like the
variance of NSUD at dollars. the denomination is deceiving.

cheers,

five

------------------------------------

vpFREE Links: http://members.cox.net/vpfree/Links.htm

Yahoo! Groups Links

#15

...........................................................................

Can simulations be done in Video Poker for Winners? I do not see it in any menu. I see bankroll, but not sim. Am I overlooking it?

···

--- In vpFREE@yahoogroups.com, "bobbartop" <bobbartop@...> wrote:

I regularly do high speed simulations with WinPoker, usually in sets of a million hands each, different games, and I am not surprised to often see the results being more than a percent off, up or down, from what "it's supposed to be".

#16

>
> I regularly do high speed simulations with WinPoker, usually in sets of a million hands each, different games, and I am not surprised to often see the results being more than a percent off, up or down, from what "it's supposed to be".

...........................................................................

Can simulations be done in Video Poker for Winners? I do not see it in any menu. I see bankroll, but not sim. Am I overlooking it?

Yes, I meant Video Poker for winners, not WinPoker

···

--- In vpFREE@yahoogroups.com, "deuceswild1000" <deuceswild1000@...> wrote:

--- In vpFREE@yahoogroups.com, "bobbartop" <bobbartop@> wrote:

#17

There's an auto-play under 'options', but it's not very fast. Unless I am missing something, in which case I hope someone tells me otherwise. This is the main reason I still keep a shortcut to WinPoker on my desktop.

···

--- In vpFREE@yahoogroups.com, "deuceswild1000" <deuceswild1000@...> wrote:

Yes, I meant Video Poker for winners, not WinPoker
>