what is long term video poker play?

Arcimedes, here's what I don't understand, so please explain it to me: when a VP machine has 99.5% return over the long term, how much "lee way" or a "window" or "variance" should I allow in that figure of 99.5%? In other words, isn't 99.5% really 99.5% ?? Or is 99.4% also considered to be 99.5%?

Don't you reach a number of 99.5% by taking into account ALL of the possible hand results, as Wizard did, in his probability chart using more than 19 trillion possible returns?

So, if the probability chart is based on 19 trillion different possible returns, how can fewer than 19 trillion give you a 100% accurate reflection of that 99.5% game? I think you'll agree you cant.

In business and in economics we do extrapolate numbers. But extrapolations made with variables can be wrong. It seems to me that every hand in VP is another variable. Hence the long term is a minimum of that 19+ trillion hands. You might indeed hit that 99.5% playing fewer, just like in craps I might throw boxcars only one out of 36 times on my first 36 rolls of the dice.
Ok...let's start with something simple. What is the probability of getting 'Heads' when you flip a coin?
Quote

Originally posted by: MoneyLA
Arcimedes, here's what I don't understand, so please explain it to me: when a VP machine has 99.5% return over the long term, how much "lee way" or a "window" or "variance" should I allow in that figure of 99.5%? In other words, isn't 99.5% really 99.5% ?? Or is 99.4% also considered to be 99.5%?


In real play your return will fall into a range of ER +- x, where x is some value which decreases as you play more and more hands. If you played 19 trillion hands then x is still non-zero. The ER doesn't change but anyone's real results only converge on the ER. You are looking for an exact number of hands where x=0. There is none.

Quote

Originally posted by: MoneyLA
Don't you reach a number of 99.5% by taking into account ALL of the possible hand results, as Wizard did, in his probability chart using more than 19 trillion possible returns?


Let's say you are dealt 3 aces. One of the possible draws is the 4th ace and one of the other 46 cards. The results of all these 46 situations is the same. You don't need to "experience" them all and you won't get them all in any kind of succession of the other 46 cards. Maybe you get the 2S 2 times and the 6C never. So, in this case it's obvious you need not take into account all 19 trillion hand results to achieve a games' ER. The same holds true for ALL the other results as well. You only need to experience the results close to their probability of occurring. The more one plays the closer they get.

Quote

Originally posted by: MoneyLA
So, if the probability chart is based on 19 trillion different possible returns, how can fewer than 19 trillion give you a 100% accurate reflection of that 99.5% game? I think you'll agree you cant.


Simple, a typical VP game has between 9-15 results. The key is to experience the results themselves at about the probability of their occurrence to achieve the ER. You don't need to experience all variations of the same result.

Quote

Originally posted by: MoneyLA
In business and in economics we do extrapolate numbers. But extrapolations made with variables can be wrong. It seems to me that every hand in VP is another variable. Hence the long term is a minimum of that 19+ trillion hands. You might indeed hit that 99.5% playing fewer, just like in craps I might throw boxcars only one out of 36 times on my first 36 rolls of the dice.


It's better to look at it this way, take a 4-3 toss vs. a 5-2 toss vs. a 6-1 toss. There are 6 total chances to get that ONE 7 result. The result will be the same if you throw 4-3 4 times and 5-2 2 times and never throw a 6-1. You don't need to experience all 6 types of results to achieve the statistical numbers. There are only 11 results (2-12) that are meaningful out of the total of 36 possible results.
thanks for putting in the effort to explain this. the main thing that I "take away" is this: "You are looking for an exact number of hands where x=0. There is none." So, in reality, it's "99.5% over the long term, maybe" -- is that more accurate?

Quote

Originally posted by: MoneyLA
thanks for putting in the effort to explain this. the main thing that I "take away" is this: "You are looking for an exact number of hands where x=0. There is none." So, in reality, it's "99.5% over the long term, maybe" -- is that more accurate?


Yes, even in the so-called long term it is still "maybe". It could be 99.2% or it could be 99.9%. Remember it is not just minus, it's just as likely to be plus. The results converge on the ER over time. That's why I don't like "long term" as an expression because it gives an impression that it is precise, when it is not. You can only define results in terms of probabilities. That's why I prefer the expression "over time". And, this is not only true of VP, but almost everything else.

When we were conceived the gene expression of our parents also worked on probabilities. If we end up fat or skinny is often decided at our birth. The color of our hair, our height, color of eyes, etc., etc. It's all probabilities.

Finally, when you started a business the probability of being successful was not 100%. However, the longer you were able to maintain the business the better your chances of success got. In addition, the more effort and skill you put into your business the better your chances became. Not at all unlike VP. The more effort I put into the game the better results I achieved as well.

Quote

Originally posted by: MoneyLA
Is it true that to play for the long term you must play thru 19,933,230,517,200 hands?


Long term is a subjective point based on how accurate or confident the end data needs to be as required by the user. All based on statistics which leads into probabilities. Very rarely should the results of statistics be considered absolute and 100% verifiable.

To reach an over 99.999~% confidence, you'd need to play at least 19,933,230,517,200 hands.

However, in my former line of work where I would typically require 100 data points from a specific test 1,993,323,051,720,000 plays would be a better sampling given the 19,933,230,517,200 possible variations. Even after nearly 2 quadrillion plays, you may still not have reached the 100% statistical confidence level.

To answer the original question - It depends.

Suppose you are (God forbid) in the hospital hooked up to an IV line. How do you know that the bag of IV fluid dripping into your vein is sterile? I guarantee you that the manufacturer did not test the bag dripping into you. A few IV from the same lot were tested, passed and then the entire lot was released for use. There is no 100% guarantee the IV is sterile.
Quote

Originally posted by: snidely333How do you know that the bag of IV fluid dripping into your vein is sterile?

Because he paid the Doctor $1245....oh wait a minute.

When will it end ?

When will it end?
Quote

Originally posted by: DonDiego
When will it end ?

When will it end?


Was it over when the Germans bombed Pearl Harbor?
Already a LVA subscriber?
To continue reading, choose an option below:
Diamond Membership
$3 per month
Unlimited access to LVA website
Exclusive subscriber-only content
Limited Member Rewards Online
Join Now
or
Platinum Membership
$50 per year
Unlimited access to LVA website
Exclusive subscriber-only content
Exclusive Member Rewards Book
Join Now