Video Poker Trucks Run Over Everybody

[Alan Mendelson]

ATTENTION ALL VIDEO POKER PLAYERS. ATTENTION ALL VIDEO POKER AUTHORS. ARCIMEDE$ HAS JUST CHANGED THE ENTIRE VIDEO POKER INDUSTRY:

We know VP is random.

Did you read that: Video poker is random, he says. Forget about your ability to hold cards. Forget about proper strategy. Arcimede$ says "VP is random." Throw out your software, recycle your books, stop writing your columns. Video poker is random -- random like tossing coins and throwing dice.

Game over.

[bigfoot66]

Well the actual results of video poker players do not need to be surveyed. That the results will form a bell curve around the expected return is a theoretical device that teaches us what to expect in the future of our video poker playing or other gambling. It is much less useful in explaining the past, though it does describe the past.

[arcimede$]

ATTENTION ALL VIDEO POKER PLAYERS. ATTENTION ALL VIDEO POKER AUTHORS. ARCIMEDE$ HAS JUST CHANGED THE ENTIRE VIDEO POKER INDUSTRY:

Did you read that: Video poker is random, he says. Forget about your ability to hold cards. Forget about proper strategy. Arcimede$ says "VP is random." Throw out your software, recycle your books, stop writing your columns. Video poker is random -- random like tossing coins and throwing dice.

Game over.

I hardly changed the industry. Those in the industry already know VP is random. Alan, here's a little advice. When you're in a deep hole, stop digging.

[Alan Mendelson]

No Arc, video poker is not a random game as long as you are able to choose the cards you hold. If video poker were a random game, there would not be a skill element to it. You of all people talk about the skill of playing video poker.

[arcimede$]

No Arc, video poker is not a random game as long as you are able to choose the cards you hold. If video poker were a random game, there would not be a skill element to it. You of all people talk about the skill of playing video poker.

Alan, you're missing the point. Parts of the game are completely random like the deal/draw. There is skill involved, however when you look at groups of players over a large number of hands the skill factor simply is one random feature of the overall sample of players. If you want to look at all players then you will have a skewed bell curve but it is still shaped like a bell. It all gets down to the assumptions and what information you are trying to pull out of the data.

[Alan Mendelson]

Alan, you're missing the point. Parts of the game are completely random like the deal/draw. There is skill involved, however when you look at groups of players over a large number of hands the skill factor simply is one random feature of the overall sample of players. If you want to look at all players then you will have a skewed bell curve but it is still shaped like a bell. It all gets down to the assumptions and what information you are trying to pull out of the data.

Yes, Arc, I agree with you when you say "It all gets down to the assumptions and what information you are trying to pull out of the data." Because as we all know, there are lies, damned lies and statistics.

I think that wraps up this discussion about bell curves.

[arcimede$]

Yes, Arc, I agree with you when you say "It all gets down to the assumptions and what information you are trying to pull out of the data." Because as we all know, there are lies, damned lies and statistics.

I think that wraps up this discussion about bell curves.

As long as you accept that a bell curve is a useful tool in analyzing VP, then I agree. However, I suspect you're still very confused and probably don't understand the fundamental mathematics involved.

[Alan Mendelson]

Arc, I have 30 hours in Economics from the Maxwell School at Syracuse. I only got a "minor" in Economics and not a dual major because I only took one semester of psychology and did not fulfill the "science" requirement for the dual major. Trust me, I know plenty about bell curves.

Arc, are bell curves useful tools? Yes they are. But we plot bell curves based on research. We don't automatically apply bell curves to different data and assume that the different data is appropriate for the bell curve we use. To put it simply: we don't take a bell curve for apples and apply it to oranges.

Now, you're a math guy, and Im an economist. You think bell curves are universal, and I'm saying they're not. And if you want to continue arguing this you can, but to me the argument is similar to "trial and error" vs "the scientific method."

[arcimede$]

Arc, I have 30 hours in Economics from the Maxwell School at Syracuse. I only got a "minor" in Economics and not a dual major because I only took one semester of psychology and did not fulfill the "science" requirement for the dual major. Trust me, I know plenty about bell curves.

So? It's not what you learned many decades ago that's relevant. It's what you say today. So far you show little understanding.

Arc, are bell curves useful tools? Yes they are. But we plot bell curves based on research. We don't automatically apply bell curves to different data and assume that the different data is appropriate for the bell curve we use. To put it simply: we don't take a bell curve for apples and apply it to oranges.

Alan, right away you show you don't understand the topic all. You don't "apply" a bell curve. The bell curve is simply the result of charting data. However, we know up front when we will see a bell curve because of the type of data being charted.

Now, you're a math guy, and Im an economist. You think bell curves are universal, and I'm saying they're not. And if you want to continue arguing this you can, but to me the argument is similar to "trial and error" vs "the scientific method."

No one said anything about bell curves being universal. They apply primarily to random data with a normal distribution.

[Alan Mendelson]

But Arc there is no data to chart. Get your data, chart it, and then we'll talk.

[arcimede$]

But Arc there is no data to chart. Get your data, chart it, and then we'll talk.

Alan, we can chart data because we can generate data based on what we know about VP games. Good grief man, it's what's known as a tool. You use it to understand. But first you have to spend a little time trying to understand instead of spewing verbal garbage all over the place.

You do realize that VP is a "solved game" ... right? We can examine every aspect of the game including all possible deals and draws. With that level of data we can then produce graphs that just happen to take the form of a bell. We don't need to collect any data to understand a game. Now, how you apply that knowledge takes some care. It is possible to make invalid claims if one is not careful to understand what the data means. The point is your continual rants do not apply.

[Alan Mendelson]

Perhaps I am not making myself clear. The original issue, as I remember it to be, was whether or not the results of the universe of video poker players on their play matches a bell curve. This "number" is the "end number" at the end of a trip, or a session, or a year or a lifetime of play. It is not a "number" representing the results of any deal, or draw, or hand.

Now, when that number is plotted on a chart, the question is -- does it represent a bell curve?

My understanding is that 95% of casino visitors lose money. I don't know what percentage of players lose it all when they visit, do you? I don't know what percentage of players leave with a win or how much of a win, do you?

So if we don't know what percentage of players lose it all when they visit, or over a year or a lifetime, and we don't know how much the winners win or what percentage of their bankroll that winning amount represents, how can you plot anything let alone a chart that resembles a bell curve?

Now, I'm not saying a bell curve would not exist -- I'm saying there is no date that shows a bell curve.

A bell curve is based on random events -- but (and you acknowledge this) video poker is not a 100% random game like throwing dice or flipping coins even though video poker has some random elements to it. So because video poker is not just a random game it would not be appropriate to apply without appropriate research a bell curve to the results of video poker players.

We also discussed live poker tournaments in which 90% of players receive nothing for their entry fee, and all of the payments go to 10% of the players with the final few players getting the big percentage of the payoffs. And clearly there is no bell curve in poker tournaments. Poker tournaments are just one example where "bell curves" do not exist.

You will have to prove to me that bell curves exist in the results of video poker play. If you have the data, plot it and show it.

That sums up my position. Tell me where I am wrong. I'm listening.

I think part of the problem here is that you are looking for a bell curve in terms of the universe of "hands" and I am looking for the existence of a bell curve for the "cash outs" of players. I really think you and I have been debating two different things.

[Vegas Vic]

You will have to prove to me that bell curves exist in the results of video poker play. If you have the data, plot it and show it..

I haven't been involved in this particular discussion because Arci continuously channels his inner Rob Singer with one insult after another, which I find quite distracting, but it appears that Alan has summarized it with this request. My only thought is this: data (real or fabricated) can be plotted to meet any objective. Can he post something? Well, that's up to him. Does it prove anything? That's for the reader to decide.

[Alan Mendelson]

Here's an exercise for those of you interested in bell curves, and plotting data: take a look at the results for this one tournament at the WSOP. Look at the number of entries, look at how many places were paid, and look at the slope of the payoffs leading to the payoff for the champion of the tournament. Then, tell me where there is a bell curve? And remember, all those entrants who didn't get a payoff got a return of zero.

Here are the results: http://www.wsop.com/tournaments/payo...=887&tid=12138

[Count Room]

The structure of having top-heavy prizes in a tournament (only the top 10% cashes in) is precisely the structure that draws the most entrants (and more money for the casino). Many moths are drawn to a flame, most get burnt to a crisp, and a precious few bask in the warmth and glory?

[Alan Mendelson]

The structure of having top-heavy prizes in a tournament (only the top 10% cashes in) is precisely the structure that draws the most entrants (and more money for the casino). Many moths are drawn to a flame, most get burnt to a crisp, and a precious few bask in the warmth and glory?

A very good observation, Count Room. At most of the California casinos, when cash games are "advertised" in the casino, the announcer will almost always comment about how large the "bad beat jackpot" is. Players will gladly sit at a table losing blinds just for the slim chance that a jackpot will be hit at their table and the "table share" could be worth ten times their buy-in, or a hundred times their buy-in if they happen to be the losing hand in the bad beat.

In fact another "lure" for tournament players is the "guarantee" which represents the seed money that the casino puts up. A smart tournament player will seek out the tournaments where the field is less than the guarantee which can give a good player a significant edge on the time spent in the tourney.

That happened to me once, but by accident because I was unaware that the guarantee was larger than justified by the field. It was at The Commerce, and I finished at the final table. The casino needed about eight tables to meet the guarantee, and fewer than five full tables played. While the number of players paid was reduced, it meant more money for each of the pays.

[arcimede$]

Perhaps I am not making myself clear. The original issue, as I remember it to be, was whether or not the results of the universe of video poker players on their play matches a bell curve. This "number" is the "end number" at the end of a trip, or a session, or a year or a lifetime of play. It is not a "number" representing the results of any deal, or draw, or hand.

Now, when that number is plotted on a chart, the question is -- does it represent a bell curve?

No, it will not match a bell curve because none of that data is normalized. What would match a bell curve is the % return of the universe of players. It still wouldn't be very useful because there's still too much variability in the data. Why would anyone want to compare penny and nickel players with APers?

My understanding is that 95% of casino visitors lose money. I don't know what percentage of players lose it all when they visit, do you? I don't know what percentage of players leave with a win or how much of a win, do you?

So if we don't know what percentage of players lose it all when they visit, or over a year or a lifetime, and we don't know how much the winners win or what percentage of their bankroll that winning amount represents, how can you plot anything let alone a chart that resembles a bell curve?

Now, I'm not saying a bell curve would not exist -- I'm saying there is no date that shows a bell curve.

That's because this has nothing to do with random events. You need to understand what you are doing. It would help if you made even a small effort to understand the topic you re discussing.

A bell curve is based on random events -- but (and you acknowledge this) video poker is not a 100% random game like throwing dice or flipping coins even though video poker has some random elements to it. So because video poker is not just a random game it would not be appropriate to apply without appropriate research a bell curve to the results of video poker players.

No one claims it should be applied to results directly. That would make no sense. And, I already have stated this so I don't have any idea why you keep bringing it up.

We also discussed live poker tournaments in which 90% of players receive nothing for their entry fee, and all of the payments go to 10% of the players with the final few players getting the big percentage of the payoffs. And clearly there is no bell curve in poker tournaments. Poker tournaments are just one example where "bell curves" do not exist.

Bell curves would exist. They just don't apply directly to results.

You will have to prove to me that bell curves exist in the results of video poker play. If you have the data, plot it and show it.

That sums up my position. Tell me where I am wrong. I'm listening.

I think part of the problem here is that you are looking for a bell curve in terms of the universe of "hands" and I am looking for the existence of a bell curve for the "cash outs" of players. I really think you and I have been debating two different things.

Well, it would help if you actually read and tried to understand what I have written before going off half-cocked.

[arcimede$]

Here's an example that might help. Let's say someone held a VP tournament. It lasted for a couple of weeks and every entrant got to play 100K hands of FPDW. At the end of the tournament the players that won the most money were paid just like your live poker tournament players. So, you do a couple of graphs. In one case you graph the money won in the tournament. Would it look like a bell curve? Nope not even close.

Now, you plot the actual % return of the player's. OMG, now you have a bell curve. A few players will lose money. Most will win and a few will do extremely well. The vast majority will sit in the middle (close to the ER) thus forming the high center of the bell.

You see, the return of the players is based on random events. The payoff was a one time, predetermined amount and had nothing to do with randomness in it's creation.

[Alan Mendelson]

Now you're saying that the bell curve represents the "percentage return" of video poker play. Wow. You really can pick from among the universe of possible answers to try to justify your claims.

Even if you think a bell curve represents the "percentage return" I'd still like to see your data, because true research is based on data. And the fact is there is no research, and for all we know the "curve" looks like a "W" or a "/" or "______/" or "\____/".

It was a nice discussion. At least you didn't call me an addict and I respect your knowledge.

EDITED TO ADD: I just thought of a way to test your theory. There is a casino here in Southern Cal that runs weekly tournaments... usually slots but sometimes video poker. I could ask them for their print out of results. I think they have 200 players per tournament.

But then I thought about this: when VP machines (like slot machines) are switched to tournament function no machine goes negative. You start with zero points and you only add credits as you hit wins. So in tournament function there would not be the kind of curve you expect. The curve would more likely show a slowly rising base with the winners showing some "surge in points" like this: _____________/

I've played in a couple of video poker and slot tournaments and that is pretty much what the final point tally showed. Everyone was about the same and then the few "winners" had the big points because they hit the royals.

[arcimede$]

Now you're saying that the bell curve represents the "percentage return" of video poker play. Wow. You really can pick from among the universe of possible answers to try to justify your claims.

Wrong again. A bell curve represents a sampling of random values. Since the % return represents a sampling of random values it should show up as a bell curve. Remember, the bell curve is just a graphical view of the data. It's the data itself that is important.

Even if you think a bell curve represents the "percentage return" I'd still like to see your data, because true research is based on data. And the fact is there is no research, and for all we know the "curve" looks like a "W" or a "/" or "______/" or "\____/".

At low numbers of random events the shape will not be a bell. It becomes a bell as the number of samples increases.

It was a nice discussion. At least you didn't call me an addict and I respect your knowledge.

Singer has to believe everyone is an addict or he has to admit he has a weakness. He is unable to do the latter, that makes every other VP player an addict in his eyes.

EDITED TO ADD: I just thought of a way to test your theory. There is a casino here in Southern Cal that runs weekly tournaments... usually slots but sometimes video poker. I could ask them for their print out of results. I think they have 200 players per tournament.

But then I thought about this: when VP machines (like slot machines) are switched to tournament function no machine goes negative. You start with zero points and you only add credits as you hit wins. So in tournament function there would not be the kind of curve you expect. The curve would more likely show a slowly rising base with the winners showing some "surge in points" like this: _____________/

I've played in a couple of video poker and slot tournaments and that is pretty much what the final point tally showed. Everyone was about the same and then the few "winners" had the big points because they hit the royals.

It would form a bell if the y-axis was the number of players and the x-axis any given credit range. There would be a few players way up there in credits and a few quite low. The rest would be bunched in the middle. It probably wouldn't be a perfect bell as there would not be enough hands played, but it would be somewhat bell shaped.