Higher Free Odds to lower variance? (moved from Craps)

First post here.

I am a PhD student in a quantitative discipline, and I am interested in studying the quantitative properties of craps.... everything from the statistics to the physics. I play the game for fun (in Shreveport where I can get 100x odds on a $5 min table).

I am interested in collaborating with someone (Ahigh, are you out there?) to use my money and real-casino rolls in conjunction with software to analyze the optimal sets, etc. for my throw. I do not claim I can "influence" the dice, but after ~75 hours of live craps play on my own throw, I have reason to suspect a bias. Short-term variance perhaps, but I am willing to pay to keep learning.

The unique aspect of where I play is that the odds structure allows for a miniscule house edge. I don't have the bankroll to play $500 odds each roll, but I typically play around 20x odds, which, with the appropriate bankroll, can lower overall variance.

Anyway, as I'm already rambling, I just want to reach out and see if anyone might be interested in partnering up.

Quote: DicePhD

First post here.
<snip>
to analyze the optimal sets, etc. for my throw. <snip>

nice
you really need to post in a different forum

Please have this moved to the Dice Setting forum where it belongs

"If ye have faith as a grain of mustard seed, nothing shall be impossible unto you."

Thank you and have more fun
Sally

Quote: DicePhD

20x odds, which, with the appropriate bankroll, can lower overall variance.

The odds bet increases variance.

Mustang: Maybe you can clarify this for me, since no one in the past has. It involves "variance," what is meant by it. I see the term used in contexts I don't think it applies. I see "variance" as referring to how much a dice roll "varies" from the expected outcomes of the so-called pyramid of possibilities.

High variance would mean a series of dice rolls depart more from the pyramid predictions, resulting in fewer 7 outs and more points hitting. Low variance would mean they conform more to pyramid predictions, resulting in more 7 outs and fewer points hitting. This is my definition, the mathematical definition of variance. Please don't get into EVs, standard deviations, etc. I am math challenged. ; )

The OP, as many others, talks about lowering "variance" by increasing odds. Doing so doesn't apply to the "variance" as I understand it. So, what does the OP, et al., mean by increasing odds lowers "variance"? The only "variance" I can discern is that it lowers (varies from) the HE.

What do you understand about variance, not only mathematically, but how it's used in posts that seem to use it outside of the probabilities of certain numbers showing according to the pyramid?

Quote: DicePhD

but I typically play around 20x odds, which, with the appropriate bankroll, can lower overall variance.

Ok
prove your statement

Quote: RS

The odds bet increases variance.

Ok
prove your statement

Quote: Sonny44

What do you understand about variance, not only mathematically, but how it's used in posts that seem to use it outside of the probabilities of certain numbers showing according to the pyramid?

I have no opinion about variance other than it is just another word.

now, that is the new angle
Post a craps question and have it centered around Dice setting

I find that rude and offensive and did not want to read about dice setting stuffs
like opening up a video where the link posted by a super super ++ ploppie
contained massive amounts of totally offensive language WITHOUT any notice that existed in the damn video.

DIs have their own universe in my opinion

and here their
own BEAUTIFUL forum for discussion

have more fun

are you still offended?
Sally

Dicephd




If you are going to start with say a $5 pass line bet with 20 odds that is a starting position of $105
with no other bets on the table. You would need at least a gamblng account of $10,000- $11,000 just
to start. Now if you start with 20 odds and placed two come bets with similar odds you would need to
have $30,000 or more.

If you want to put as much as you can in your favor, start with a good class, one that will teach not
only a way to bet affectively, money management, but also throwing work if you have the desire to
take a look in that direction. Ahigh is not the place to start.

IN terms of the desire to go over the math in 17 different directions.. that is a waste of time. The
only math of the game of craps that means anything is were you on the numbers that were being
rolled.

If you started with $300, and cashed out $400, now see thats math.

Dicesetter

Quote: DicePhD

First post here.

I am a PhD student in a quantitative discipline, and I am interested in studying the quantitative properties of craps.... everything from the statistics to the physics. I play the game for fun (in Shreveport where I can get 100x odds on a $5 min table).

I am interested in collaborating with someone (Ahigh, are you out there?) to use my money and real-casino rolls in conjunction with software to analyze the optimal sets, etc. for my throw. I do not claim I can "influence" the dice, but after ~75 hours of live craps play on my own throw, I have reason to suspect a bias. Short-term variance perhaps, but I am willing to pay to keep learning.

The unique aspect of where I play is that the odds structure allows for a miniscule house edge. I don't have the bankroll to play $500 odds each roll, but I typically play around 20x odds, which, with the appropriate bankroll, can lower overall variance.

Anyway, as I'm already rambling, I just want to reach out and see if anyone might be interested in partnering up.

LOL Swell, Max odds no doubt my best bet on this situation. Easy come easy go. Not enough BR? I just happen to know someone with an extra $2500. He to is an Ahigh fan. Boy oh boy what a wonderful coinkydink, unfortunately he just woke up and found out he got suspended and you showed up just in time, I'm sure you will fill the gap quit nicely. You have similar interests. Welcome nice to have you, we need more craps guys like you.

I to, have reason to suspect a bias in lots of things. PhD student in quantitative discipline writing, how perfect, your teacher would call you out for misspelling minuscule. However, I wouldn't know.

Quote: DicePhD

First post here.

I am a PhD student in a quantitative discipline, and I am interested in studying the quantitative properties of craps.... everything from the statistics to the physics. I play the game for fun (in Shreveport where I can get 100x odds on a $5 min table).

I am interested in collaborating with someone (Ahigh, are you out there?) to use my money and real-casino rolls in conjunction with software to analyze the optimal sets, etc. for my throw. I do not claim I can "influence" the dice, but after ~75 hours of live craps play on my own throw, I have reason to suspect a bias. Short-term variance perhaps, but I am willing to pay to keep learning.

The unique aspect of where I play is that the odds structure allows for a miniscule house edge. I don't have the bankroll to play $500 odds each roll, but I typically play around 20x odds, which, with the appropriate bankroll, can lower overall variance.

Anyway, as I'm already rambling, I just want to reach out and see if anyone might be interested in partnering up.

Do you by chance know a guy who knows a girl with the golden touch?

Quote: rainman

Do you by chance know a guy who knows a girl with the golden touch?

Don't be ridiculous.

Quote: mustangsally

I have no opinion about variance other than it is just another word. Sally

Disappointed in you, Mustang.

Ok. Well, I had forgotten just how inviting the anonymity of the internet is.

I read through about 20 pages of someone with the golden touch girl and now I understand most of the inside jokes. Timing coincidences notwithstanding, perhaps I can establish myself as a different person, though I'm sure there is no satisfactory way of doing so to appease the masses. If there is, just let me know.

Now, to try and address some of the issues from my post:

1- I am not a dice influencer, nor do I really believe that DI is a real thing. I don't set the dice. I don't think that a new stick man means 7-out. Things don't "seem" to happen "always" when I'm at the craps table. I play craps for fun, and I am extremely interested in the mathematics of the game. I do understand distributions and p-values and most of all I understand that anomalies can exist much longer than you can stay solvent.

2- Regarding variance, I define it as the sum of the squared distances from the expected value. In small samples, sample variance can be a flawed (biased) statistic. The way I thought about it, just from an intuitive aspect, was that the lower the percentage of house edge, the lower the negative bias, and therefore the lower downside variance, in percentage terms.

Quote: RS

The odds bet increases variance.

I agree that the more you bet, the more volatile your bankroll will be, but I think it's important to look at upside and downside variance in separate ways. Forget the house edge for a second and pretend you have a choice between two bets, both actuarially fair: one where you win $5 if you roll a 6 and lose $1 if you don't (single die), and another where you win $1 if you don't roll a 6 and lose $5 if you do. Both bets have zero expectation and because they are symmetric, have identical variance, but clearly all of the variance on the first bet is "upside" and the latter is "downside". Well, not all of the variance, but my point is that one of these looks like purchasing a call option (small losses, larger, less frequent upside) and the other is like selling a call option.

Surely these things can't both be looked at identically.

And if you lower HE by increasing odds as much as possible, it would stand to reason that some of this volatility is shifted to the positive side of the distribution, as in decreasing the relative percentage of the downside variance?

Quote: dicesitter

Dicephd




If you are going to start with say a $5 pass line bet with 20 odds that is a starting position of $105
with no other bets on the table. You would need at least a gamblng account of $10,000- $11,000 just
to start. Now if you start with 20 odds and placed two come bets with similar odds you would need to
have $30,000 or more.

If you want to put as much as you can in your favor, start with a good class, one that will teach not
only a way to bet affectively, money management, but also throwing work if you have the desire to
take a look in that direction. Ahigh is not the place to start.

IN terms of the desire to go over the math in 17 different directions.. that is a waste of time. The
only math of the game of craps that means anything is were you on the numbers that were being
rolled.

If you started with $300, and cashed out $400, now see thats math.

Dicesetter

I suppose my bankroll is around $25,000 but I am a losing player overall. Sometimes I play small bets and 20x+ odds, sometimes I play small bets and take lots of sucker bets ($54 or $137 across for example) and just have fun. Dice is not an income source... it is a hobby that I lose money doing, but have tons of fun. I try and give myself a chance to be at the table for those long, anomalous rolls. My longest personal roll was probably around 25 rolls with no 7, and I've been at the table for a roll of about 35 with no 7. On those rolls I make a killing, and then bleed to death in the meantime. But it's a relaxing thing for me...a social thing. I have no delusions of "beating" an unbeatable game. I am fascinated by just digging into the numbers. Simulating results, and I even have fun when I lose (probably 70% of my sessions). I tip the dealers every hand, which is clearly -EV.

The reason I like Ahigh's videos is that it can take a static set of rolls and simulate what the p/l would've looked like with different starting orientations. It's interesting to me.

Quote: Sonny44

Mustang: Maybe you can clarify this for me, since no one in the past has. It involves "variance," what is meant by it. I see the term used in contexts I don't think it applies. I see "variance" as referring to how much a dice roll "varies" from the expected outcomes of the so-called pyramid of possibilities.

High variance would mean a series of dice rolls depart more from the pyramid predictions, resulting in fewer 7 outs and more points hitting. Low variance would mean they conform more to pyramid predictions, resulting in more 7 outs and fewer points hitting. This is my definition, the mathematical definition of variance. Please don't get into EVs, standard deviations, etc. I am math challenged. ; )

The OP, as many others, talks about lowering "variance" by increasing odds. Doing so doesn't apply to the "variance" as I understand it. So, what does the OP, et al., mean by increasing odds lowers "variance"? The only "variance" I can discern is that it lowers (varies from) the HE.

What do you understand about variance, not only mathematically, but how it's used in posts that seem to use it outside of the probabilities of certain numbers showing according to the pyramid?

Sonny44,

I'm going to take a non-math shot at answering your question, but it does necessitate using math terms to some extent. Hang in there for a few short paragraphs.

"Variance", mathematically, is simply the square of Standard Deviation on any game. Standard Deviation is how flat the various results are compared to the mean result of a particular bet. The OP and many other people, including me on occasion, misuse "variance" when they really mean "volatility". Volatility is a descriptor that includes variance. There are ways of reducing volatility, including hedging, some AP moves, doey-don't, stuff like that. But the HE and the SD stay the same on each component of any bet on a craps table. Since the SD doesn't vary/bet, neither does the variance.

Standard deviation is not determined by any single result. Rather, it's determined by an added series of results divided by the number of events. That result of that is the mean. How wide the swings are away from the mean in either direction is the deviation. And again, the variance is the square of that (which exaggerates the swing) and illustrates the volatility of any single bet. Again, you can change the volatility of your betting by combining/hedging/multiple bets. You can't change the variance by doing that.

I think it's easier to explain using a Roulette (American, with O and OO) example. If you bet on just 1 number (17, of course-joke), your volatility is extremely high. The HE on that number is 5.whatever%; it happens, at the mean, 1/38 times. We all know there are streaks, but the Standard Deviation on a game is determined over millions or billions of spins, and that number won't change. Since SD won't change, variance won't change. But "good" variance would happen if you won more than 1 in 35 times, or a little less than 1/2 of the time, which would be a net profit for you if you flat bet, in the relatively short span of a couple hours betting.

Go to the other extreme and bet all 38 numbers. Your volatility is nothing. But the chance of any ONE of those 38 numbers winning is still the same. You still get paid 1:35. The HE didn't change; the house made a net profit of 2 bets, the same as if you played 17 38 times and won once. The SD of getting that number didn't change, and so the variance didn't change. But a person who did this could rightly claim they had "won" every spin they'd ever bet on. They just lost more MONEY than they won every time because of the built-in house edge. True odds (no HE) would be 1:37. Even then, the SD of getting any one number, and thus the variance, is still exactly the same.

Hope I got that right. And hope it helps.

Quote: AxelWolf

LOL Swell, Max odds no doubt my best bet on this situation. Easy come easy go. Not enough BR? I just happen to know someone with an extra $2500. He to is an Ahigh fan. Boy oh boy what a wonderful coinkydink, unfortunately he just woke up and found out he got suspended and you showed up just in time, I'm sure you will fill the gap quit nicely. You have similar interests. Welcome nice to have you, we need more craps guys like you.

I to, have reason to suspect a bias in lots of things. PhD student in quantitative discipline writing, how perfect, your teacher would call you out for misspelling minuscule. However, I wouldn't know.

I did a little searching to understand all the vitriol. I had a misspelled word. Clearly discount the rest of what I say.

But even after reading about the mythical $2500 bet that was destined never to happen, I don't fully understand how I fit that mold. Perhaps I overplayed the idea that I'd like to see my p-values from real rolls (again, the only thing of interest was that the particular configuration of 7s was anomalous). To be clear, the funniest thing about the golden touch girl was that there existed someone out there who was "intelligent" enough to overcome the laws of physics but not intelligent enough to figure out how to "bet more on 8, less on 4, more on 12".

In the spirit of trying to ingratiate myself into this forum, I will let my posts separate me from golden touch pimp guy over time.

I won't even try and critique the following "sentences":

'I to, have reason to suspect a bias in lots of things. PhD student in quantitative discipline writing, how perfect, your teacher would call you out for misspelling minuscule.'

And I certainly wouldn't attempt to extend the 5? 6? 7? mistakes in grammar to an attack on your character or anything. Clearly, it's just the internet. But thanks for your help.

Quote: DicePhD

And if you lower HE by increasing odds as much as possible, it would stand to reason that some of this volatility is shifted to the positive side of the distribution, as in decreasing the relative percentage of the downside variance?

this can get over my head pretty quickly, but at least you are using English.

>decreasing the relative percentage of the downside variance?

I suppose this is what odds players are trying to do by playing free odds. When I play I know I am taking a situation [passline betting] that means a 49.5% odds of winning, roughly, to no significant difference from 50-50, roughly, if I can bet enough odds [edit].

I have often wondered if DI, if possible, could maybe push a very small advantage over the top since miniscule edge now needs to be beat.

But here is the rub on that: you want less variance in positive expectation, not more [edit]

This is quite shocking for you to suggest that you are decreasing the variance in some manner.

Quote: DicePhD

I agree that the more you bet, the more volatile your bankroll will be, but I think it's important to look at upside and downside variance in separate ways. Forget the house edge for a second and pretend you have a choice between two bets, both actuarially fair: one where you win $5 if you roll a 6 and lose $1 if you don't (single die), and another where you win $1 if you don't roll a 6 and lose $5 if you do. Both bets have zero expectation and because they are symmetric, have identical variance, but clearly all of the variance on the first bet is "upside" and the latter is "downside". Well, not all of the variance, but my point is that one of these looks like purchasing a call option (small losses, larger, less frequent upside) and the other is like selling a call option.

Surely these things can't both be looked at identically.

And if you lower HE by increasing odds as much as possible, it would stand to reason that some of this volatility is shifted to the positive side of the distribution, as in decreasing the relative percentage of the downside variance?

Again, I'm fiddling above my head (see my last). But the terminology you're using is a lot of what's causing contention. (The other is circumstantial and coincidental; there is no evidence beyond the interest in dice and the timing of your joining the board.)

The HE of any particular bet remains the same, and is additive for multiple bets, no matter what you do. By playing high odds, you are reducing the absolute amount of your exposure to the Vig, assuming you're playing a full multiple odds of your base bet amount. And your base bet amount is the correct amount to allow you to receive full odds in return (no rounding). But, and it will seem like splitting hairs to say so, you are doing NOTHING to change the HE itself. That is fixed and immutable. Each bet has one, including any base bet on which you may place odds. That they offer a 0% HE on the odds for a bet that has a built-in HE just means they are willing to be satisfied with the long-term results of a coin flip beyond what they will earn on the base bet, every time you bet it (over time).

Quote: beachbumbabs

you are doing NOTHING to change the HE itself

I think you mean EV.

Quote: odiousgambit

I think you mean EV.

the -EV and the HE in craps and roulette are synonymous, to the best of my understanding.

Quote: odiousgambit

this can get over my head pretty quickly, but at least you are using English.

>decreasing the relative percentage of the downside variance?

I suppose this is what odds players are trying to do by playing free odds. When I play I know I am taking a situation [passline betting] that means a 49.5% odds of winning, roughly, to no significant difference from 50-50, roughly, if I can bet enough odds [edit].

I have often wondered if DI, if possible, could maybe push a very small advantage over the top since miniscule edge now needs to be beat.

But here is the rub on that: you want less variance in positive expectation, not more [edit]

This is quite shocking for you to suggest that you are decreasing the variance in some manner.

I think the crossover between actual statistical terminology and dice voodoo is partly to blame, but I appreciate a civil discourse in English :)

If we think of trying to use short-term results to estimate a long-term success (or failure) rate, the real culprit is bias. Not bias from throwing dice, but in the statistical sense. I argue, perhaps not so eloquently, that house edge is a source of bias, and if we can lower it (in relative terms), then we lower the bias associated with any projection. I think this is pretty much objective fact, and the only place where we can pseudo-affect the HE is on the odds, if we look at a "pass" bet as the combination of both the pass line and the odds. Nothing beyond craps 101 here, but I'm trying to make the following extension:

Take a hypothetical example where you make some other bet on the craps table... say you place the 6 for $60 every hand. For the sake of argument assume you play 100 hands every day for 100 years so you have a sufficient sample. Take your (admittedly arbitrary) "daily" bankroll wins and losses and calculate the variance thereof. The reason that variance is so attractive is because it is easily standardized, so the total level of wins and losses is not relevant to the risk. I'm thinking in percentage terms. To think of it visually and less statistically, just plot the resulting daily bankrolls as a histogram and see what the "shape" looks like.

We all know the place bets have defined, static house edge. This is intertwined with EV of any given bet or strategy or whatever. What I wonder is whether or not changing the house edge has any effect on bankroll variance, or if it merely impacts the per-roll/per-resolution expectancy. I'm not even certain the answer, but that's the point of the thread.

Clearly, if all the sudden we arbitrarily change the payout of the place bets so the house how has an edge of, say, 5% for the placed 6, this is bad for our trivial "strategy". It would seem that in addition to lowering the expectation of each roll, when we summed up our "portfolio" of days, the losses would exhibit higher variance. By increasing the likelihood of negative days and decreasing the payoff to "wins", it would seem that overall variance goes up, but the majority of it coming from the variance of our losses, or downside variance.

If that hypothetical is accepted, the opposite should hold true as well. Variance still increases, but downside variance goes down (as we decrease the effective edge, by considering odds+pass line a combined bet).

Just spitballing...

Quote: beachbumbabs

the EV and the HE in craps and roulette are synonymous, to the best of my understanding.

Sort of. EV is, as I would define it, 1-HE

For example, if you have a .5% house edge in roulette on the black/white or whatever, then you are expected to recover 99.5 cents for every dollar you bet.

Or, your expected value of betting on red is .995% (I don't think this is the right number, just using it as an example), which = 1 - .005

Quote: DicePhD

2- Regarding variance, I define it as the sum of the squared distances from the expected value. In small samples, sample variance can be a flawed (biased) statistic. The way I thought about it, just from an intuitive aspect, was that the lower the percentage of house edge, the lower the negative bias, and therefore the lower downside variance, in percentage terms.

great! So, on a table allowing 0-100x odds, why do you feel like 20x provides the optimal level of variance? Variance is minimized at 0x, house edge is minimized at 100x. It's easy to see the arguments for betting at those levels. What is the mathematical basis for selecting another bet level?

Quote: rdw4potus

great! So, on a table allowing 0-100x odds, why do you feel like 20x provides the optimal level of variance? Variance is minimized at 0x, house edge is minimized at 100x. It's easy to see the arguments for betting at those levels. What is the mathematical basis for selecting another bet level?

I don't think there is anything special about 20x. I think it's an artifact of non-mathematical processes- single chips, some bankroll considerations, etc.

But I think you've hit on an interesting aspect, which is honing in on what type of "goal" is optimal. I once read an interesting discussion on the biggest sucker bet at craps. A math PhD was trying to make the case that the only way to answer that was to determine the initial goal. It was a detailed discussion, but interesting nonetheless. The idea was if your goal was to turn $100 into $1000 or go broke trying, it might not make sense to take a bunch of low-edge, low-EV bets... it might be best to just bet $1 on boxcars and parlay it, hoping to hit once.

I suppose Kelly would tell me what level of odds a given bankroll could sustain though, if I wanted to optimize.

Your expected value (EV) is the edge the player has (PE or Player Edge). It's a negative number, unless you're doing AP.

EV = PE, ie -0.5%
HE = -EV, ie: 0.5% in blackjack
ER (expected return) = 1-HE, ie: 100% -0.46% = 99.54% (Jacks or Better)

DicePHD



I am glad it is interesting to you, and if you have around $25,000 why not give it to me instead of the casino where
it is sure to end up.

I understand all this math stuff seems interesting like Ahigh's P values which are a bunch of nonsense in terms
of actually playing craps. Ahigh wanted some numbers so i went down to my table and threw a bunch of rolls
and sent them to Ahigh.....

He accused me of making them because they were way to good in terms of "p" value....pure nonesense, they
showed some outside number bias.... period.


If....if ... you are really interest in numbers, learn to throw consistantly, then take a look at your records to see if
they tell you anything about your roll. Now that is far more interesting and relevant than looking at a bunch of numbers
from some one else that you cant possibly duplicate.

I think Ahigh loves the game of craps, and i admire him for his dedication, but he, nor i have any of the answers your
looking for... they all have to come from you and what you can learn about the game and the throw. Now i am not talking
about dice control, long rolls are luck, but there are a lot of things that can make you have better luck.


good luck on your craps play....

dicesetter

I cab run that measly 25 grand into some really money. Meet me in Vegas and I will show you what a real DI can do.

Quote: dicesitter

DicePHD



I am glad it is interesting to you, and if you have around $25,000 why not give it to me instead of the casino where
it is sure to end up.

I understand all this math stuff seems interesting like Ahigh's P values which are a bunch of nonsense in terms
of actually playing craps. Ahigh wanted some numbers so i went down to my table and threw a bunch of rolls
and sent them to Ahigh.....

He accused me of making them because they were way to good in terms of "p" value....pure nonesense, they
showed some outside number bias.... period.


If....if ... you are really interest in numbers, learn to throw consistantly, then take a look at your records to see if
they tell you anything about your roll. Now that is far more interesting and relevant than looking at a bunch of numbers
from some one else that you cant possibly duplicate.

I think Ahigh loves the game of craps, and i admire him for his dedication, but he, nor i have any of the answers your
looking for... they all have to come from you and what you can learn about the game and the throw. Now i am not talking
about dice control, long rolls are luck, but there are a lot of things that can make you have better luck.


good luck on your craps play....

dicesetter

The purpose of collaborating is to utilize existing tools to analyze my own throws and to just enjoy studying the game. I worked in the hedge fund world for a number of years... lucrative and excruciatingly boring and terrible. We made +EV and -EV bets all the time. I played millions of hands of poker, and was a slightly winning player in tournaments.

The main reason I won't give you $25k is because I don't know what kind of entertainment and mental stimulation I will get for it. I know that given the way I play, $25k will end up in the casino almost certainly, and it will take a decent amount of time to do so. I have no delusions. But this is what I get in return:

Mental stimulation (I'm in the 99th percentile of all dorks)
Camaraderie
Nonsense/bullshitting with dealers (we have a good time)
The utility that comes from tipping and helping them out (I spread $6 across for the dealers each hand and let them ride/press/bank as they choose... most hands they get $1-2 in the box, but the record is turning that $6 into $275 across and dropping along the way. Box lady guessed that hand we dropped about $1,000 from that $6 alone.) I don't care about the exact level of HE there... it was fun.
"Free" stuff- rooms, nice steak dinners, TVs, other giveaways, etc. (again, I'm not an idiot... these are what casinos call "reinvestment" and I understand why... but in terms of recovery value, they have a positive effect).
An outside chance of a ridiculous roll or two and the associated fun that comes from it. (The same roll where we dropped about $1k for the dealers, I made about $9k...I don't look at this as anything other than an extension of my time-to-ruin, but this is a good thing, since I have a finite life expectancy).
A few more lottery tickets into enormous giveaways, etc.

Ultimately, all of these factors are worth more than $25,000 to me over the course of a few years. The utility of giving you $25k is negligible. But I'm always willing to listen to offers.

Quote: beachbumbabs

the -EV and the HE in craps and roulette are synonymous, to the best of my understanding.

EV = Bet * HE in my book

thus HE = EV/Bet [not -EV]

An HE can be changed in Craps by using free odds. Or as has been explored recently, many other bet combinations.

The EV of a pass line bet of $10 is immutably about -14 cents. However, the combination when playing 20x odds [as has been mentioned in the thread] gets the HE lowered to 0.099% . Yet the expected loss on that bet is still 14 cents. That is why I think you are thinking of EV. The HE certainly can be changed!

Quote: DicePhD

Variance still increases, but downside variance goes down

I guess I can see what you are saying, but it sounds very theoretical and unproved. Sounds like Skew, but no one seems to say Skew really matters. Till you. Unfortunately, over my head.

I am still thinking there is the paradox that once EV becomes positive, the variance goes from an effect of lowering HE, approaching 0% HE, to being "the enemy" once +EV is achieved for the player.

Quote: odiousgambit

EV = Bet * HE in my book

thus HE = EV/Bet [not -EV]

An HE can be changed in Craps by using free odds. Or as has been explored recently, many other bet combinations.

The EV of a pass line bet of $10 is immutably about -14 cents. However, the combination when playing 20x odds [as has been mentioned in the thread] gets the HE lowered to 0.099% . Yet the expected loss on that bet is still 14 cents. That is why I think you are thinking of EV. The HE certainly can be changed!

I might be confusing things by normalizing everything to $1 bets so that you can see the percentages and cents in unison.

But by this formulation, the HE would be something like 99%, which is a matter of notation, but I think it might confuse things.

Expected value is simply the average of a random variable. So, think about a random variable being the result of a pass line bet. Every time you win, your $5 becomes $10 and losses become $0. In the limit, as the # of trials increases to infinity, you will have a bunch of $10s and $0, the average of which will be about $4.93 (or 7 cents of house edge per $5 bet... same as the 14 cents for $10).

So, there should be a uniform definition that the expected value be in the denomination of the bet. But by this definition, HE = $4.93/$5 which is ~.986 = 98.6%, but certainly we wouldn't say the house has an edge of ~98.6%. Their edge is 1 - this number, or 1.4%.

The edge on the actual $5 pass line bet can't be changed by odds, but the combined "strategy" has the same edge of $.14 regardless of how much is put on odds. The odds simply decrease the denominator in the above fraction.

I think we are all saying the same thing in general, which might make it more confusing :)

Quote: odiousgambit

I guess I can see what you are saying, but it sounds very theoretical and unproved. Sounds like Skew, but no one seems to say Skew really matters. Till you. Unfortunately, over my head.

I am still thinking there is the paradox that once EV becomes positive, the variance goes from an effect of lowering HE, approaching 0% HE, to being "the enemy" once +EV is achieved for the player.

I agree it's theoretical... and I'm hoping to research this with more concreteness in the near future.

Skew is definitely in play here, and is partially accounted for in the bias. Essentially the HE shifts the distribution to the left (a mean of less than zero), but this in an of itself says nothing about the tails of the distribution or the relative symmetry thereof. A highly skewed strategy would be to bet minimum on high/low and parlay one win. You would have a ton of small losses and a rare enormous win. You could create a left tail as well by essentially placing across for 1 roll and coming down after 1 hit. You'd have a small win "most" of the time, but when you lost, your losses would be relatively large. (this is the 'picking up nickels in front of a steam roller analogy... an extreme version of which would be playing about 32 numbers in roulette each spin).

Higher order moments (skew and kurtosis) are of some interest, but are less understood by the "average" person. Essentially these are just the (sum of) cubed and quadratic differences from the mean, much like variance is the squared difference.

If and when I can get my hands on the code to do this efficiently, the effect, if any, of lowering relative house edge on bank roll upside/downside variance can be settled. I'll happily report.

Quote: odiousgambit

EV = Bet * HE in my book

thus HE = EV/Bet [not -EV]

An HE can be changed in Craps by using free odds. Or as has been explored recently, many other bet combinations.

The EV of a pass line bet of $10 is immutably about -14 cents. However, the combination when playing 20x odds [as has been mentioned in the thread] gets the HE lowered to 0.099% . Yet the expected loss on that bet is still 14 cents. That is why I think you are thinking of EV. The HE certainly can be changed!

Touche, EV = Bet * HE. Errrrr, rather, EV = Bet * -HE.


Although, I'd say the scenario you gave does not change the HE. The HE on the pass line does not change because odds are backing it up. The combined HE (on all the action) changes, although, I think that's an unfair way to describe it, as the PL and the odds are two separate bets.

Dicephd


I cant argue that giving me $25,000 wont make you feel better, but it sure as hell will make
me feel better.

I can surely appreciate the fun of the game, but for better or worse, the fun i get is from
taking ( or trying to take) their money . Remember that movie " color of money" Eddy said he
discovered that money won is twice as good as money earned. That's how i feel.

Having my own business since i was 22, i found i can make money, but beating them
on their table is another level of satisfaction.

But i am in no hurry, you want to think the $25,000 over for a tad, i have time (laughing)


dicesetter

Quote: dicesitter

Dicephd


I cant argue that giving me $25,000 wont make you feel better, but it sure as hell will make
me feel better.

I can surely appreciate the fun of the game, but for better or worse, the fun i get is from
taking ( or trying to take) their money . Remember that movie " color of money" Eddy said he
discovered that money won is twice as good as money earned. That's how i feel.

Having my own business since i was 22, i found i can make money, but beating them
on their table is another level of satisfaction.

But i am in no hurry, you want to think the $25,000 over for a tad, i have time (laughing)


dicesetter

I can appreciate all of these things for sure. I devoted countless hours to the mathematics of poker, and worked my way to a +EV player. I enjoyed the competition and still do. Without an online outlet for that (until it comes back), that part of me is infrequently put to work.

I think daily fantasy sports can be beaten (again, you're playing against a large mass of uninformed opponents), and I'm working on models to see if I can scrape EV out of that high-variance world.

My approach to "beating them" in craps is very different. The first weekend I ever played craps was in AC and I did not know it was happening (I was at a different hotel), but the infamous lady who rolled for 3 or 4 hours happened that same weekend. I don't know how many rolls it was with all of the nonsense surrounding a table like that, but my "goal" is to just be at the table when something like a 1.5 hr roll happens. I've been witness to 3 rolls that were at least 30+ in length (rolls, not minutes) and my aggressive -EV addiction gives me a MASSIVE right tail (that's a statistical term, not a sexual euphemism). If I had been at a table like that lady... even if it was "only" 100 rolls or so. I'd probably end up making 6 figures. That's the implied lottery ticket I buy by playing a bunch. I don't think it will ever happen, but the competitor in me wants to be at the table if it does. I've been fortunate enough to take all of the orange a table had (admittedly not many), and it was pretty fun.

It's really amazing how freeing it can be to enjoy craps but spend zero mental energy on trying to devise a "system". If you luck-rolled for 2 hours with me at the table, I might just give you $25k anyway :)

Take it from an unemployed PhD in chemical engineering...stop thinking this way.

And...does anyone want to hire me for a REAL job? Slot gaming math for various gaming companies are welcome...

Quote: tringlomane

Take it from an unemployed PhD in chemical engineering...stop thinking this way.

And...does anyone want to hire me for a REAL job? Slot gaming math for various gaming companies are welcome...

I have some real jobs for smart folks. We prefer guys who can do more than one thing, but sharp in casino math is a start.

Understand fun?

http://finance.yahoo.com/news/nanotech-partners-james-industries-bring-130000118.html

Quote: Ahigh

I have some real jobs for smart folks. We prefer guys who can do more than one thing, but sharp in casino math is a start.

Understand fun?

http://finance.yahoo.com/news/nanotech-partners-james-industries-bring-130000118.html

Plus he has a pinball machine in his office.

Quote: Ahigh

I have some real jobs for smart folks. We prefer guys who can do more than one thing, but sharp in casino math is a start.

Understand fun?

http://finance.yahoo.com/news/nanotech-partners-james-industries-bring-130000118.html

Is that janitor job still open ? I am willing to move !

Welcome.

May I suggest you start with Persi Diaconis a professor of Statistics of Ordinary Events at Stanford. He was a magicians assistant when he thought a Cuban casino had cheated him. He also gave a campus lunch time talk on loaded dice and how the "advantage" was eaten up by the errors of his graduate assistants who were recording the results.

Quote: DicePhD

I can appreciate all of these things for sure. I devoted countless hours to the mathematics of poker, and worked my way to a +EV player. I enjoyed the competition and still do. Without an online outlet for that (until it comes back), that part of me is infrequently put to work.

Anyone claiming to be good at online poker but uses this as an excuse, I cant take them seriously. They cant be serious players themselves.

Since it's +EV you should have no problems paying for Axel's guide to online poker playing post Black Friday 2014 .

Quote: AxelWolf

Anyone claiming to be good at online poker but uses this as an excuse, I cant take them seriously. They cant be serious players themselves.

Since it's +EV you should have no problems paying for Axel's guide to online poker playing post Black Friday 2014 .

As much as it hurts me that you don't take me seriously, I fail to see the point here. Sure there are ways to play online, from mild to extreme, but I suppose you are missing the opportunity cost argument here, or maybe just not fully understanding my personal utility function.

There are enough records online to show, provided you can be convinced of who I am, that I made a modest amount playing poker online. Nothing spectacular, and certainly not enough to make a living off it, but it was +EV for me to play. But that is a necessary and far from sufficient condition to play now. Consider any number of factors:

-I am currently getting a PhD. This is a massive time commitment, not conducive to putting in the required hours to make good money at poker.
-My personal "optimal" was always one or two-tabling MTTs and SNGs. At my "peak" I was a winning player on FTP at $117 hyper-turbo SNGs, 1-2 tables at a time...since the play is mostly formulaic, I tried 3-5 tabling with disastrous results. And a decent ROI for 1-2 tabling doesn't make up for the income that I can make living my life how I currently do it.
-I still play in live tournaments on occasion, which is +EV but spread out over time because I don't have the time to grind the WSOP circuit.

and so on...

I realize forums like this are just full of blowhards and "APs" and life-hackers galore, but still, what is the point of being confrontational? I was trying to explain my fascination for playing craps for the enjoyment I get from the mathematics of it all, and used poker as another example of why math keeps me interested. Now we have devolved into a worthless side conversation about how, given that I "claim" to be a +EV poker player but I don't play after Black Friday, you can't take me seriously.

I try to give a decent amount of leeway considering how much I already see forum regulars having to deal with the incessant "hey, look at my system to beat the casino!" type posts or "trust me, I can turn a random physical system into a well-oiled machine that lets me defy the laws of physics and large numbers simultaneously". But I am neither of these. I came here for my own holy grail...productive, insightful discussion of a hobby of mine. A fool's errand I see.

From the outset I've said a few things about myself:

1- I play craps for fun
2- I am a losing player (and have no problem admitting it)
3- I do not play optimally in terms of house edge...I like to play a certain way that is most entertaining for me
4- I am not amazingly rich, but I have enough money to enjoy playing a -EV game for all of the non-monetary benefits I get from it.
5- I used to play online poker a lot, but I don't anymore, even though I was a winning player.

And in return I get things like

1- Poorly worded (and grammatically shameful) responses about a misspelled word in one of my posts, clearly evidence that I'm not in an advanced degree program.
2- Challenges to the validity of my inconsequential statements regarding other non-dice hobbies.

What is the purpose?

I also play golf. I'm pretty decent at it. I played competitively in high school and I don't play as much anymore but I'm probably a 6-7 handicap and if I played regularly I could get that down to about a 3-4. I assume you will want to challenge the truthfulness of that statement, too? We both know it's not provable in any satisfactory way, but perhaps we can bicker about that as well?

Oh, and golf does not make me money. New equipment, green fees, etc. It's certainly -EV from a purely monetary standpoint, so why would anyone ever do such a thing? Could it be, just possibly, that when we make behavioral decisions about how to spend our time, we factor in things other than the expected gain or loss of fiat currency? I know this is a big stretch, and I would never be so bold as to make the claim in earnest, but is it possible? In some version of reality?

If so, and I realize this is a big ask, could it be that maybe I do things with my time that might not pay me money (golf, craps) or (gasp!) refrain from doing some things that are expected to make me money (poker)? And if I'm one of these weirdos, does it stand to reason that regardless of your implicit desire to believe what I say, my actions might be the same?

I really hope by now you have said "who cares?" to all of this. Because that's my point.

I thought the topic of upside/downside variance might be interesting. I mistakenly worded my first post in a way that made me seem like someone who thinks fictional stories about sharpshooters are substantive, and now I'm internet-bitching at a stranger because he/she doesn't take me seriously in my endeavor to not play online poker.

The internet is a wonderful thing.

Quote: DicePhD

I really hope by now you have said "who cares?" to all of this. Because that's my point.

You need to start out with that attitude and just hang on to it. What he said wasn't that bad, if unnecessary. You *are* a little hard to figure out, and doggone it it can obsess a fellow!

Quote:

I thought the topic of upside/downside variance might be interesting

I'm waiting for some of the better math guys we have here to weigh in on that. They haven't. I'm puzzling over that a bit, then again it is possible to block the entire category of "dice setting" at this site, and that might explain it.

Axel,

Dice is right. Let's give him a bit of room to get to know him, ok?

Odie,

I'm pretty sure a lot of the forum has this sub-forum blocked, as you suggest. Perhaps a math word problem that doesn't involve "systems" could be taken to the math sub-forum?

Quote: beachbumbabs

Axel,

Dice is right. Let's give him a bit of room to get to know him, ok?

Odie,

I'm pretty sure a lot of the forum has this sub-forum blocked, as you suggest. Perhaps a math word problem that doesn't involve "systems" could be taken to the math sub-forum?

How about a pure examination of what he is saying by some of our better math guys and gals? In other words, take dice setting out of the picture and make the question this: is it really possible to imagine increased variance in the normal sense to be beneficial to +EV play, from the theoretical things that our PhD candidate is talking about?

Here's my plebeian view: once an advantage is gained, no matter how tiny, reduced variance is desirable, since zero variance would mean for sure that edge would be realized every time. Example, a giant roulette wheel with 10,000 numbers that pays 9999:1* . If a player [or group of players] happens to bet one unit on every number, the variance is gone, and the house will earn exactly its tiny edge of one unit [I realize I have switched to House Advantage, but the concept is the same]. If however they rely on a random distribution of outcomes, on many days and perhaps for many weeks/months they could be showing a loss. The demoralization we see with our BJ APs going half-years in the hole is another vivid example of a group that would like to lower variance.

Our guy says, there is a type of variance that is different, that will in fact aid the side that has the edge. That's over my head.

Quote: odiousgambit

How about a pure examination of what he is saying by some of our better math guys and gals? In other words, take dice setting out of the picture and make the question this: is it really possible to imagine increased variance in the normal sense to be beneficial to +EV play, from the theoretical things that our PhD candidate is talking about?

I can't mathematically think of a way to show that the title of this thread to be possibly true because the opposite is true. Higher odds leads to higher variance. And also like you're hinting at, "advantage players" generally would prefer lowering variance.

The question, does higher odds lead to lower variance, or does higher odds lead to less variance????

Craps is a game of math, there is no question about it, and ever since the game was invented there are an
unending string of math experts which have tried to find a way to just walk up to the table and bet this or
bet that, and you will have casino money rolling out of your pockets.

This really is no different than proponents of dice control writing books and classes indicating the same thing.

In the end, none of this stuff has stopped the money from rolling into the pockets of the casino. Just ask
Ahigh about that, he has put in as much time on this as anyone.

So if math does not produce endless riches and dice control does not, what is the answer.....maybe it comes down
to a single word, the word the author of this thread started with....... variance.

If you had to describe craps in a single word..... variance would be it. every set of dice is different, every table is
different, and every player is different, all playing conditions are different, and if that is not bad enough, each of us
is different from one day to another. Nothing is constant.... so how in the hell do you win.

Well i would suggest if you win it is not every day, and not the same amount and not the same way. If you want to
win more than you lose, if that is possible, it has to be that you can adjust to what you see, have different ways
to throw the dice ( i am not talking about dice setting) i mean throw softer or harder or a different place on the table.
watch for short table trends, shooters that are having longer rolls. etc. In short you have to find some way to see
the variance happening at the time and make it work for you.

You cant fight it, you cant ignor it, and there sure as hell no math formula that sees it, adjusts to it, so all you have to
do is bet the same way everyday and on every table.

We all look for the pot at the end of the rainbow and there is none there.

It takes really hard work to win at this game, no different than anything else in life.

If you want to take your $400 to the table each week and just play, and say i had fun for the $400 so we are even
and i am happy, what Mustangsally says is correct. this is an easy game. You want to win..not so easy.

Dicesetter

Quote: DicePhD

As much as it hurts me that you don't take me seriously, I fail to see the point here. Sure there are ways to play online, from mild to extreme, but I suppose you are missing the opportunity cost argument here, or maybe just not fully understanding my personal utility function.

There are enough records online to show, provided you can be convinced of who I am, that I made a modest amount playing poker online. Nothing spectacular, and certainly not enough to make a living off it, but it was +EV for me to play. But that is a necessary and far from sufficient condition to play now. Consider any number of factors:

-I am currently getting a PhD. This is a massive time commitment, not conducive to putting in the required hours to make good money at poker.
-My personal "optimal" was always one or two-tabling MTTs and SNGs. At my "peak" I was a winning player on FTP at $117 hyper-turbo SNGs, 1-2 tables at a time...since the play is mostly formulaic, I tried 3-5 tabling with disastrous results. And a decent ROI for 1-2 tabling doesn't make up for the income that I can make living my life how I currently do it.
-I still play in live tournaments on occasion, which is +EV but spread out over time because I don't have the time to grind the WSOP circuit.

and so on...

I realize forums like this are just full of blowhards and "APs" and life-hackers galore, but still, what is the point of being confrontational? I was trying to explain my fascination for playing craps for the enjoyment I get from the mathematics of it all, and used poker as another example of why math keeps me interested. Now we have devolved into a worthless side conversation about how, given that I "claim" to be a +EV poker player but I don't play after Black Friday, you can't take me seriously.

I try to give a decent amount of leeway considering how much I already see forum regulars having to deal with the incessant "hey, look at my system to beat the casino!" type posts or "trust me, I can turn a random physical system into a well-oiled machine that lets me defy the laws of physics and large numbers simultaneously". But I am neither of these. I came here for my own holy grail...productive, insightful discussion of a hobby of mine. A fool's errand I see.

From the outset I've said a few things about myself:

1- I play craps for fun
2- I am a losing player (and have no problem admitting it)
3- I do not play optimally in terms of house edge...I like to play a certain way that is most entertaining for me
4- I am not amazingly rich, but I have enough money to enjoy playing a -EV game for all of the non-monetary benefits I get from it.
5- I used to play online poker a lot, but I don't anymore, even though I was a winning player.

And in return I get things like

1- Poorly worded (and grammatically shameful) responses about a misspelled word in one of my posts, clearly evidence that I'm not in an advanced degree program.
2- Challenges to the validity of my inconsequential statements regarding other non-dice hobbies.

What is the purpose?

I also play golf. I'm pretty decent at it. I played competitively in high school and I don't play as much anymore but I'm probably a 6-7 handicap and if I played regularly I could get that down to about a 3-4. I assume you will want to challenge the truthfulness of that statement, too? We both know it's not provable in any satisfactory way, but perhaps we can bicker about that as well?

Oh, and golf does not make me money. New equipment, green fees, etc. It's certainly -EV from a purely monetary standpoint, so why would anyone ever do such a thing? Could it be, just possibly, that when we make behavioral decisions about how to spend our time, we factor in things other than the expected gain or loss of fiat currency? I know this is a big stretch, and I would never be so bold as to make the claim in earnest, but is it possible? In some version of reality?

If so, and I realize this is a big ask, could it be that maybe I do things with my time that might not pay me money (golf, craps) or (gasp!) refrain from doing some things that are expected to make me money (poker)? And if I'm one of these weirdos, does it stand to reason that regardless of your implicit desire to believe what I say, my actions might be the same?

I really hope by now you have said "who cares?" to all of this. Because that's my point.

I thought the topic of upside/downside variance might be interesting. I mistakenly worded my first post in a way that made me seem like someone who thinks fictional stories about sharpshooters are substantive, and now I'm internet-bitching at a stranger because he/she doesn't take me seriously in my endeavor to not play online poker.

The internet is a wonderful thing.

I didn't read the entire post but the first few paragraphs made sense and seem fair. I believe you, I apologize for my skepticism.