New Project Ideas

How about a definitive study on dice setting? There have been various "small sample" studies and "challenges", but by your own admission, "One of the most frequently asked questions I get, and certainly the most frequent about craps, is whether dice setting is for real." (from the WoO site referenced previously).

How about arranging a large sample, "shootout" for proponents of various methods? Their roll-by-roll results can be tracked and you can provide analysis and commentary on the results. To make it even more interesting, they could even bring their own stakes and play for real cash (putting up real dough would have the advantage of performances being under actual playing conditions, but it may (will) severely limit the sample size.)

Quote: thecesspit

Your argument breaks down in the phrase "the only logical objective could be to lose as little as possible". The objective could be to have the greatest chance of doubling up or going home. The objective could be to make a 30% gain. The objective could be to have fun. Or it could be to increase your variance as you like the thrill of highs and lows.


None of these are invalid onjectives and they may all have different strategies.

And it is NOT immaterial on taking the odds if your objective is to lose as little as possible. Taking odds is 0EV but adds to the variance. If you want to protect your bank as hard as possible, don't take odds. The house % changes when you take odds, but the actual amount doesn't change. (However I would argue if you were betting table minimum and want to increase your action, the first increase should be taking odds... $5+$10 behind is probably better than $5 on the pass and hitting $5 come bets for preserving your bank roll longer).

And this is why I am interested in the variance and patterns over time rather than just knowing the EV of perfect play. It is of course the Wizard's site and time and I should go do the work myself ;)

Variance has no effect on EV. And I did say that the way to accomplish the objective of losing as little as possible was to make minimum bets on the Pass or Don't Pass. Making massive, modest, or no odds bets will not affect the fulfillment of that objective. Doing that will indeed increase your variance, but the sum of all those wagers will still equal zero--so the only consideration is how "losing" your losing bets are.

And BTW--there ARE no "patterns over time" in independent-trial games. It's an artifact of our primate brains to so desperately search for patterns where none exist.

Quote: thecesspit

Your argument breaks down in the phrase "the only logical objective could be to lose as little as possible". The objective could be to have the greatest chance of doubling up or going home. The objective could be to make a 30% gain. The objective could be to have fun. Or it could be to increase your variance as you like the thrill of highs and lows.

Let me elaborate. "Logical objective" linked to a STRATEGY. Any strategy related to doubling up, or winning 30%, or winning any amount, would be no viable strategy at all. Therefore, only a strategy related to minimizing losses would make sense. This, in turn, presumes that one is in some way compelled to play the game in the first place, since the OPTIMAL strategy is to simply not play. So the implied statement is, "Okay, if you MUST play, then..."

Wiz:

I think it would be interesting to have "variance" explained in
layman terms. I'm sure that would not be very difficult for
you. If you already have, can you point me to where ?

For example, my wife believes she can play longer on a 100 play
penny DW VP game with a 98.5 % paytable than I can play on a 99.7 %
(NSUD) paytable for the same buy-in, say $ 50. How many hands
do I have to play to essentially have the same variance ?

She usually plays $ 5 per hand (so 5 pennies per hand x 100 hands)
and I am playing 5 coins on a quarter machine. At our favorite
casino, I think the comp values are much better for having more
play through, but I can't quantify that.

I hope you keep your websites going, I enjoy them very much. These
forums are also a lot of fun.

Quote: Ayecarumba

How about a definitive study on dice setting?

I'd love to put that topic to rest as well. However, it would take at least 10,000 rolls to get close to proving or disproving anything. It would simply take too much time.

Quote: Wizard

I'd love to put that topic to rest as well. However, it would take at least 10,000 rolls to get close to proving or disproving anything. It would simply take too much time.

Well, you didn't ask for "easy"..hehe. In any case, I would still encourage you to consider it. The findings have the potential to revolutionize the way Craps tables are constructed... everywhere. Or they could put a real dent in dice control book sales.

Assuming six throws per minute during the sample periods, and counting in rest breaks and meals, I think the data collection would take about 40 hrs. tops, perhaps less if multiple samples could be collected simultaneously.

I actually think the hardest part would be getting dice control proponents to submit to tracking. They have more to lose if they cannot perform as promised.

Quote: Ayecarumba

I actually think the hardest part would be getting dice control proponents to submit to tracking. They have more to lose if they cannot perform as promised.

That will be hard. Harder still will be to get other dice control proponents to accept any results that don't favor dice control. They can always claim the volunteers weren't that good to begin with.

It's the Niven Principle (one of many) as applied to gambling: no ultimate sotry to end all stories ever ends anything. THat is, you take a science fiction premise and examine its ultimate logical conclusion, which ought to kill that premise. This often results in great stories, but good premises aren't easy to kill.

Dice control is a great premise: you can beat the casino at craps, consistently, and make a very good living at it; best of all the casinos don't go after dice setters.

I'd also like to see a serious study on dice control, especially one done against adequate controls. By that I mean against 1) regular shooters, 2) shooters who set dice but don't throw in a controlled fashion and 3) randomly generated "throws."

Quote: Nareed

Dice control is a great premise: you can beat the casino at craps, consistently, and make a very good living at it; best of all the casinos don't go after dice setters.

There is no folly that is too foolish to be believed by somebody who truly wants to believe it.
Dice controllers beat the casino at craps ONLY by separating suckers from their money at seminars before those suckers can take their wallets to the craps table.
If it really worked no one, but no one, would be selling seminars when they could make far more at the craps tables.

Quote: JohnnyQ

Wiz:

I think it would be interesting to have "variance" explained in
layman terms. I'm sure that would not be very difficult for
you. If you already have, can you point me to where ?

For example, my wife believes she can play longer on a 100 play
penny DW VP game with a 98.5 % paytable than I can play on a 99.7 %
(NSUD) paytable for the same buy-in, say $ 50. How many hands
do I have to play to essentially have the same variance ?

She usually plays $ 5 per hand (so 5 pennies per hand x 100 hands)
and I am playing 5 coins on a quarter machine. At our favorite
casino, I think the comp values are much better for having more
play through, but I can't quantify that.

I hope you keep your websites going, I enjoy them very much. These
forums are also a lot of fun.

Not only will she not play longer than you, she will bust out long before you do assuming the same bankroll and no cashback. The fact that she is on a 100-play does not make that much of a difference; the fact that she is betting $5.00 a hand on an inferior paytable is. I can't show you the math (I'm too stupid), but the difference in variance between a $1 single-line (5 credits) and an 100-play penny machine is not as great as people think it is. The fact of the matter is that she is betting $5 on every spin, which is a lot more than $1.25. Also, because she is playing with on a paytable that is 1.2% poorer than yours, her bankroll will need to be bigger by an order of magnitude to have the same risk of ruin. Sorry for the limited analysis, but the variance is not even close. In the words of caretaker Grady from The Shining, your wife needs to be "corrected, Sir."

Quote: mkl654321

Variance has no effect on EV. And I did say that the way to accomplish the objective of losing as little as possible was to make minimum bets on the Pass or Don't Pass. Making massive, modest, or no odds bets will not affect the fulfillment of that objective. Doing that will indeed increase your variance, but the sum of all those wagers will still equal zero--so the only consideration is how "losing" your losing bets are.

I am well aware that variance and EV are different. This is exactly my point. The EV just tells you the centre point of where you can be after x-trials. The variance will tell you the probability of how far from that centre point you are.

IF you goal is to take $200 and play $5 Craps for AT least 2 hours, placing x10 odds on on the Pass Line bets is not a good strategy to achieve this.

If your aim is to make $400 or go bust, placing those big odds is (probably) the best way of doing so.

If your bankroll was $2000, then I'd agree the odds bets are a immaterial. Simply put, if the variance can force into busto land, then it's important.

Quote:

And BTW--there ARE no "patterns over time" in independent-trial games. It's an artifact of our primate brains to so desperately search for patterns where none exist.

Yes there are patterns. There is the bell curve of probabilities. I am not talking about the pattern of results of each independent trial. I am talking about what is the pattern of possible results. Going away from an area the Wizard is comfortable with, I've produced simple curves of results for Martingale, progression and flat betting roulette. They look different. DUCY?

Quote:

Let me elaborate. "Logical objective" linked to a STRATEGY. Any strategy related to doubling up, or winning 30%, or winning any amount, would be no viable strategy at all. Therefore, only a strategy related to minimizing losses would make sense. This, in turn, presumes that one is in some way compelled to play the game in the first place, since the OPTIMAL strategy is to simply not play. So the implied statement is, "Okay, if you MUST play, then..."

I don't understand your objection to an objective being to attempt to double your money? Why is this not a logical objective, and why can it not have a strategy that optimises your chances to his that objective?

Not playing is the optimal way to guarantee you preserve your bankroll with 0 risk. People may be prepared to risk some of their bank roll, with a negative expectation, but with some chance to increase it.

Otherwise, 'optimal' strategy on roulette would be to cover both red, black and green. And on Craps to play pass/don't pass. But you've given up all the variance... and as I've stated elsewhere, trading variance for EV is one of the things a casino offers.

We are possibly using the word strategy in different ways here.