The Arnold & The Arnold from the Don't - DiceData

The Arnold & The Arnold from the Don't
Two Strategies - 100,000 Bots - In Depth Analysis - DiceData
https://www.youtube.com/watch?v=lZOlBZ_xl_0

https://www.youtube.com/watch?v=aZKfLkPcdbE

Kind of a weird cool strategy that even Jeremy of Color Up was shocked at how much it was winning. Would love to see this explored by the DiceData guy and see if there's some weirdness to it for the average darkside craps player to smooth out wins/losses until you want to press your 'homerun' strategy.

If you're gonna start with $400 and set a win goal of $100 at a $10 table and never lose the $400, sounds like a plan. Eventually after 4 wins you could double up those numbers, and after another 4 wins, double up again. But I've had experience getting stranded with the high initial DC bet ladder going up and up and never coming down.

At the end he finally does some simulations, the only way to be sure it works. Alas I have to criticize him, there is no reason not to do millions of rolls and he does not do that. It looks like he instead does '10 shooter' trials by which he seems to mean 9 "hands" , that is, a beginning shooter and then nine 7-outs requiring a new shooter. He does do this multiple times, but clearly the number of trials is pathetically inadequate

We also have to trust he isn't cherry picking.

A seemingly nice guy, he fails to mention that each bet shown is -EV and thus cannot be combined into something +EV. Yes, he should come out and say this. ... Doesn't... and thus joins the ranks of what I will call the Exploiters without Integrity.


Does he know these things?

He gets given strategies to try out for a video and more times than not he loses badly at them through sheer bad luck, not from any fault of the strategy. He doesn't go over the HA of the bets for each strategy. He has done videos early on explaining the HA of each bet so he assumes you already know the HA of all these bets.

The 10 shooter data at the end is a new feature this past year and he gets to pick a few favorite strategies to compare. It tends to show where the strategy he just used compares to the Dolly, the Molly, and another he picked out. So if the Don'ts are winning will the strategy you just saw win or lose in comparison? The program he picked up only does so many shooters, so 10 is a reasonable number to expect from that program.

Quote: ChumpChange

He doesn't go over the HA of the bets for each strategy. He has done videos early on explaining the HA of each bet so he assumes you already know the HA of all these bets.
link to original post

Going over the HA is not the same as saying "that each bet shown is -EV and thus cannot be combined into something +EV. " ... everybody knows there is a house edge, but millions can be fooled into thinking a certain strategy can overcome it

He's making money with these videos, let's not praise him to the moon

.
there can be no better strategy at craps then betting pass and taking maximum odds or betting don't pass and laying maximum odds

it is the only type of table game bet that without any type of AP play gets the house edge down to below 1%

edit - woops - I almost forgot about basic strategy at blackjack which also brings the HE to less than 1%

on the linked page from the Wizard calculates the HE when taking just 2 times odds with a pass line bet as being just 0.572% and with don't pass and laying just 2 time odds as being just 0.455%

obviously, with 5 times or greater odds the HE would come even further

the main point of the vid is that this type of hedging reduces volatility and allows for maximum time at the table -

but if a player has a sufficient bankroll making pass line bets with maximum odds or don't pass laying maximum odds he will eventually get into the long run and his total loss will reflect the small edge he is playing against


https://wizardofodds.com/games/craps/appendix/1/

.

.
kinna amusing and ironic - to me anyway

even though taking or laying 5 times odds lowers the HE on the total action by a great deal -

in the long run a player betting $100 on the pass line and $500 on free odds -

will have the exact same long run loss as a player who never made any odds bets at all - but just bet $100 on the pass line________________________(-:/

in the short run, all the player did with the 5 times free odds bet was increase the risk of a much larger loss, but also the possibility of a much larger profit

I guess that's what gambling is all about - for a great many

.

I can either do a progression on the PL or take odds on the PL. The hot shooters required for multiple points are too far apart while winning 6 PL hits on one turn seems easier for the progression.

Quote: ChumpChange

I can either do a progression on the PL or take odds on the PL. The hot shooters required for multiple points are too far apart while winning 6 PL hits on one turn seems easier for the progression.
link to original post

if a bet was 50/50 such as a coin toss (and a PL bet is of course less than a 50/50 chance)

the chance of flipping heads successfully 6 times in a row is 1 in 64 or 1.5625%

on average, in the long run, if the PL bet was a 50/50 chance, in 64 attempts; 63 times you would lose in your attempt to win 6 times in a row; and one time you would win

don't wanna fuss with the math to figure out how much less exactly the chance would be on the PL compared to a coin flip

.

Quote: lilredrooster

.
kinna amusing and ironic - to me anyway

even though taking or laying 5 times odds lowers the HE on the total action by a great deal -

in the long run a player betting $100 on the pass line and $500 on free odds -

will have the exact same long run loss as a player who never made any odds bets at all - but just bet $100 on the pass line________________________(-:/

in the short run, all the player did with the 5 times free odds bet was increase the risk of a much larger loss, but also the possibility of a much larger profit

I guess that's what gambling is all about - for a great many

.
link to original post

All correct.

The only thing I'll add is that 'the long run' for the betting you are describing, ie free odds at 5x, is very, very long till you get into the area of 'the exact same long run loss as a player who never made any odds bets at all' ... In fact it takes so many rolls it is past what anyone can do in a lifetime save for something crazy like playing 8 hrs a day, 7 days a week, for 7 decades [or somesuch] and maybe not even then

and even then, it would be similar outcome by percentages, which allows for huge variation in the actual amount.

I get your point. My point is, playing with 5x or greater free odds will send you into such wild swings that the experience just won't resemble something that keeps 'returning to the mean'. Almost nobody plays this way on a consistent basis, and the higher the free odds multiplier, the more that statement is true

Quote: odiousgambit

Quote: lilredrooster

.
kinna amusing and ironic - to me anyway

even though taking or laying 5 times odds lowers the HE on the total action by a great deal -

in the long run a player betting $100 on the pass line and $500 on free odds -

will have the exact same long run loss as a player who never made any odds bets at all - but just bet $100 on the pass line________________________(-:/

in the short run, all the player did with the 5 times free odds bet was increase the risk of a much larger loss, but also the possibility of a much larger profit

I guess that's what gambling is all about - for a great many

.
link to original post

All correct.

The only thing I'll add is that 'the long run' for the betting you are describing, ie free odds at 5x, is very, very long till you get into the area of 'the exact same long run loss as a player who never made any odds bets at all' ... In fact it takes so many rolls it is past what anyone can do in a lifetime save for something crazy like playing 8 hrs a day, 7 days a week, for 7 decades [or somesuch] and maybe not even then

and even then, it would be similar outcome by percentages, which allows for huge variation in the actual amount.

I get your point. My point is, playing with 5x or greater free odds will send you into such wild swings that the experience just won't resemble something that keeps 'returning to the mean'. Almost nobody plays this way on a consistent basis, and the higher the free odds multiplier, the more that statement is true
link to original post

This seemed high, so I did a bit of math. You may be overstating it a bit but you are pretty much right.

I assumed that you just make pass line bets at the same amount, and always take 5x odds on any point.

I also assumed that there are about 2 rolls/minute. This is obviously dependent on how busy the table is, how fast the shooters are, and how competent and quick the dealers are at paying off bets, but 2 rolls/minute seemed like a decent average? I'm really not sure.

I also assumed that you went to the casino 3x/week and played for 4 hours each time you were there (so 12hr/week of play).

I figured that it would take almost 8 years (!!) before your (negative) expectation caught up to your standard deviation. That means that you have about a 1/6 chance of being even or ahead after 8 years of this. That's really not too bad for habitual -EV flat betting.

A recreational gambler who just goes to Vegas a few weekends a year and only plays a couple of hours a night while he's there could easily be ahead after a lifetime of this.

Us bubble craps players can rack up 20,000 rolls in 250 hours at 80 rolls an hour.

Quote: ChumpChange

Us bubble craps players can rack up 20,000 rolls in 250 hours at 80 rolls an hour.
link to original post

I was assuming 120 rolls/hr (2/minute). Is that too fast?

Typical full tables are much slower. Typical near empty tables can go quicker. I'm stuck with a machine that has countdown timers.

Deleted Dup post.

i did simulations with Wincraps a long time ago, plotting a reasonable lifetime of such play for a dedicated craps player, and using 10 or so different players experiencing a different set of rolls. At 10x or more the swings were astounding, some players getting rich, maybe 4 of 10 even [I blogged about it but have trouble accessing this myself]. Maybe a couple came out about even. Plenty lost big, some so horribly you just would not want to be in their shoes.

What each and every one of them experienced, though, was horrific and horrifying wild swings in bankroll. OK, thrilling too when it was going well. The fact is, no one in the real world can stand emotionally to go through these kinds of swings. Very few last long trying ... maybe nobody

Quote: SkinnyTony

Quote: odiousgambit

Quote: lilredrooster

.
kinna amusing and ironic - to me anyway

even though taking or laying 5 times odds lowers the HE on the total action by a great deal -

in the long run a player betting $100 on the pass line and $500 on free odds -

will have the exact same long run loss as a player who never made any odds bets at all - but just bet $100 on the pass line________________________(-:/

in the short run, all the player did with the 5 times free odds bet was increase the risk of a much larger loss, but also the possibility of a much larger profit

I guess that's what gambling is all about - for a great many

.
link to original post

All correct.

The only thing I'll add is that 'the long run' for the betting you are describing, ie free odds at 5x, is very, very long till you get into the area of 'the exact same long run loss as a player who never made any odds bets at all' ... In fact it takes so many rolls it is past what anyone can do in a lifetime save for something crazy like playing 8 hrs a day, 7 days a week, for 7 decades [or somesuch] and maybe not even then

and even then, it would be similar outcome by percentages, which allows for huge variation in the actual amount.

I get your point. My point is, playing with 5x or greater free odds will send you into such wild swings that the experience just won't resemble something that keeps 'returning to the mean'. Almost nobody plays this way on a consistent basis, and the higher the free odds multiplier, the more that statement is true
link to original post

This seemed high, so I did a bit of math. You may be overstating it a bit but you are pretty much right.

I assumed that you just make pass line bets at the same amount, and always take 5x odds on any point.

I also assumed that there are about 2 rolls/minute. This is obviously dependent on how busy the table is, how fast the shooters are, and how competent and quick the dealers are at paying off bets, but 2 rolls/minute seemed like a decent average? I'm really not sure.

I also assumed that you went to the casino 3x/week and played for 4 hours each time you were there (so 12hr/week of play).

I figured that it would take almost 8 years (!!) before your (negative) expectation caught up to your standard deviation. That means that you have about a 1/6 chance of being even or ahead after 8 years of this. That's really not too bad for habitual -EV flat betting.

A recreational gambler who just goes to Vegas a few weekends a year and only plays a couple of hours a night while he's there could easily be ahead after a lifetime of this.
link to original post

I don't believe it would take a tremendously long time for a player's experience to match his true expectation when betting 5 times free odds

the payouts - 2/1, 3/2, and 6/5 are not very extreme -

if the payouts were much more extreme such as 15/1 or 25/1 that would be much different

I would speculate that after 1,000 resolved bets the player is very likely to be very close to his true expectation much of the time

and I don't think it matters much if it was 2 times free odds or 100 times free odds - as long as the bets are consistently the same times odds

the key thing, I believe, is that the payout percentages are the same no matter how many times odds the player takes

although, I admit I don't have the math chops to demonstrate this decisively

.

Quote: lilredrooster

I don't believe it would take a tremendously long time for a player's experience to match his true expectation when betting 5 times free odds

it wasn't an intuitive thing for me either

Quote:

the payouts - 2/1, 3/2, and 6/5 are not very extreme -

if the payouts were much more extreme such as 15/1 or 25/1 that would be much different

I would speculate that after 1,000 resolutions the player is very likely to be very close to his true expectation much of the time

and I don't think it matters much if it was 2 times free odds or 100 times free odds - as long as the bets are consistently the same times odds

the key thing, I believe, is that the payouts percentage are the same no matter how many times odds the player takes

although, I admit I don't have the math chops to demonstrate this decisively

.
link to original post

Just use a simulation program, like Wincraps, that allows you to do millions of rolls without too long a wait, and you'll see

At one time I was playing in PA fairly often and you could get 10x odds at a $5 table. Even though it was 10x, I had fun doing this and would do it . I'm thinking that sort of thing is gone now. Even if it is still available, I think I was saved by my inability to really do it much ... too much travel. So, sounds doable for a 7 cent HA per passline bet, but generally you saw no one else doing it exclusively, max odds without doing other bets, unless you met up with somebody from here

I'd bet minimum with single or double odds and bump up the odds as I got ahead if I was at a 10X table. After I tripled my buy-in, I'd reset and add a come-bet with double the buy-in and try again. If I triple-up again, I'd move on to 3 come bets with double the buy-in on a new session. After that I'd double all the values and start over. Win 3 sessions more and double again. Win 3 sessions and make the last 3 sets of sessions 5X what the first session was. If I get through all that I'll have around 180X my original buy-in. So my $400 buy-in on a $10 table could turn into a near $72K cash-out after all that winning! When I get to a $6,400 original buy-in, I could end with over a million dollars from those sessions.

Quote: lilredrooster

Quote: SkinnyTony

Quote: odiousgambit

Quote: lilredrooster

.
kinna amusing and ironic - to me anyway

even though taking or laying 5 times odds lowers the HE on the total action by a great deal -

in the long run a player betting $100 on the pass line and $500 on free odds -

will have the exact same long run loss as a player who never made any odds bets at all - but just bet $100 on the pass line________________________(-:/

in the short run, all the player did with the 5 times free odds bet was increase the risk of a much larger loss, but also the possibility of a much larger profit

I guess that's what gambling is all about - for a great many

.
link to original post

All correct.

The only thing I'll add is that 'the long run' for the betting you are describing, ie free odds at 5x, is very, very long till you get into the area of 'the exact same long run loss as a player who never made any odds bets at all' ... In fact it takes so many rolls it is past what anyone can do in a lifetime save for something crazy like playing 8 hrs a day, 7 days a week, for 7 decades [or somesuch] and maybe not even then

and even then, it would be similar outcome by percentages, which allows for huge variation in the actual amount.

I get your point. My point is, playing with 5x or greater free odds will send you into such wild swings that the experience just won't resemble something that keeps 'returning to the mean'. Almost nobody plays this way on a consistent basis, and the higher the free odds multiplier, the more that statement is true
link to original post

This seemed high, so I did a bit of math. You may be overstating it a bit but you are pretty much right.

I assumed that you just make pass line bets at the same amount, and always take 5x odds on any point.

I also assumed that there are about 2 rolls/minute. This is obviously dependent on how busy the table is, how fast the shooters are, and how competent and quick the dealers are at paying off bets, but 2 rolls/minute seemed like a decent average? I'm really not sure.

I also assumed that you went to the casino 3x/week and played for 4 hours each time you were there (so 12hr/week of play).

I figured that it would take almost 8 years (!!) before your (negative) expectation caught up to your standard deviation. That means that you have about a 1/6 chance of being even or ahead after 8 years of this. That's really not too bad for habitual -EV flat betting.

A recreational gambler who just goes to Vegas a few weekends a year and only plays a couple of hours a night while he's there could easily be ahead after a lifetime of this.
link to original post

I don't believe it would take a tremendously long time for a player's experience to match his true expectation when betting 5 times free odds

the payouts - 2/1, 3/2, and 6/5 are not very extreme -

if the payouts were much more extreme such as 15/1 or 25/1 that would be much different

I would speculate that after 1,000 resolved bets the player is very likely to be very close to his true expectation much of the time

and I don't think it matters much if it was 2 times free odds or 100 times free odds - as long as the bets are consistently the same times odds

the key thing, I believe, is that the payout percentages are the same no matter how many times odds the player takes

although, I admit I don't have the math chops to demonstrate this decisively

.
link to original post

I didn't believe it either. Then I did the math and realized that I was wrong. It takes about 170,000 resolutions for the standard deviation to catch up to the expectation. What I mean is that in a 170,000 resolution sample, betting $100 with 5x odds, your expectation is -$240,404.40 but standard deviation for that sample is $240,129.72 (for a sample that large we can use the normal distribution approximation and it works just fine). So you have about a 1/6 chance of being even or ahead after this long. 170,000 resolutions is about 573,750 rolls. At 2 rolls per minute that's 4781 hr of play. If you played craps as a full time job (40hr/week) with no holidays or vacations it would take you well over 2 years to play that much. That is much more craps than most people play in a lifetime, and even after that much you aren't THAT sure that you are down. 1/6 is a reasonable chance to be ahead.

This is why we have the ability to do math like this, so we don't have to base what we believe on hunches. Again, my intuition was to agree with you but the numbers say otherwise, so I was wrong.

The reason that this happens is that the outcomes actually are quite extreme compared to the edge. With a $100 bet and 5x odds the house edge for a single resolution is $1.41 but your max win is $1100 (5x odds on 4 or 10 wins $1000, plus $100 on the pass line). This should give you a hint that what you said about 1000 rolls can't be right ... the difference between a win and a loss on a single point of 4 or 10 is $1700 ($600 loss vs $1100 win). That's more than the edge on 1000 resolutions.

.
got it - you're right

.

Is the variance on those 5X odds bets really 0? Seems they could be something like up to 1 to 3X +/-$1.2 million or something.

Quote: ChumpChange

Is the variance on those 5X odds bets really 0? Seems they could be something like up to 1- 3X +/-$1.2 million or something.
link to original post

The variance is certainly not 0. Variance of 0 means you get the same result every time.

I'm not sure what you're asking or referring to.

If you are breaking even at 170K PL decisions at +1 SD on the PL but are winning with +1 SD with 5X Odds decisions, you'd still be ahead $1.2 million. Not saying Odds and PL wins go together at all but if they did, you could get 6X more rolls-decisions before hitting the tipping point of negativity.

Quote: ChumpChange

If you are breaking even at 170K PL decisions at +1 SD on the PL but are winning with +1 SD with 5X Odds decisions, you'd still be ahead $1.2 million. Not saying Odds and PL wins go together at all but if they did, you could get 6X more rolls-decisions before hitting the tipping point of negativity.
link to original post

What? No. The numbers that I gave are combined.

Your total wins and losses over 170k resolutions have an expectation of about -$240k and a standard deviation of about $240k.

If you didn't make the odds bets your expectation would be the same but your standard deviation would be much much lower (somewhere in the ballpark of $41k, which is just the square root of 170k multiplied by your $100 bet). In other words your chances of being ahead after this long if you just bet pass with no odds are basically 0 (6 standard deviations is basically no chance).

All that the odds bet does is add a ton of variance. Which, for a -EV game is arguably a good thing. I doubt anyone would play if you just handed over $1.41 after every resolution, regardless of the result.

So just flat betting $100 on the PL with no odds and getting a +1 SD win will hit the break even point at just under 140K rolls or losing all of your $41,200+ of winnings to the HA. I've really got to make a winning play before I hit 100K rolls (1,000 hours at 100 rolls/hr).

Quote: ChumpChange

So just flat betting $100 on the PL with no odds and getting a +1 SD win will hit the break even point at just under 140K rolls or losing all of your $41,200+ of winnings to the HA. I've really got to make a winning play before I hit 100K rolls (1,000 hours at 100 rolls/hr).
link to original post

I'm not sure where you are getting $41,200 in winnings from.

The bottom line is that you are playing a -EV game and can expect to lose. If you throw in odds the variance goes way up and you have a better chance of winning and also a better chance of losing a lot more than you would otherwise.