Battle of the Betting Systems

Quote: darkoz

Quote: EvenBob

Quote: darkoz

Quote: EvenBob

Bet selection is everything, without the correct selection you have nothing. If you have the correct bet selection every progression will work but then you won't need it. How's that for irony.
link to original post

There is literally no irony there.

Just two people with different losing systems discussing why theirs is better.

They can't even agree on what to call it. One calls it a system while the other calls it a method.

Now that is ironic!
link to original post

If you do not see the clear irony in what I said then you need to look irony up in the dictionary because it doesn't mean what you think it means.
link to original post

Now that is ironic.

EB who doesn't understand irony says other people don't understand irony.
link to original post

I see the irony in your statement that I don't understand irony. Ironically you need to find a dictionary. In college I actually wrote a paper on irony, talk about ironic.

https://youtu.be/XfpB0kDLEts?t=19

ThomasK, 14.97% is not the house hedge. In your example, you are leveraging my initial 1 unit to expose up to 7 units to the house edge.

The casino will not give me this leverage for 'free'. The (2/38) house edge continues to be imposed on subsequent bets and extracted from my expected returns.

I have about a 10.6% chance of being up 7 units. I could also risk 1 unit on a square bet (4 numbers) and have a 10.5% chance of being up 8 units.

The square bet seems to offer the player much better odds than paroli. Why?

With paroli, the player would have to be up 1 unit, then 2 additional units, before having the opportunity to bet 4 units to be ahead a total of 7 units. Even when he loses the third bet, you can't calculate his expected returns as if he also lost the first two. When he generates a win on the first step and the second but loses the third, he is effectively 'spending' his profits to make that third bet. He may not be able to realize those initial profits under the rules of paroli, but they are wins for purposes of calculating total returns.

Whereas the paroli player seems to be disadvantaged in terms of chances to turn 1 starting unit into 7, he should have the same exact EV as other roulette players in terms of total wins divided by total units wagered. The paroli player leverages his 1 unit bankroll to 'buy' more spins and obtain the ability to make a higher average wager per spin. He increases his exposure to the house edge but he doesn't change it. Because he's putting more money into play, he will also be rated higher for comps compared to someone who flat bets 1 unit.

Quote: JackSpade

ThomasK, 14.97% is not the house hedge.

I will leave this for the moment. Let's see whether I will be able to convince you.

Quote:

In your example, you are leveraging my initial 1 unit to expose up to 7 units to the house edge.

It is not about exposing 7 units, because you never gave me 7 units. It is about turning your 1 unit into 7 units. How this is done is of no interest to you, because you hold me responsible for the details. The only thing you will experience is that in (18/38)^3 of the times you sent me off, I brought back these extra 7 units together with the 1 unit you gave me. In all other cases I returned to you with nothing.

Quote:

The casino will not give me this leverage for 'free'.

Since there is no harm done to the casino, there is nothing you have to fear. I will play regular roulette at a regular roulette table along the regular rules defined by the casino.

Quote:

The (2/38) house edge continues to be imposed on subsequent bets and extracted from my expected returns.

The (2/38) only effects my play and therefore dictates how often I will be able to return the extra 7 units to you. All you have to do is sit there and wait for me to return with these 7 units and your original 1 unit wager.

Quote:

I have about a 10.6% chance of being up 7 units. I could also risk 1 unit on a square bet (4 numbers) and have a 10.5% chance of being up 8 units.


The square bet seems to offer the player much better odds than paroli. Why?

The equations for the expected value are much more volatile than one intuitively would think. Please see my comparison of both chances of winning applied to the amounts of 7 and 8 units won.

p(winning)		7 units	8 units
(18/38)^3		=((18/38)^3)*7+(1-(18/38)^3)*(-1)	=((18/38)^3)*8+(1-(18/38)^3)*(-1)
		-0.14973	-0.04345
4/38		=(4/38)*7+(1-(4/38))*(-1)	=(4/38)*8+(1-(4/38))*(-1)
		-0.15789	-0.05263

Quote:

With paroli, the player would have to be up 1 unit, then 2 additional units, before having the opportunity to bet 4 units to be ahead a total of 7 units.

That is how it works and that is what I am trying to achieve at the table. Your task is to enjoy yourself and wait for me returning. You won't care about these details.

Quote:

Even when he loses the third bet, you can't calculate his expected returns as if he also lost the first two. When he generates a win on the first step and the second but loses the third, he is effectively 'spending' his profits to make that third bet. He may not be able to realize those initial profits under the rules of paroli, but they are wins for purposes of calculating total returns.

Sitting in the lounge you won't know what I am doing and you won't care either. How could you know that, for example, I found a very friendly dealer at an empty table who was glad to get something to do and happily appreciated my action? I would explain to him that I was trying to win three rolls in a row, and to relief him from all this juggling the chips we would count the next rolls together and he would only pay me 7 chips if three reds came up on the next three rolls.
Now no (intermediate) profits are spent. Yet, the total payout will remain the 7 units won plus the 1 unit wagered.

If a player doesn't have the chance to come to such an agreement with her dealer, the intermediate chips won are only tokens that allow her to continue the Paroli progression. Only on the third step these tokens are "converted" to the resulting amount she will take home.
I already tried to explain this in my comment on the Wizard's AMA episode #78.

Quote:

Whereas the paroli player seems to be disadvantaged in terms of chances to turn 1 starting unit into 7, he should have the same exact EV as other roulette players in terms of total wins divided by total units wagered. The paroli player leverages his 1 unit bankroll to 'buy' more spins and obtain the ability to make a higher average wager per spin. He increases his exposure to the house edge but he doesn't change it.

As explained in the previous section, the right to continue the progression is effectively not a question of "buying" spins, with the connotation of spending money for the next spin. As long as the player hasn't completed the Paroli sequence yet, only the fact that she won the previous spin qualifies her to play the next spin. The question is which tools are used to keep track of the sequence:

• With an idling dealer the player might simply count the number of rolls.
• If the casino offered three step Paroli as a possible betting option they could employ special chips, which have no value for the player, to keep track of the Parolis on the table.
• Or on regular tables with regular chips the player can simply define the intermediate winnings as marker chips which she is not allowed to cash and spend, because she hasn't finished her Paroli, yet.

Quote:

Because he's putting more money into play, he will also be rated higher for comps compared to someone who flat bets 1 unit.
link to original post

This is completely true. I answered this to Mental previously. I haven't done the math yet, but it seems to be worth the while. I'm thinking of combining the a priori approach for the progression itself with the a posteriori calculation for comps.
I guess the percentage of comps needed to break even, would be a good figure, which could be compared to the percentages casinos actually offer.
What do you think?

ThomasK,

You could devise any sequence of steps the $1 has to go through -- maybe after a win on roulette, it goes into a slot machine, then whatever remains bets banker on baccarat using a fibonacci progression -- and come up with a number that represents my chances of coming out X units ahead. The house edge only applies to actual wagers made.

I think roulette can be good game for comp grinding because it allows the player to make high probability bets with the expectation of a low volatility session. I certainly wouldn't use paroli or martingale. I would employ a gentle negative progression (raising by less than 1 unit after a loss).

I'm not an expert on comp hustling, but I've heard that buying in for a large amount helps even if you only intend to put a small portion of it into play. You will often get rated according to the size of your first bet, so if you are playing a system where you are willing to bet up to 4 units, then you might want to start with 4 units before resetting to 1 unit.

Some casinos place more weight on time spent at the table, so it's best to play on a full table that tends to slow the game down. Taking lots of bathroom breaks can extend the time your chips remain at the table as well.

Quote: JackSpade

ThomasK,

You could devise any sequence of steps the $1 has to go through -- maybe after a win on roulette, it goes into a slot machine, then whatever remains bets banker on baccarat using a fibonacci progression -- and come up with a number that represents my chances of coming out X units ahead. The house edge only applies to actual wagers made.

I think roulette can be good game for comp grinding because it allows the player to make high probability bets with the expectation of a low volatility session. I certainly wouldn't use paroli or martingale. I would employ a gentle negative progression (raising by less than 1 unit after a loss).

I'm not an expert on comp hustling, but I've heard that buying in for a large amount helps even if you only intend to put a small portion of it into play. You will often get rated according to the size of your first bet, so if you are playing a system where you are willing to bet up to 4 units, then you might want to start with 4 units before resetting to 1 unit.

Some casinos place more weight on time spent at the table, so it's best to play on a full table that tends to slow the game down. Taking lots of bathroom breaks can extend the time your chips remain at the table as well.
link to original post

JackSpade,

I'm struggling with the comp topic as well and it seems that the effect is too small to move the game considerably. But I'll keep fiddling ...

Somehow that tinkering sparked off a crazy idea in me about tipping the dealer and it seems that Martingale can produce some value for it. The idea is the following:

Let's assume we still are in the casino an you will send me off to play at the table.

Suppose you hand me 7 red chips equaling $35, and I simply put them all on red for one spin and return back to you either with nothing or with 14 reds, which is a profit of additional 7 units worth $35. The probability for winning is 18/38. Here is the calculation of the house edge based on payouts:



5.26% is what you expect from a double zero roulette anyway, so I guess you are fine with that figure and that you are aware that in the long run your bankroll will be grinded down by roughly 5 cents per dollar wagered.

Now I will change my play at the table and I will do a three step Martingale.

Still, you will hand me 7 units worth $35.
This time I will return to you either with nothing or with $39, which is $1 less than the 8 units I should have brought. The probability of this happening is (1-(20/38)^3), i.e. the probability of not losing all three possible rolls of the Martingale, as discussed earlier. The house edge, again calculated based on payouts, now is 4.82%, which should still be close enough to be accepted by you as the "regular" house edge for double zero roulette:



Still your bankroll keeps being grinded down by roughly 5 cents per dollar wagered.

The difference is that I now tip the dealer $1 every time I win the Martingale.

What is your take on that?

If I'm trying to grind $20 in points, I need to coin-in $12,000 on a machine. So if I bet $30 a spin on roulette, after 400 spins, that will total $12K of coin-in with a 5.26% HA of $631. My problem is I bet $3 so I need 4,000 spins to lose the same amount. If I spent $30/hand on BJ at 0.5%, I'd only lose $60 to the HA after 400 hands.

If I was doing a 10 in a row progression that totaled $1365, on roulette that'd be $72 of HA and on BJ it'd be $7 of HA. If I lose my 10th bet and win $765 for the series, I'd lose 10% of my total win to the HA on roulette, and 1% to the HA on BJ.

Quote: ThomasK

The difference is that I now tip the dealer $1 every time I win the Martingale.

What is your take on that?
link to original post

TLDR, but each time you win the martingale, you are only profiting one unit, so to give $1 away is a bit generous.

My Craps system is to bet one unit on do and one unit on don't. If the point is made, I take 6x and 3-4-5x odds. All bets resolve as a push and so are null bets unless a 12 is rolled. My EV on non-null bets is -50%.This seems to me the way TK would look at it. He would replace my system with a slot machine that takes two coins to play and always gives back 1 coin.

It is a pretty bad system that returns -50% on every non-null bet. I used to do this a lot with a $75 unit bet and it all worked out pretty well for me with 0,4% cashback and a huge leaderboard prize.

Quote: Mental

My Craps system is to bet one unit on do and one unit on don't. If the point is made, I take 6x and 3-4-5x odds. All bets resolve as a push and so are null bets unless a 12 is rolled. My EV on non-null bets is -50%.This seems to me the way TK would look at it. He would replace my system with a slot machine that takes two coins to play and always gives back 1 coin.

It is a pretty bad system that returns -50% on every non-null bet. I used to do this a lot with a $75 unit bet and it all worked out pretty well for me with 0,4% cashback and a huge leaderboard prize.
link to original post

What’s the point of the odds bets? Not comped and will always net zero

I assume you mean “if a point is established” not “made”

Quote: Ace2

Quote: Mental

My Craps system is to bet one unit on do and one unit on don't. If the point is made, I take 6x and 3-4-5x odds. All bets resolve as a push and so are null bets unless a 12 is rolled. My EV on non-null bets is -50%.This seems to me the way TK would look at it. He would replace my system with a slot machine that takes two coins to play and always gives back 1 coin.

It is a pretty bad system that returns -50% on every non-null bet. I used to do this a lot with a $75 unit bet and it all worked out pretty well for me with 0,4% cashback and a huge leaderboard prize.
link to original post

What’s the point of the odds bets? Not comped and will always net zero

I assume you mean “if a point is established” not “made”
link to original post

Yes, thanks for the correction of my craps terminology. The odds bet is comped on E-Craps. The reason for the $75 unit was to avoid shutting down for W2-G's. They changed to a hybrid E-Craps with a human shooter and removed the $1200 W2-Gs after I left the area. They also cut the cashback by a factor of four.

Quote: ChumpChange

If I'm trying to grind $20 in points, I need to coin-in $12,000 on a machine. So if I bet $30 a spin on roulette, after 400 spins, that will total $12K of coin-in with a 5.26% HA of $631. My problem is I bet $3 so I need 4,000 spins to lose the same amount. If I spent $30/hand on BJ at 0.5%, I'd only lose $60 to the HA after 400 hands.

If I was doing a 10 in a row progression that totaled $1365, on roulette that'd be $72 of HA and on BJ it'd be $7 of HA. If I lose my 10th bet and win $765 for the series, I'd lose 10% of my total win to the HA on roulette, and 1% to the HA on BJ.
link to original post

Thank you ChumpChange for your examples.
I'm not all that familiar with comps, therefore I'd like to ask some questions, if you don't mind:

Does that mean that $20 equal 12000 points, so 1 point would be worth 0.17 cents (=$20/12000)?

And 1 point would be awarded per $1 wagered?

Would this also mean that the house would give as comps 3.17% (=$0.0017/$0.0526) of the expected loss per dollar wagered on double zero roulette?

This might then also mean that 33.33% (=$0.0017/$0.005) of the expected loss per dollar wagered in Blackjack would be given as comps.
Is this reasonable or did I completely mess up the math?

Could please explain what kind of progression you use, because I don't seem to be able to figure it out based on the values of $1365 and $765? Thank you very much.

Quote: OnceDear

Quote: ThomasK

The difference is that I now tip the dealer $1 every time I win the Martingale.

What is your take on that?
link to original post

TLDR, but each time you win the martingale, you are only profiting one unit, so to give $1 away is a bit generous.
link to original post

Somehow the math says that it not only is possible to be generous but that furthermore the average loss per dollar wagered still is nearly half a cent lower than that of a single bet on red (=0.0526-0.0482=0.0044).



(20/38)^3 is the probability of losing all three steps, i.e. losing 7 units.
(1-(20/38)^3) therefore is the probability of winning one unit (=payout 8 units).
1 unit equalling $5.

Quote: Mental

My Craps system is to bet one unit on do and one unit on don't. If the point is made, I take 6x and 3-4-5x odds. All bets resolve as a push and so are null bets unless a 12 is rolled. My EV on non-null bets is -50%.This seems to me the way TK would look at it. He would replace my system with a slot machine that takes two coins to play and always gives back 1 coin.

It is a pretty bad system that returns -50% on every non-null bet. I used to do this a lot with a $75 unit bet and it all worked out pretty well for me with 0,4% cashback and a huge leaderboard prize.
link to original post

The paytable of the equivalent slot machine would look like this:

pays	probability
1	2.78%
2	30.56%
11	16.67%
12	22.22%
13	27.78%

Why isn’t the equivalent doey don’t slot take 13 coins and look like this:

pays	probability
12	2.78%
13	97.22%

Quote: unJon

Why isn’t the equivalent doey don’t slot take 13 coins and look like this:

pays	probability
12	2.78%
13	97.22%

link to original post

unJon,

Thank you. You're absolutely right. I blew it this time. Put the wrong parameters together.

Am I allowed to remove the post?

Quote: ThomasK

Am I allowed to remove the post?
link to original post

Edit (if you can) and spoiler it.

The "wrong" thought process can be useful reference in future.

Quote: ThomasK

Thank you ChumpChange for your examples.
I'm not all that familiar with comps, therefore I'd like to ask some questions, if you don't mind:

Does that mean that $20 equal 12000 points, so 1 point would be worth 0.17 cents (=$20/12000)?

And 1 point would be awarded per $1 wagered?

No, it means $6 coin-in equals 1 point and 2,000 points equals $20 of points (or $10 of free play on Thursdays). If playing regular slots with a ~10%-15% HA, they allow $3 coin-in per point.

Quote: ThomasK

Am I allowed to remove the post?
link to original post

You should be allowed to remove this whole thread.

Quote: ThomasK

Quote: Mental

My Craps system is to bet one unit on do and one unit on don't. If the point is made, I take 6x and 3-4-5x odds. All bets resolve as a push and so are null bets unless a 12 is rolled. My EV on non-null bets is -50%.This seems to me the way TK would look at it. He would replace my system with a slot machine that takes two coins to play and always gives back 1 coin.

It is a pretty bad system that returns -50% on every non-null bet. I used to do this a lot with a $75 unit bet and it all worked out pretty well for me with 0,4% cashback and a huge leaderboard prize.
link to original post

The paytable of the equivalent slot machine would look like this:

pays	probability
1	2.78%
2	30.56%
11	16.67%
12	22.22%
13	27.78%

The house edge should be 32.69%.

The equivalent slot machine would require 13 coins in in order to play passline and don't pass in parallel.

pays	probability	expected payout
1	2.78%	0.16667
2	30.56%	0.02778
11	16.67%	0.44444
12	22.22%	1.22222
13	27.78%	0.61111
total:	100.00%	8.75000
based on 13 coins in:
	return:	0.67308
	house edge:	32.69%

Equivalence of the paytables.

The probabilities of same payout amounts add up to the probabilities given in the paytable published on the machine. The player outside the slot machine is not able to distinguish which of the random outcomes caused which of the payouts of a certain amount.

pays	probability	expected payout		outcome	payout(pass)	payout(don't)
2	8.33%	0.16667		2, 3: craps	0	2
1	2.78%	0.02778		12: craps or push	0	1
2	22.22%	0.44444		7, 11: frontline winner	2	0
11	11.11%	1.22222		point is 4 or 10: 7 out	0	11
11	5.56%	0.61111		point is 4 or 10: won	11	0
12	13.33%	1.60000		point is 5 or 9: 7 out	0	12
12	8.89%	1.06667		point is 5 or 9: won	12	0
13	15.15%	1.96970		point is 6 or 8: 7 out	0	13
13	12.63%	1.64141		point is 6 or 8: won	13	0
total:	100.00%	8.75000
based on 13 coins in:
	return:	0.67308
	house edge:	32.69%

link to original post

I'm terribly sorry. Tense, tired, and in some haste I completely blew my reply above.
Many thanks again to unJon for verifying my math.
This should now be the correct information:

The paytable of the equivalent slot machine would look like this:

pays	probability
12	2.78%
13	97.22%

Quote: Dieter

Quote: ThomasK

Am I allowed to remove the post?
link to original post

Edit (if you can) and spoiler it.

The "wrong" thought process can be useful reference in future.
link to original post

Thanks for the advice.
I replied now with a full quotation.

"Edit" wasn't available any more. How long after posting is editing allowed?