Applying betting systems with an edge
Flatbetting yields this advantage no matter what we try to do. We must stand the 1% edge fluctuations as well.
What about applying a mild MM or system to make an attemp to increase this 1% or win easier?
I was thinking about Oscar´s grind system, first introduced by PHD Allan Wilson in his "Gambler´s guide".
This system has got 99.8% in winning 1 units with 1000 BR and 99.95% with 5000 br. These %s are with -2.7%
What if we play it with +1% edge?
regards
ybot
it's still a +1% edgeQuote: ybotHi, we take for granted we win easily with an edge, for instance 1% playing even money bets.
Flatbetting yields this advantage no matter what we try to do. We must stand the 1% edge fluctuations as well.
What about applying a mild MM or system to make an attemp to increase this 1% or win easier?
I was thinking about Oscar´s grind system, first introduced by PHD Allan Wilson in his "Gambler´s guide".
This system has got 99.8% in winning 1 units with 1000 BR and 99.95% with 5000 br. These %s are with -2.7%
What if we play it with ?
regards
ybot
If you already have a defined edge, then there is such a system that tells you exactly how much to bet with your BR/variance/RoR/etc... That's Kelly Criterion.
KC is a classical tool to play.
OG systems longs for +1 to reset. I wonder how it worked with an edge.
We know that +1% edge is a fact. The quest is to get more than 1% or get it with lower fluctuations.
Cowing a +1% edge takes huge task. In case we play table limits we must stand it with a big BR.
In this example we play even chances with +1%, it doulbles the edge of playing 9 numbers with 1% edge on average.
But, is it the same to play 18 numbers with +1% and +2% on 9 numbers? How do fluctuate?
What if we played 24 or more numbers with +1%?
Math problem: if a spin of this +EV roulette game takes 60 seconds, how long would it actually take for you to bankrupt a casino's $10M bank starting with $100 in your pocket. Assume nobody else is playing.
There are variables that could give +1% or more playing from 12 to 30 numbers in roulette, such as bias play or VB.
The more numbers covered the better.
Quote: RomesAh, well it's quite simple then. When cowing a 1% edge across 9 numbers clearly one must consider BR requirements. However if you move to 18 numbers and it's a Thursday evening then you're more likely getting about 1.24%, not 2%. When the 3rd full move of the year arises though, this is the time to strike at the full 24 number mark. You see, when the moon is full (for the 3rd time only) then and only then will the gravitational affect on your betting chips be enough to gander the full 2% advantage. If that doesn't work then you can always get a cottage on an island and contemplate the meaning of the number 17.
?
Quote: MathExtremistMath problem: if a spin of this +EV roulette game takes 60 seconds, how long would it actually take for you to bankrupt a casino's $10M bank starting with $100 in your pocket. Assume nobody else is playing.
Average between 19 and 20 hours when using the most effective betting strategy.
But that can't be right because the edge is completely immaterial when it comes to winning or losing at roulette.
No, actually if you're playing a bias or clocking the wheel -- and you're right about the sector on the wheel where the ball will land -- you want to bet on as few numbers as you can. More numbers just increases your loss. But if you can clock a wheel and be right about the target 1/2 of the wheel 60% of the time, you don't just have a +1% edge, it's much larger.Quote: ybotMathextremist, a nice example of +1% with 0 variance, something impossible.
There are variables that could give +1% or more playing from 12 to 30 numbers in roulette, such as bias play or VB.
The more numbers covered the better.
Quote: ybotWe know that +1% edge is a fact. The quest is to get more than 1% or get it with lower fluctuations.
If edge is 1% then you should aspire to get 1% of more action. That's what KC does. MM doesn't increase it beyond 1%
Target 18/37 numbers 60% of the time is more than +1% edge
I made a question before, is it better or worse 18 numbers with +1% or 9 numbers with +2%. They take different BRs
To go to the extremes, you might have a 10% edge in 1 single number, you must play for your entire life to get a decent profit, or 18 numbers with 1% edge
What should be the best mathematical way to attack this situation with the lowest risk and BR? 1 number, 9 numbers, 18 numbers, 24 numbers, 30 numbers. Each of them with an edge
warm regards
ybot
Quote: OnceDearIf edge is 1% then you should aspire to get 1% of more action. That's what KC does. MM doesn't increase it beyond 1%
Oscar system is a combination of positive and negative progresion on 18 numbers with very high success rate.
BJ AP players are used to play with such a low edge, AP roulette players do not.
The purpose of this thread is to share experiences with math guys and APs
regards
Quote: ybotOscar system is a combination of positive and negative progresion on 18 numbers with very high success rate.
BJ AP players are used to play with such a low edge, AP roulette players do not.
The purpose of this thread is to share experiences with math guys and APs
regards
Deleted
Quote: TwoFeathersATLJust so I got all this right, which moon are we in now?
What do you mean by the moon?
I can appreciate that. Sharing experiences/thoughts is what drives innovation and ideas. Unfortunately to me it sounds like you're "discussing" (or promoting) a system that has to do with changing your bets, not the game/rules/etc/etc. Changing your bets alone will never change the house edge nor flip your negative expectation positive. Thus, this apparent roulette system where you bet positive & negative progressions on 18 numbers is a losing system. If you're just doing this for fun, or to try something new, then cool and knock your socks off. I just hope you're not toting that a system such as this is "good" for the player, changes the odds at all, or will help the player win, because it won't.Quote: ybotOscar system is a combination of positive and negative progresion on 18 numbers with very high success rate.
BJ AP players are used to play with such a low edge, AP roulette players do not.
The purpose of this thread is to share experiences with math guys and APs
regards
If you want to post the entire system I can mathematically show you how it's a negative expectation... If you want the mathematical response from an AP.
EDIT: I found the "system"... results to follow... http://www.roulettestrategy.net/strategy/oscars-grind/
Quote: ybotOscar system is a combination of positive and negative progresion on 18 numbers with very high success rate.
Define success.... Don't bother.
Quote:BJ AP players are used to play with such a low edge, AP roulette players do not.
Roulette players do not play with an edge. If you know one who does, then tell us what gives him that edge (Clue: It isn't money management. It would have to be actual wheel bias or accurate Visual Ballistics
Quote: ybotWhat do you mean by the moon?
Was a bad joke, I deleted after 15 seconds.
I would ask that you do the same.
Then I'll do the same, etc.
It's like a system.....
Quote: RomesI can appreciate that. Sharing experiences/thoughts is what drives innovation and ideas. Unfortunately to me it sounds like you're "discussing" (or promoting) a system that has to do with changing your bets, not the game/rules/etc/etc. Changing your bets alone will never change the house edge nor flip your negative expectation positive. Thus, this apparent roulette system where you bet positive & negative progressions on 18 numbers is a losing system. If you're just doing this for fun, or to try something new, then cool and knock your socks off. I just hope you're not toting that a system such as this is "good" for the player, changes the odds at all, or will help the player win, because it won't.
If you want to post the entire system I can mathematically show you how it's a negative expectation... If you want the mathematical response from an AP.
EDIT: I found the "system"... results to follow... http://www.roulettestrategy.net/strategy/oscars-grind/
I do not play mathematical systems, I flatbet, the worst enemy of Aps is fluctuation.
You might find out that I only play with an edge
Quote: OnceDearDefine success.... Don't bother.
Roulette players do not play with an edge. If you know one who does, then tell us what gives him that edge (Clue: It isn't money management. It would have to be actual wheel bias or accurate Visual Ballistics
Do you really believe that Bob Nersesian´s clients are all BJ players?
In Roulette there are 36 numbers, and two zeros (0 & 00). Your odds of hitting your First 18 bet are as follows:
EV(first 18) = (18/38)(1) + (20/38)(-1) = .4737 - .5263 = -.0526
...which is indicative of the well known house edge on even money bets of 5.26%, as shown here on the Wizards Roulette Page: https://wizardofodds.com/games/roulette/basics/
We can break the "System" down to it's 4 states. At any given time you're either betting 1 unit, 2 units, 3 units, or 4 units. For the sake of this example let's pretend our units are $10. Thus, 1 unit = $10, 2 units = $20, and so on.
EV(first 18 1 unit) = (18/38)(10) + (20/38)(-10) = 4.737 - 5.263 = -.526
EV(first 18 2 unit) = (18/38)(20) + (20/38)(-20) = 9.474 - 10.526 = -1.052
EV(first 18 3 unit) = (18/38)(30) + (20/38)(-30) = 14.211 - 15.790 = -1.578
EV(first 18 4 unit) = (18/38)(40) + (20/38)(-40) = 18.947 - 21.053 = -2.104
As you can clearly see here... At no point do you have a positive expectation. Why? Simply look at the negative aspects of the game. If you bet on ANY "even money" bet you have 18/38 chances to win because of the 0 & 00. Likewise, you have 20/38 chances to lose the bet. The ONLY thing that changes in these states is the amount of money you are betting... which given the house edge really only UPS the players expected loss.
Unlike what the page describing the system says, the obvious flaw (to any AP/math minded player) is that you're playing a negative expectation bet, regardless the size of your bet. In things like blackjack, players use the information and effect of card removal to understand when the edge CHANGES to the player. Thus, they don't win because they bet more, they win because they have the advantage over the house.
To compare it to this roulette situation, that would be like me saying... Every 10th spin the player WINS even money bets if 0 or 00 come up. This would CHANGE the odds so that you WIN 20/38 and lose only 18/38, yielding a positive 5.26% for the PLAYER. Then, by betting more one could expect to make more money because the actual odds of the game and advantage have changed.
Oscar's Grind Roulette System is mathematically flawed and will with 100% certainty lead to the player LOSING, in the long run.
Flat betting is fine, if you have an advantage. Without visual ballistics, ball steering, or biased wheels though you don't have an advantage on any bet of a regular game of roulette. You don't have to tells us what, obviously, but are you implying you have one of these or "something" that changes the original house edge of the game... that has nothing to do with betting?Quote: ybotI do not play mathematical systems, I flatbet, the worst enemy of Aps is fluctuation.
You might find out that I only play with an edge
Quote: RomesSo we can choose any "even money" bet... suggestions are Red/Black, Odd/Even, and first/last 18. Since ybot seems to be a fan of the 18, let's go ahead and assume we're going to be betting on the first 18 for our system trial.
In Roulette there are 36 numbers, and two zeros (0 & 00). Your odds of hitting your First 18 bet are as follows:
EV(first 18) = (18/38)(1) + (20/38)(-1) = .4737 - .5263 = -.0526
...which is indicative of the well known house edge on even money bets of 5.26%, as shown here on the Wizards Roulette Page: https://wizardofodds.com/games/roulette/basics/
We can break the "System" down to it's 4 states. At any given time you're either betting 1 unit, 2 units, 3 units, or 4 units. For the sake of this example let's pretend our units are $10. Thus, 1 unit = $10, 2 units = $20, and so on.
EV(first 18 1 unit) = (18/38)(10) + (20/38)(-10) = 4.737 - 5.263 = -.526
EV(first 18 2 unit) = (18/38)(20) + (20/38)(-20) = 9.474 - 10.526 = -1.052
EV(first 18 3 unit) = (18/38)(30) + (20/38)(-30) = 14.211 - 15.790 = -1.578
EV(first 18 4 unit) = (18/38)(40) + (20/38)(-40) = 18.947 - 21.053 = -2.104
As you can clearly see here... At no point do you have a positive expectation. Why? Simply look at the negative aspects of the game. If you bet on ANY "even money" bet you have 18/38 chances to win because of the 0 & 00. Likewise, you have 20/38 chances to lose the bet. The ONLY thing that changes in these states is the amount of money you are betting... which given the house edge really only UPS the players expected loss.
Unlike what the page describing the system says, the obvious flaw (to any AP/math minded player) is that you're playing a negative expectation bet, regardless the size of your bet. In things like blackjack, players use the information and effect of card removal to understand when the edge CHANGES to the player. Thus, they don't win because they bet more, they win because they have the advantage over the house.
To compare it to this roulette situation, that would be like me saying... Every 10th spin the player WINS even money bets if 0 or 00 come up. This would CHANGE the odds so that you WIN 20/38 and lose only 18/38, yielding a positive 5.26% for the PLAYER. Then, by betting more one could expect to make more money because the actual odds of the game and advantage have changed.
Oscar's Grind Roulette System is mathematically flawed and will with 100% certainty lead to the player LOSING, in the long run.
Oscar system and any other are 100% losers, no doubt.
In my scenario no matter we are in an american or european wheel, I have at least 1% advantage over the house(2.7 or 5.26).
The expectation is positive, we long for win more or easier(less risk) than 1%
Well, as mathextremist pointed out... if you have a guaranteed 1% edge no matter your bets (via rebate or something), and your main concern is to smooth out variance, then you should simply bet every number on the board.Quote: ybotOscar system and any other are 100% losers, no doubt.
In my scenario no matter we are in an american or european wheel, I have at least 1% advantage over the house(2.7 or 5.26).
The expectation is positive, we long for win more or easier(less risk) than 1%
Quote: RomesFlat betting is fine, if you have an advantage. Without visual ballistics, ball steering, or biased wheels though you don't have an advantage on any bet of a regular game of roulette. You don't have to tells us what, obviously, but are you implying you have one of these or "something" that changes the original house edge of the game... that has nothing to do with betting?
We know that the ways to get an edge in roulette is bias, vb or dealer´s signature.
I already have got the edge
Have you ever played with 1% edge?
Quote: RomesWell, as mathextremist pointed out... if you have a guaranteed 1% edge no matter your bets (via rebate or something), and your main concern is to smooth out variance, then you should simply bet every number on the board.
Variance/fluctuation becomes a 1% edge in an ordeal
You must place hundreads of bets with +1% edge, you are a human , you spend time and expenses.
Quote: MathExtremistIf you somehow had a roulette game where all the inside numbers had a 101% RTP, just bet all of them every spin for 0 variance and a constant 1% increase in bankroll. For what it's worth, an example of such a game would be a 100-segment wheel where inside numbers paid 100-to-1 (that is, 101-for-1). Bet on all of them, lose 99, win 100 = +1 per 100 wagered with 0 variance. Do that for 100 spins, get to 200 bankroll, then bet 2 units on each; do that for 50 spins, then bet 3 units on each, etc. After a short while you'll break the bank.
Math problem: if a spin of this +EV roulette game takes 60 seconds, how long would it actually take for you to bankrupt a casino's $10M bank starting with $100 in your pocket. Assume nobody else is playing.
This also makes other unlikely assumptions like 1)the table limits are $1 - $100,000; 2)no heat; and 3)the dealers can spin, collect the bets, and pay out those odd amounts like $97,949 ("Sir, just drop me $51, and I'll slide you $98,000!?") in the prescribed 60 seconds.
Now, I have a question (just curious -- I don't know the answer): Is there a point when I can "jump the gun" (e.g. bet $1 on 10 spots, and $2 on the other 90 spots at the point where my bankroll is $190) at each bet level and expect (>50% chance) to break the $10M bank in a shorter amount of time? What's the minimum time where my success rate is 50%?
Quote: ybotWe know that the ways to get an edge in roulette is bias, vb or dealer´s signature.
I already have got the edge
Have you ever played with 1% edge?
I wouldn't waste my time on a 1% edge in such a slow game.
Quote: IbeatyouracesI wouldn't waste my time on a 1% edge in such a slow game.
Liar, liar, pants on fire!
That was a joke, no insult.
Still not sure it was a completely honest answer though. I mean there are times when there's nothing better to do...;-)
With 0 variance, you could give me 1/10 that and I'd play it. Start with $10,000 and make $10/spin? That's $600/hour risk free. If you're doing better than that elsewhere, good for you. I don't make $600/hour. Maybe someday...Quote: IbeatyouracesI wouldn't waste my time on a 1% edge in such a slow game.
Don't focus on the percentage edge, focus on the win per hour. A 10% edge over the house isn't that important if you can only bet it at $1/play, but a 1% edge over the house is worth millions if the casino will book five or six figures per bet.
The flip-side is true too -- many people think that 3/4/5x odds on the passline is the "best bet in the casino" at 0.374% house edge, while others argue "no, it's the don't pass with full odds at 0.273% edge. But the truth is that making a $10 line bet costs about $0.14 for both pass and don't bettors regardless of how much in odds you take (or lay), so if you stand at the table for an hour and make 35 bets, you're looking at around $5 in expected loss per hour. Odds don't change that expected loss per hour, they just increase your variance.
Quote: MathExtremistWith 0 variance, you could give me 1/10 that and I'd play it. Start with $10,000 and make $10/spin? That's $600/hour risk free. If you're doing better than that elsewhere, good for you. I don't make $600/hour. Maybe someday....
Certainly, but I'm talking real world scenarios. Put it this way, I rarely, if ever, count cards any longer.
Quote: ybotYou might find out that I only play with an edge
How might we. When might we. Bring it on.
Give that guy the $25 for "Ad*van*tage Play Dreaming". He's "earned" it.Quote: MathExtremistI don't make $600/hour. Maybe someday...
You know, those "internet dollars" really add up once you get started.
Yep.Quote: OnceDearBring it on.
I think he has a legitimate question and I understand why he wants to cut down variance and see the 1% grow.
If he has 1% he's probably not going to "bring it on"
Quote: AxelWolfI think you guys are being hard on this guy. He clearly said he knows system betting can overcame the house edge. He believes he has a 1% edge on roulette. It's not impossible, there could be a legitimate bias, there could be a promotion etc etc.
I think he has a legitimate question and I understand why he wants to cut down variance and see the 1% grow.
If he has 1% he's probably not going to "bring it on"
I looked back through his posts. It's pretty clear that he is hoping to monetize his previous analysis of a few thousand spins where he went pattern hunting https://wizardofvegas.com/forum/questions-and-answers/math/23007-edge-on-roulette/#post478515
To that extent he is, indeed asking a valid question.
But he needs to go back to the answers he had to that post. He simply does not have the evidence to support his hypothesis that he has found a biased wheel..
Quote: OnceDearI looked back through his posts. It's pretty clear that he is hoping to monetize his previous analysis of a few thousand spins where he went pattern hunting https://wizardofvegas.com/forum/questions-and-answers/math/23007-edge-on-roulette/#post478515
To that extent he is, indeed asking a valid question.
But he needs to go back to the answers he had to that post. He simply does not have the evidence to support his hypothesis that he has found a biased wheel..
I Just want to try other ways to cow a 1% edge wheel. You could state there is no chance to win more than 1% in the long run
This thread, probably should belong to math section.
At This sub-forum "gambling with a edge" is suposed to write APs
Regards
You lost me at Pattern hunting.Quote: OnceDearI looked back through his posts. It's pretty clear that he is hoping to monetize his previous analysis of a few thousand spins where he went pattern hunting https://wizardofvegas.com/forum/questions-and-answers/math/23007-edge-on-roulette/#post478515
To that extent he is, indeed asking a valid question.
But he needs to go back to the answers he had to that post. He simply does not have the evidence to support his hypothesis that he has found a biased wheel..
I was just assuming he actually did have a 1% edge due to some other reasons, not a system where he only thinks he has a 1% edge.
Oh well.
Quote: Romes
EDIT: I found the "system"... results to follow... http://www.roulettestrategy.net/strategy/oscars-grind/
This oscar system introduced by a former and pioneer AP phd Allan Wilson caught his interest by 1965. He analyse many posibilite. He ended with a negativa expectation 1.46 for craps ando 2.7 forma french roulette.
Wilson madera a challengies to matemáticamente to build an algorithm
Wilson's challengies(this was oscar's system related) were taken by one of the Best mathematicians Stewart Ethier, who write a full chapter in a book named "finding the edge" edited by 3 other prestigious mathematicians.
Ethier answered another Wilson's challenge too, in 1981
Just to add more information about this realm
Regards
Ybot
Whomever suggested the Oscar System was not an AP nor a very good mathematician. The expectation is far worse than -2.7 for french roulette even. In each state you're betting 1-4 units, so the BEST you can do is state 1 where you're betting 1 unit and expect to lose the HE of 2.7. In the other states that expected loss is multiplied.Quote: ybotThis oscar system introduced by a former and pioneer AP phd Allan Wilson caught his interest by 1965. He analyse many posibilite. He ended with a negativa expectation 1.46 for craps ando 2.7 forma french roulette.
Wilson madera a challengies to matemáticamente to build an algorithm
Wilson's challengies(this was oscar's system related) were taken by one of the Best mathematicians Stewart Ethier, who write a full chapter in a book named "finding the edge" edited by 3 other prestigious mathematicians.
Ethier answered another Wilson's challenge too, in 1981
Just to add more information about this realm
Regards
Ybot
For oscar system in french roulette you Will lose 2.7% of total money wagged
We are not deciding wether we win or lose. I try to win more than 1% using oscar system. I am probably wrong
Math guys can know it.
Oscar's "system" is by far NOT the best way to maximize your winnings if you do have a 1% edge.
Quote: ybotRomes, we have got an advantage, it is a fact
Hmmmm. Define fact in this context.
Am I correct that you are referring to the system that you derived from your 10,000 spins recorded here?
https://wizardofvegas.com/forum/questions-and-answers/math/23007-edge-on-roulette/2/#post478643
You observed those spins and looked for a method of betting that you would have had you win in all of 20 subsets of those spins, and where you would ( had you bet a certain way, have made 1% profit?
Be told. 10,000 spins analysed till the cows come home would not tell you anything about the next 10,000 spins.
Now. here's a question for you. You analysed 10,000 spins as though it was 20 groups of 500. So now, were half of those spins clockwise and half anti-clockwise? If so, and if you retained that information, then try applying your system to two lots of 5,000 and compare clockwise with anticlockwise. One direction would have given you better results. Then split the analysis by whether they were recorded on an even number day or an odd number day. Odd, or even, one set will be better. Find other ways to split the observed spins, such as morning or evening observations, male or female croupier etc. In each case, one set will be better, for sure.
Then put all the facts together for a super refined system. You might find for example that the very best strategy would be to wager on a certain area of the wheel, in the morning on odd days, that were not wednesdays or fridays and where the ball went anticlockwise and the croupier was a female brunette.
Then test your derived best system with real money. You will tend to lose 1/37 or 1/38 depending on the game..
Even you MUST realise, that with unbalanced wheel with consistent croupier, that clockwise and anticlockwise will have very different ballistic properties. Just like different croupiers will show different characteristics.
When we look at the historical facts and measurements of the universe, then it's very easy to find apparent correlations. The harder you look, the more patterns you will see. But that doesn't means the correlations are facts. Regression analysis doesn't PROVE that relationships exist, especially when you are selecting subsets of old data and throwing away all the outcomes that were not favourable..
http://uk.businessinsider.com/real-maps-ridiculous-correlations-2014-11?r=US&IR=T
.... Unless of course, you can go back in time and bet on those previously recorded 10,000 spins. Best strategy then would be to put your life savings and all winnings on each next winning number in turn. But you would attract heat doing that and may need to duck back into your time machine.
The link you placed is referred to a wheel I studied some time ago.
All trials were counterclockwise.
We know that at 10k there is some positive fluctuation working, probably 1 to 1.5sd taking a section as a mass.
The more numbers involved, the closer to know the tru edge.
In that thread I tested a fixed way to play, no progresion or method, the chip value vary from 1 number to the other but does no change all the test. That was flatbetting. The math questions was to calculate the chance to duplicate another 20 successful sessions playing the same way.
I often split data randomnly to test what you stated.
We know that Clock and counterclock have different results
Math helps to determine the chance of rnd of certain event to accelate a decition or not.
regards
Quote: ybotOnceDear, I see you know this business.
The link you placed is referred to a wheel I studied some time ago.
All trials were counterclockwise.
I've done some analysis myself on an air wheel, where there was no random nature from croupier ( google air wheel roulette )
My analysis was based on looking for the angle between position of ball release to next landing slot. Eg looking for some tendency to have a bell shaped distribution of angles between start and end on a wheel where the ball tended to fairly reliably rotate between about 17 and 25 times, with wheel rotating about half of that in the opposite direction. I was hoping to find a sweet spot such as maybe a SLIGHT tendency to land 35 and a bit rotations away from the last number. I found no such tendency, or rather, there was a tendency for the sweet spot angle to never trend towards anything, just vary like crazy.
Quote: OnceDearI've done some analysis myself on an air wheel, where there was no random nature from croupier ( google air wheel roulette )
My analysis was based on looking for the angle between position of ball release to next landing slot. Eg looking for some tendency to have a bell shaped distribution of angles between start and end on a wheel where the ball tended to fairly reliably rotate between about 17 and 25 times, with wheel rotating about half of that in the opposite direction. I was hoping to find a sweet spot such as maybe a SLIGHT tendency to land 35 and a bit rotations away from the last number. I found no such tendency, or rather, there was a tendency for the sweet spot angle to never trend towards anything, just vary like crazy.
Looking for dealer´s(air ball) signature is a long task.
Imagine, VB players calculate desacelation and knee point when there are 2 to 4 revs left, you wanted to predict something harder.
Quote: ybotLooking for dealer´s(air ball) signature is a long task.
Imagine, VB players calculate desacelation and knee point when there are 2 to 4 revs left, you wanted to predict something harder.
I don't see it that way. I was hopeful that the signature of the airball would be vastly more consistent that any croupier's signature. Plus I could observe/record the airball 24/7 and that would just be one signature to deal with ( actually two: Clockwise and anti-clockwise). Change of live croupier would have meant switching in for a whole new set of variables every 30 mins or so. Finally, because the airball always simply started the roll from the last landing pocket and the result was machine readable, I was able to substantially automate my data collecting and analysis.
If you could find a game with 1% edge, then no betting system will increase nor decrease that edge. However, what you should be doing to grow your bankroll is to bet in proportion to it. If the bet were on an even money game, then I would recommend betting 1% of your bankroll every time. The greater the variance, the less your bet size should be.
Quote: WizardIf you could find a game with 1% edge, then no betting system will increase nor decrease that edge. However, what you should be doing to grow your bankroll is to bet in proportion to it. If the bet were on an even money game, then I would recommend betting 1% of your bankroll every time. The greater the variance, the less your bet size should be.
I suspect that the Op believes he has identified a section of the wheel, maybe relative to some start point, where there is an overall 1% advantage but where a smaller part of that section has a better edge. But of course, being a smaller section, it will have bigger variance. I reckon each of his flat bets would consist of small bets to the outside edge of his section and bigger bets towards the middle of his section.
Again, I may be well off the mark, but I suspect he wants to maximise profits without substantial or increasing risk of ruin, maybe concentrating his wager to the middle of his section.
Fixed percentage would be my way to go, but I'd favour a less greedy 1/2%
Actually, I suspect his wagers are far from even money bets, maybe concentrating on 1/4 of the wheel.
When I observed randomised results with fixed edge on a chart of half Kelly compared to full Kelly, the full Kelly scared me more.
In either case, he's going to need to play 10s of thousands of wagers to get anything like the full picture of a smoothly growing fortune. I'll dig out my workbook and attach it later.
At 1% of bankroll, rounding to the nearest dollar, he will be substantially flat-betting for the first several thousand wagers, anyway.
