I don't want to shoot this down straightaway; from a personal viewpoint it might well be a worthwhile investigation, but the singles and series may have you going around in cirlces for a very long time.

It's an old concept known as Marigny de Grilleau. It's the only system-esque method I personally ever really put through the wringer. Predictably it all came out very even in the end.

Though I should say bet selections of this nature are still pretty popular when combined with a soft progression and can bring some success. They can go for a long time, compared to most systems which fail even within a few thousand bets.

These methods haave also been well-documented at the old VLS by the member Lucky Strike. He is a smart guy who spent a lot of time exploring different betting rules including regression with "ecart". Do a search there for Marigny de Grilleau I'm sure you'll find a lot of material.

Through research, in the end you will find that no matter how you look at the ECs, everything comes down to 50/50 random chance. The simple fact is, a single has a 50% chance of staying a single and a 50% chance of becoming a series of 2.

Any series has a 50% chance of continuing or breaking. There is no deciphering when or how these events will happen. Over the longer term you'll find the "law of series" in any sample for example: 1000 singles, 500 series of 2, 250 series of 3, 125 series of 4, 112 series of 5, 66 series of 6 etc. The existence of this tells us straight off that it's not exploitable.

Moreover, distribution of the singles and series are subject to the same fluctuation any EC bet is. The bet selections are basically FTL or OTL wrapped up in a huge amount of complexity. So, why not just pick randomly?

When applying regression to look for potential dominances this is also difficult, since regression can happen over hundreds or even thousands of spins, you could well go broke before making any profit. My opinion is you can't really get anywhere without empirical evidence, I think everything should be combined with conditional probabilities.

If you can apply that to singles or series it may be much easier to identify the most effective methods of betting, but this just makes it even more complicated. But, on the other hand, at least this method is centred on events that happen, rather than events that are rare.


[mod] I've spaced this for easier reading than a solid block of text. [/mod]

I don't know if it's for academic reasons or if you think you have an edge, but it's worth noting a few points anyway.

Firstly, it's impossible to gain an advantage with a roulette system, old news I know.

The system would obviously have to be simulated with flat bets, or even without any betting, as long as you end up knowing the overall strike rate. A strike rate above the expected may suggest an edge. After that you can use the z-score to find out the probability of whether the results are due to luck or something else. While the house edge would simply undermine the premise behind any bet selection, it's feasible to think you don't need an edge to win consistently, which gives hope for a holy grail system. With a long simulation you can test for a few vital bits of data (you can't get these testing real play small sessions). Mainly you can ascertain, officially or not, the variance and volatility. There are formulas for that, but they may produce a number that can't be interpreted. You can look at the variance as how far your strike rate differs from the expected; you can look at losing runs and how the strike rate dips at those moments it performs badly. And the volatility is how that variance would affect your bankroll; you can discover the max drawdown and tell whether the system is economical and what kind of bankroll you need. In short you could gauge variance and volatility without calculating it and still tell if your system has any kind of merit, or if you're going to go bust as soon as you start playing. If the variance is low, there's no reason not to think you can't play through it with some kind of grinding progression because, having tested it thoroughly, you would know that it's likely to recover quite quickly. Contrast that with not testing a system properly, most players don't know the volatility and go bust in their first serious drawdown.

For roulette it's not so complicated.

The EV for an even chance is (chance of winning v chance of losing), where 1 is your stake.

(18/37 * 1) + (19/37 * - 1) = 18/37 - 19/37 = -0.027 * 100 = -2.70%

Any odds in single zero roulette will return -0.027 EV.

--

To test your own bet selection, you first need to calculate your probability of winning a single bet by simulating it across an adequate sample size, ie 10,000 bets.
You can then calculate the expected value suggested by the results. If there is an edge, your probability of winning an even chance would be higher. You can simplify to this:

(0.51 - 0.49) = 0.02 * 100 = 2%

Again that is chance of winning v chance of losing.

--

Calculating the EV for sports is a bit more complicated.
The official way does involve factoring in your potential winnings and losses. I'll quote from a useful article.

Quote

The formula for calculating Expected Value is relatively easy – simply multiply your probability of winning with the amount you could win per bet, and subtract the probability of losing multiplied by the amount lost per bet:
(Probability of Wining) x (Amount Won per Bet) – (Probability of Losing) x (Amount Lost per Bet)
To calculate the expected value for sports betting, you can fill in the above formula with decimals odds with a few calculations:
1. Find the decimal odds for each outcome (win, lose, draw)
2. Calculate the potential winnings for each outcome by multiplying your stake by the decimal, and then subtract the stake.
3. Divide 1 by the odds of an outcome to calculate the probability of that outcome
4. Substitute this information into the above formula.
For example, when Manchester United (1.263) play Wigan (13.500), with a draw at 6.500, a bet of $10 on Wigan to win would provide potential winnings of $125, with the probability of that happening at 0.074 or 7.4%.
The probability of this outcome not occurring is the sum of Man Utd and a draw, or 0.792 + 0.154 = 0.946. The amount lost per bet is the initial wager – $10. Therefore the complete formula looks like:
(0.074 x $125) – (0.946 x $10) = -$0.20
The EV is negative for this bet, suggesting that you will lose an average of $0.20 for every $10 staked.

Quote from: Archie on August 15, 2013, 06:14:29 PM
The internet is a very poor place to seek out actual expertise. 

That is according to your opinion. Besides, the value of information is subjective to the reader, depending on your level of intelligence ie if you can understand it, and whether it is actually relevant to your cause. Of course, some information is not accurate; that is not unbelievable, there are literary, conceptual and factual mistakes in all places. In a way I also happen to see your point (in the context of casino games) - everyone can claim to be an expert without actually proving anything.

Quote from: Archie on August 16, 2013, 01:41:18 PM
Perhaps, you mean "dumbed up", ie, put into a form in which pseudo rocket scientists can "make sense" of it?

Your example still does not really address why proportional betting would not be mathematically logical when you have an advantage. Moreover losing 12 or 15 bets in a row seems more applicable to a -EV casino game as oppose to sports betting, which this thread is meant to be about. It also doesn't represent a sound investment of time or money. It would be wise to test for variance of the bet and also its volatility, with that in mind the individual would know what losing streaks to expect and how much cash was needed to get through them. As a sidenote, that is why long simulations of betting systems are important, and testing "real" play samples aren't. A 5% stake, then, might prove to be too high, though 5% as a maximum is generally recommended for sports. Up to that point, entering a game without that information appears ignorant, and even with an advantage it would still be possible to lose. It's possible to contrive a statistical system for sports betting that offers a long term advantage, with the right information you can create your own probabilistic odds and then proceed to find trades which are misvalued. Back testing the selection method against previous results would enable you to fine tune your formula until it generates favourable results. According to the EV, proportional betting would absolutely be the best staking plan.

That's the whole thing with trading odds, isn't it?
What they don't tell you is that you still need the insight to find a trade that is misvalued for your particular needs.

It's not a new concept at all and while it's sound enough, the execution might be something else.

Quote from: Archie on August 15, 2013, 06:14:29 PM
Now he's loses very close to half his life's savings.

You have dumbed your example down to such a level I'm surprised you even think it's still valid.
The old timer is just another ignorant gambler playing a game blind, it's not surprising he's about to lose his entire life savings.
In this case proportional betting isn't going to save him, but neither is anything else for that matter.
Quitting, maybe, before being entirely broke would be the best course of action.

Personally I have no problem with the term baloney, or other such forms of rebuke.

But I am not sure if there is a purposeful conflict of interests going on, whether it's simply argumentative or otherwise. For example proportional betting with a +EV is entirely logical and the most effective method of growing an investment. There's really no argument against that case.

Quote from: Archie on August 15, 2013, 01:43:09 PM
Not a single one has satisfactorily answered my specific statistically related questions about what it is that they claim to do.

The age old question abounds: Why would they tell you? I suspect you think they are full of it since they dodge your questions. Most probably are, there are several threads on this and other boards that would testify to that. Betfair charges fees on transactions of course, but that doesn't mean the odds still aren't much more favourable than a bookies. It also commissions the use of "official" software for scalping trades. You can watch video demos of any of them and see with your very own eyes how the concept works. Not that it's as simple as abc. For the record my initial posts weren't anything to do with arbitrage, but betting systems that produce an accurate degree of probability through statistics. There are a number of methods this can be done, they are documented online in various places. Don't ask me where, I am not willing to find them for you. 

What is baloney to you? That gambling winnings in the UK are tax free? That there are professional sports bettors, or that there are systems that exists which have stood up to rigorous back testing and yield a positive EV for years and years? People agree on that, and the maths supports it. There are probably hundreds of professional betfair traders, maybe thousands, and it's important to understand that these people bet against each other, not the house, therefore in most cases the odds are very favourable. Before you start disgorging bilge about quantum mechanics, multiverses and strange worlds at the Earth's core and at the top of giant beanstalks, bear in mind it is possible to predict with fair accuracy, even without a system, the outcomes of football matches. There is so much capital involved, with a reasonable risk, that the probability of ruin (though it exists), is a mere blip that allows for a huge degree of confidence not available in any other activity. There are some betfair trades available called scalping, which is a form of arbitrage, that allow for profits to be made on an trade regardless of the outcome.

The only real point it needs to cover is that the EV has to be positive.
If it is, Kelly works, it is entirely logical. You can find the formula anywhere.

Not that I am against books or anything...

It is also worth noting that it is an investment strategy, as oppose to a progression ie it does not work for negative expectation bets.

You can make an excellent tax-free living sports betting in the UK, especially on Betfair where the odds tend to be better than a bookies.
I only know one professional gambler but he's been doing it for more than 10 years and is very well off. His risk of ruin is now practically zero since his system is totally bulletproof.
He bets on Premier League football, it is really quite predictable. And with 760 games per season, there is more than enough opportunity.

Personally I think knowledge of the game is essential. Statistically, it probably isn't, if you don't care about the emotional side of betting.

With a system that discloses accurate predictions you can gain a +EV from your bets, and use something similar to the Kelly Criterion to calculate the optimum wagering strategy. Kelly does have its critics and it can suggest to place large bets, some prefer to use a half or even quarter Kelly. I would certainly not use any progression in the mould of the Martingale, the risk of ruin is great and requires a large bankroll. At least with Kelly, you only really bet what you can afford to at that time. There is also the mental distress of losing a series of Martingale bets. If you gamble properly you shouldn't even need that anyway. Weddings is correct in saying that bets should be proportional to your BR, which the Kelly Criterion teaches. If a system is tested and proven, it will work long term.

What's the world coming to when the Martingale gets 38 pages of replys?
You can play this faster on every spin by following the last or betting the opposite.
But that would be archaic and ill-advised.

Quote from: spike on August 10, 2013, 07:46:15 PM
Its all about variance. Small edge = large variance.

True that, but variance as the norm for your ECs is low.
So, mathematically, a player should not need an edge to be successful with a betting system, even if the EV is negative.
Casinos can still be beaten from time to time by lucky players due to the volatility of the games. The opposite applies too but the casinos have more money and know how much they need to cover their payouts. Ultimately the volatility kills any progression or mechanical flat bet. Logically then an effective bet selection should be based on the player's confidence in the next series of spins. But can you ever be more confident than 50/50? We'll have to wait until Gizmotron releases his memoirs to find out...

Quote from: Bayes on August 10, 2013, 05:24:29 AM

How does the wheel "know" if any player is wagering real money or not?

It doesn't need to though does it? In the end you still supposedly lose 2.7% of your investment.
If you sit out five spins and see BBBBB then bet R, the probability of winning is still 48.7%, not perceived 98.7% (ie the gambler's fallacy).
However if you had bet all spins, and say were at -5,  by the fifth black the probability of winning with your red would be 98.7%.
Though the odds are still 18/37, always offering negative value. Apologies for the crude example, but that is just about as simple as it gets.
It's proved through all long-term system testing, that only actual bets are relevant.

BTW, I think it is possible some kind of tracking is of worth, but not in the way you would typically wait for some kind of set condition.

Quote from: Priyanka on August 09, 2013, 04:14:55 PM
Is it considered to be yielding a higher payoff to variance ratio compared to placing 3 units on EC?

I wonder, judging by the first question in that post....do you mean, does betting multiple locations affect overall variance, or is each location affected individually? Or can you reduce variance by using some kind of secondary bet? Not really, overall you are covering 21 numbers in your example, to which variance applies. BTW, in my opinion it's best to choose a bet type, maybe ECs or dozens and master their use. These tend to be easier to play than inside numbers or streets etc.

Quote from: Turner on August 09, 2013, 10:14:39 PM

What should we be doing then?...in your opinion

Sports betting. [smiley]aes/wink.png[/smiley]

Quote from: Archie on August 09, 2013, 01:53:08 PM
By definition, each cause has its effect (injective or not). But that doesn't define reality.
 

"... they are all defeated by fluctuation." - Variance is secondary to the mean.


"... with a low enough variance to allow the safe use of progressions." - If you can't alter the mean, then what of the variance.

I don't believe I provided a definition of reality? Each cause has its effect, is it not on a physical level...what's the point?
Who is altering anything? Maybe it is not wise to take the literal definition of an extraneous variable. Perhaps we should just say it's wise to consider all factors, even the unwanted ones?

Priyanka,

There is a great mythology about techniques such as dummy bets, virtual tracking, entry points and all the rest of it. It is actually quite an interesting topic; simple yet complicated at the same time. For many, it can also be hard to accept that virtual bets hold no merit at all and are, in fact, a total waste of time. So, the side of you that says it's a useless effort it correct, it's not just opinion, it could easily by proved (and has been time and again). I'll try to explain why.

The simple fact is, there is no benefit in tracking. It doesn't change anything. A good example of this is to take a system similar to the use of statistical ecart, which I had tested a lot a few years ago. The actual system doesn't matter but I'll simplify it by saying, it's a system that more or less translates as waiting for a certain standard deviation in even chance bets and then entering the game to bet for a statistical balance. This might seem logical but it's based in a fallacy and it's really just a posh way of looking for a condition like: in any 60 spins, if any even chance has hit less than 15 times, begin betting that even chance from spin 61 until you have gained 5 units. In that sequence you would encounter an SD of at least more than 3.5. Anything above 3.0 is considered a rare event. Now, if you indeed entered the game at that point and began betting, 50% of the time the SD would increase to 4.0 or 4.5, maybe even 5.0. In that case you lose. The other 50% the SD will decrease to 3.0, or 2.5, or 2.0, in which case you would hit your target. If an SD of 3.0 is rare, then how can it possibly increase half of the time (and it is, exactly, half of the time)? The answer is, the maths of the game only applies to actual bets on which you wager real money. That is the reality, which people tend to ignore.

So, when you enter the game at a perceived 3.5 SD, it is a false SD, and the real SD is still at 0-1, within the realms of normality. Then you find that the probability only begins to change when you begin to bet, hence you win 50% of the time, and lose the other 50% of the time. Also bear in mind that when you enter a game with a particular system, you are in it for life. You can't reset it by playing short sessions or by playing in different casinos. In fact, you can't even reset it by playing a different system or two systems at the same time. The variance just follows you around forever, it doesn't know where you are playing or for how long. This is related to an interesting concept known as the personal permanence.