What is this thing called standard deviation?

[john_solitude]

Some players asked me why the same systems or strategies can report so many different results?

The answer is a statistical phenomenon known as 'standard deviation'.

I will not go into the actual calculation of standard deviation (you can find a step-by-step way to calculate standard deviation in our free Roulette Fact and Fiction guide in the download section of our website nolinks://nolinks.john-solitude.be), but I will try to explain the main principal of standard deviation as simple as I can:

The best way to explain is to compare roulette with flipping a coin.

If you would flip a coin 10 times, you would expect on average you'll hit tails and heads 5 times each.

However, if each time we would flip a coin 10 times (which can NOT be compared with playing even chances on roulette because of the zero, which people tend to forget about) the result would be the average, it would be very easy to beat a coin flipping game or even roulette.

Standard deviation is a way to calculate how much the result can differ from the average.

Calculating this estimate (and this is exactly what it is: an ESTIMATE) on how much a series of outcomes (spins) can differ from the average, is the best way to keep a realistic view on what your chances are of pulling of a series of bets.

One would also be surprised how much a series of results can and will differ from the average,
without even getting statistically suspicious.

Why is it called 'standard' deviation? Because it's a natural phenomenon.

It is not something fishy, rare, suspicious or what so ever. Even if all chances are perfectly equal results will differ (deviate) from the average. In the long term the outcomes will hover around the average, but no one can predict beforehand when exactly this will occur...

But, the main problem remains: even when you know this, no one can predict beforehand on the spins you will be playing which way standard deviation will swing out (in the case we are playing a 'fair' balanced wheel on which all the chances are equal that is).

So, one could also say standard deviation is the reason why people report many different results using the same systems or strategies.

Hope this was clear.

[smiley=Santa001.gif]

[admin]

one could also say standard deviation is the reason why people report many different results using the same systems or strategies.

So my friend, a person could use the deviation from the "mean" to measure his/her own "luck" at current session, right?

When you talk about St.D, I think people may relate it better to "luck": a plus (+), neutral (=),  or minus (-) from the mean will mean good, regular and bad "luck", then require people to act accordingly.

Since we are into helping people, this St.D tool can help the punter to identify his/her own +/=/- period and maybe risk a bit more when it is a plus, holding it when equal, and betting the least or simply stop betting when the negative tram is too evident.

This way at least I think the punter will have a more realistic view on how his session is performing.

Also, this has to be applied to the "raw" hits, independent from the progression ( I think ), because progressions may distort the notion of being in a negative tram when you see the chips on your stack rising.

Well, those are my 2 cents. Hopefully we can continue the conversation.

Regards,
Victor

[john_solitude]

When you talk about St.D, I think people may relate it better to "luck": a plus (+), neutral (=),  or minus (-) from the mean will mean good, regular and bad "luck", then require people to act accordingly

Hey Victor,

Yes, I do agree:
people will experience the effect of St. D. as being lucky, unlucky or something in between.

Suppose, to keep it simple for the readers, one would be playing flat bet black, and red keeps coming up above high above the average, a punter would say he is unlucky, while a statistician would say St.D. was not in his favor (negative St.D on black).

Yes, progressions, do distort the notion of being in a negative trend.

The same example:
If one would be playing flat black, but this time a punter would play a Martingale, but red keeps coming up (or zero) one could still think his/her system is winning, while one was just lucky enough to not experience a St.D that was severe enough to wipe out the bankroll. For instance, on 10 trials, red appeared 8 times in a row (+ St.D on red), but the punter covered this with a progression.

Progressions are mostly used to make sure on a hit, one gains capital after a series of bets.
The downside is: the invested capital increases (slightly or severely, depending on the progression) on each spin, and the spread (difference between min and max bet) or the size of one's bankroll would not allow to invest so much money (keep raising the bet) to continually beat St.D.
If St.D. swings out severely against the player, the player would experience a severe wipe-out.

Flat bet's seam to be safer on first hand because the invested capital remains the same on each spin.
The downside is: on some occasions a hit, would not result in a gain of capital and one could be inclined to think: I'll just sit the negative trend out and flat bet until St.D. turns into my favor.

For instance: I'll play black, and even if red keeps coming up, I'll just keep on playing black until St.D swings in my favor (black starts coming up more than average, or a a return to a positive St.D on black) and the moment this happens I'll just quit.

This would be a rather reasonable approach, if only there was not zero or double zero making the chance for either black or red not 50/50.

So either way, playing black or red, with a progression or flat bet, the result would be negative if one would keep doing this.

Only the timespan would differ: on average the flat bet player would take more time to lose his bankroll.

But, you are right: flatbets would give the punter a better perception on identifying 'luck'.

The question I was asked time and time again: which one is better 'flatbet' or a 'progression' mainly depends on the temperament of the player and the risk one is prepared to take:  

- most flatbet players will play a lot of spins (they invest the bankroll over more spins, so either way the loss or the gain comes gradual: less fluctuation)

- progression players have a more agressive approach (they invest the capital on fewer spins, so either way the loss or the gain is more pronounced: more fluctuation)

Both styles of playing can however be deceiving in terms of win-loss:  
this is where the house-edge on roulette comes in.

For instance: if one flat bets and one would have St.D in his favor, one could still perceive this as
'I am winning, this system / strategy works', while this could only be having luck
(or a positive St.D. on the bets one has made).

More on the question, is flat better worse or better than using progressions can be read here:
nolinks://nolinks.john-solitude.be/questions.html#question_flat_betting

The thing for the readers to remember is, however:
the risk one takes on roulette is not rewarded fairly by the casino.

This is also best compared to flipping a coin to explain:

Everyone probably understands if one would flip a coin the chance is 50/50 for heads or tails coming up.
If someone however would propose a bet to you and say:
"If you bet either heads or tails and you bet a dollar and you get the bet right, I'll give you a reward of 0.95 cent for your dollar", one would say: "Hey, that's not fair, because I'll take more risk than the reward'.

However, it's the same principal at work on each spin of the wheel:
The risk one takes on whatever bet you can imagine, is in the negative of the player
(the house-edge differs depending on single or double zero, and what most literature does not mention are the obliged tips for the dealer, increasing the house-edge even further).

More on the question of the house-edge can be read here:
nolinks://nolinks.john-solitude.be/abc.html#house_edge

[admin]

I get that.

By the way, do you agree with Marigny's style of play? Waiting for a big swing and then bet for at least a small compensation to take place and get a win (i.e. a la "Lucky Strike").

[john_solitude]

Hey Victor,

To answer this question, I would first like to make a clear distinction between what is commonly known as 'gamblers fallacy' and statistics (otherwise this could start a debate).

On any game with equal chances (such as coin flipping or roulette) statistically outcomes will hover around the average.

HOWEVER, the trend itself can only be analysed in rearview.

For instance, let's suppose you take the time (as we did at the time with the JSWFA) to observe a large amount of spins (without betting that is) until you notice a severe St.D on a certain combination of numbers (be it a sector or any other bet) and you decide for yourself which entrypoint you'll take
(for instance sector X did not receive any hits for x amount of time) before betting:

In fact this is a bet, like Marigny, the ecart will straighten itself out ('catch up').

This method has some merits, but also it pitfalls:

The positive is:
- one would play far less spins, because there are less bet opportunities (because it takes time to collect a sample and analyze it > which is quite fun for people who are into numbers actually and it avoids that the player is psychologically sucked into the game, which happens if one bets every spins for hours on end > in fact the casino is hoping for this scenario to happen: they DO know what mindset players will go into)

- playing as less spins as possible on a negative expectancy game is always a good thing


The downside is:
- the wheel itself has no memory of previous spins, so one can only hope the natural tendency for a balanced game to bring outcomes around the average, will occur on the spins (bets) you are playing.

Suppose one would take 1 neg. St.D as a starting point, before betting, hoping this will turn into a + St. D (this would be a trend towards the average).

Well, at this point the outcomes can still go all the way up until 3 - St.D and this can not be predicted beforehand. In fact, it could even go further than 3 - St.D on a blue monday, but then one could start to think one is playing a biased wheel (to avoid players would easily come to this conclusion I hurry to add the very large majority of casino's have installed statistical monitoring themselves to avoid sophisticated players would take advantage of technical problem).

Waiting for 3 neg St.D on a certain chance, is a time stakingly procedure:
it requires collecting a large sample (can be literally hundreds or even thousands of spins, depending on the outcomes), but that is what it would take.

Than there are two possibilities when a - 3 St.D occurs:

1) It is improbable (but not impossible) the negative trend would continue, so players hoping the trend would start to swing mildly or severely towards the average, do have the correct presumption that any balanced game will in the long run produce outcomes that hover around the average (so one could 'profit' from the natural tendency of balanced outcomes).

HOWEVER it is underestimated how many spins this can take: can be ten, hundreds or even thousands, and the wheel itself will not suddenly decide to produce the exact opposite of what it has done before.

2) It is a biased wheel (in this case, playing the 'catch up' chances would be of no use, simply because the chances are not equal. However, the casino would normally be the first to notice, because they have a larger sample: so they can decide to close the wheel or simply install a new one.

Although a professional wheel does not come cheap, it's still far cheaper and a question of good economics for the casino to take good care of their equipment and regularly replace the wheels, because a sophisticated player knowing there is bias, could cause major financial damage.

So what is my opinion on Marigny style of playing?

- I favor any system or strategy that basically is advocating to play as less spins as possible (if it leads to less invested capital in a certain timespan) than systems or strategies, which would allow you to play for hours on end without no break (players are not aware of this themselves, but if you play for hours on end, one starts to get into a different mindset that can be compared to being under the influence of a drug > one's mind will just go numb).

- Basically if one decides to play roulette, it would be better to state:
hey, I've got x amount of money I'm willing to risk, and I will take the most rational approach possible -fully excepting one is playing a negative expectancy game- and if I win or lose, I'll get out after the series of bets I was prepared to take, rather than say: I will be betting for five hours straight without interruption (because normally people will tend to go to the cashier if their money runs out before they wished to leave)

I am in favor of system or strategies that have a tendency to bet that outcomes will hover around the mean, only one should still be aware that this can take a long time to happen.

One should never come to the conclusion the wheel has a memory because it hasn't, it couldn't care less. It's only a natural tendency occurring and it can take a long time before the chaos returns to balance.

 

[john_solitude]

-