An effort to help some members get out of the worst gambing fallacy

[TwoCatSam]

OK

Are we saying the numbers produced by the wheel must--must--conform to statistics?  Either they do or don't.

Let's say they do and use an exaggerample.  We just had 100 reds hit in a row.  At some point blacks must catch up.  No?

Assuming all of the above is correct________________________

What force in the universe makes black more likely to hit?  What force in the universe says red has hit too many times and black must catch up? 

R.D. Ellison called it "statistical pressure" and hinted  that some force in the universe causes this equalization.  And how does it accomplish this goal?  Shrink red pockets? 

Samster

[iggiv]

there are only LONG TERM statistics that matter like that, not short term. You can't use short term statistics and expect to win. It is gonna be  fallacy. Classical fallacy.  That's why roulette is so hard to beat. Say in the last 40 spins we had 30 blacks (that's more realistic than 100 blacks in a row scenario). We can't bet that that very soon we will see more reds. But say within last 10000 spins it is likely that roughly 47% of them will be red and 47% black. It can be 46, 45 or 48%, but it is gonna be pretty close to 47%. This is for  single zero roulette of course.

So this CAN'T be used for gambling. And casinos know this very well.

and if we take millions of spins then these numbers would be even more close to 47.3%. These are long term statistics.

[albertojonas]

THIS stuff has been said over and over again:


after those 30 reds the thing one can do is to try and catch the trend of correction. Once Blacks start to show up clearly, try and catch that correction. This is not a Fallacy.
Cheers
AL


How about the graphic bellow for a Fallacy???

[iggiv]

u can try to catch any trend yes. The fallacy is not about trends. It is about "high probability" of
something that DID NOT SHOW up. Trend is something else.

[Ralph]

There is a law of large numbers. Even if every single event should be  much different in reasons for happen like a traffic accident of deadly outcome, the numbers are ruffly the same from one year to an other. If the number changes, something more than trivial has happen, in many areas like car construction, alcohol behaviour, or road construction, or legal reforming.

Every insurance company use this. There are en expected number of outcomes, which are predictable.  Even rare events as lethal lighten event, is in my contry not more than 1-3 a year and three is  rare, as zero is.

[Bayes]

OK

Are we saying the numbers produced by the wheel must--must--conform to statistics?  Either they do or don't.

Yes, they must (given an unbiased wheel).

Let's say they do and use an exaggerample.  We just had 100 reds hit in a row.  At some point blacks must catch up.  No?

Blacks have to catch up insofar as they must conform to statistics, but it doesn't mean that they have to catch up or equalize in the short term, that's GF. That doesn't mean that you can't use past spins and statistics to good effect though, because a sequence of spins doesn't just consist of R/B, but other ratios, such as the number of streaks vs chops, the number of series of 2 vs higher series etc. There are averages within averages, and tracking these and knowing the statistics and relations between them is more effective than just betting that black will catch up.

Assuming all of the above is correct________________________

What force in the universe makes black more likely to hit?  What force in the universe says red has hit too many times and black must catch up? 

Knowing the forces won't help you, it's enough to know that it happens and that you can rely on it.

R.D. Ellison called it "statistical pressure" and hinted  that some force in the universe causes this equalization.  And how does it accomplish this goal?  Shrink red pockets? 

Equalization occurs because of symmetry in the device. The tendency is for the outcomes to equalize because there's no reason for them not to.

Samster

[Bayes]

there are only LONG TERM statistics that matter like that, not short term. You can't use short term statistics and expect to win. It is gonna be  fallacy. Classical fallacy.  That's why roulette is so hard to beat. Say in the last 40 spins we had 30 blacks (that's more realistic than 100 blacks in a row scenario). We can't bet that that very soon we will see more reds.

You don't need to take millions of spins in order to see an approximate balance. It's enough to get a balance within certain boundaries or deviations in order to get an advantage. Take your example of 30 blacks in the last 40 spins, which is a STD of a little over 3. According to probability, the next 40 spins will generate approximate 20 reds, and in fact in ANY sequence of 40 spins this is the most likely event. 10 reds in 40 spins is a rare event, but it doesn't mean that the next 40 spins will produce 30 reds and 10 blacks, giving a balance over the 80 spins of 40 reds and 40 blacks, that's GF.

The point of using a deviation as a trigger is that by definition, rare events don't occur very often, and a deviation over a longer sequence is rarer than a deviation over a shorter sequence, so you can reasonably expect that the deviation will not be so severe (in the same direction) as it was on the first event. Suppose that in the 2nd set of 40 spins, the deviation continued, and again you only got 10 reds. The STD is now about 4.4 over the past 80 spins. You can now be even more confident that the NEXT sample of 40 spins will produce more than 10 reds. If it didn't the STD over the last 120 spins would be 5.5, (30 reds in the last 120 spins), which is practically unheard of.

Note that not only is a strong deviation unlikely to be repeated in successive samples, but a strong deviation is more likely to occur following an AVERAGE sample, which is just logical, if you think about it. This means that, more often than not, you're better off looking at past spins instead of just betting randomly, ie; past spins can indicate (but not cause) future spins.

[Bayes]

Like in chess, pushing a pawn or controlling the centre may be fundamentally correct actions....but not in the particular position you see before you.

Exactly. Simple mechanical systems don't work because they rely on the outcomes conforming to a certain pattern, but the nature of random is to change continuously, so these kinds of systems are at the mercy of every pattern which doesn't match that which the system is designed to beat.

[Bayes]

You can try to catch any trend yes. The fallacy is not about trends. It is about "high probability" of
something that DID NOT SHOW up. Trend is something else.

The way I see it is that trends and sleepers are two sides of the same coin. Both are deviations. Suppose the STD of a certain trend is 0, +3.00, or -3.00, which would you choose as your trigger for starting to bet on the trend (when it emerges)?

[albertojonas]

The way I see it is that trends and sleepers are two sides of the same coin. Both are deviations. Suppose the STD of a certain trend is 0, +3.00, or -3.00, which would you choose as your trigger for starting to bet on the trend (when it emerges)?

...may consider that the end of the trend is when sleepers correct.

[iggiv]

The way I see it is that trends and sleepers are two sides of the same coin. Both are deviations. Suppose the STD of a certain trend is 0, +3.00, or -3.00, which would you choose as your trigger for starting to bet on the trend (when it emerges)?

well, classical Gambler Fallacy (like Wikipedia tells us)  is about expecting something that DID NOT SHOW UP. Something about the "balance". Trend is something different.

about other fallacies -- these are already speculations. It is easy to call this or that "a gambler fallacy". But i think we should use well known definitions, which are recognized world-wide.
I remember myself some time ago -- anything i did not agree with-- i called "a gambler fallacy".
But that's WRONG. After reading some material on the subject i realized it. There is only one well-known Gambler Fallacy.

[Ralph]

An other fallacy is the wheel is rigged, because you lose.
And you are clever if you win.

[Robeenhuut]

Unfortunately capitalizing on "due" swings in SD (i dont like STD abbr. ) is far from a sure bet due to a fact that its impossible to predict when its going to happen and for how long. But i would take it anyway. If i could get few heavily imbalanced events at the same time i would take my chances with a positive progressions on them.

[Ralph]

If a trend is more likely to come, a positive progression is both cheaper and more profitable!
But it is to know its due.

[iggiv]

An other fallacy is the wheel is rigged, because you lose.
And you are clever if you win.

this is not about classical fallacy  already. Wheel could be basically rigged and u may be clever if u win. Though in most cases it is like u say --wrong conclusions. But again -- this is not a well-known GAMBLER FALLACY. Another way of classical Gambler Fallacy (but not in gambling)  for example is to expect a son if your wife gave birth to 4 daughters already.