### Author Topic: Does A Series Of Short Sessions Combine To Make A Long Series?  (Read 7566 times)

0 Members and 2 Guests are viewing this topic.

#### Reyth

• Global Moderator
• Hero Member
• Posts: 4174
• Thanked: 1436 times
##### Does A Series Of Short Sessions Combine To Make A Long Series?
« on: April 30, 2016, 08:16:53 PM »
Real on another place on the forum I have explained you in Jip and Janner language that special small events are together not a a large random row.!!!!!

So, I consider this knowledge to be quite important!

How is it that this difference can exist??

Why does this difference exist??

Is it true then that by playing small sessions we will get different results in the long term??

How do we divide between a small session and a long session?

What is the method whereby we can demonstrate this difference?
« Last Edit: April 30, 2016, 08:29:05 PM by Reyth »

#### Harryj

• Veteran Member
• Posts: 359
• Thanked: 174 times
• Gender:
##### Re: Does A Series Of Short Sessions Combine To Make A Long Series?
« Reply #1 on: May 01, 2016, 10:40:01 AM »
Reyth,
Any one who has carefully studied short sessions will come away with the concept that Random Chance tends to produce temporary bias. These apparent bias's exist only for a short time and then swing away. These Patterns or Trends, as they are called, create a situation that could be described as alternating results !
In the long term they pass un-noticed as a part of the normal Regression Toward the Mean. The result is that while the  long session performs as expected within the Law of Large Numbers the short session doesn't ! Instead they seem to cancel each other out.

Unfortunately the regression is rarely immediate, and often result in large deviations. So each short session MUST be considered on it's own merits. That is why long progressions, designed to take advantage of the swings between trends, often fail.

Harry

The following users thanked this post: Reyth, juice, Rinad

#### kav

• www.Roulette30.com
• Hero Member
• Posts: 2086
• Thanked: 1075 times
• Gender:
##### Re: Does A Series Of Short Sessions Combine To Make A Long Series?
« Reply #2 on: May 01, 2016, 10:59:54 AM »
I think one of the differences is this.

If you study sequences of 600 spins then you can make some research and come to some conclusions depending on the betting system. Much like I present the good and bad scenarios when describing systems.

However in a long sequence (million spins) the good and bad sequences are lost and always the result will be something like -2,7%.

For example like horus has posted in the betselection cc forum, Philip Koetsch in his book "Conquer the casinos" has tested various betting systems. He used one hundred 600 spin sequences for his test and came up with very interesting conclusions. If he just tested millions of spins sequences he couldn't reach any usefull conclusion. I quote horus post:

- - - - - - - -

He looks for what he considers to be the most useful information.
Namely:
• what are the absolute worst and best things that could theoretically happen?
• what were the worst and best things that actually did happen in 600 rounds?
• what were the average worst and best things that did happen?
• what typically happens if you don't get out while your ahead , but instead play the full round of 100 games?
• during a round, how often can you expect to get ahead by at least 10, 20 or 30 chips?
• in a 100 game round, how many times does your bankroll net-status typically reverse from losing to winning?
• how often does this net-status never reverse but remains losing throughout a 100 game round?
So all in all, it's a very detailed and interesting piece of analysis.

In the following tables we see the results for  3 dififferent methods of betting:
• Flat betting
• Linear betting: add 1 chip to the previous bet every time you win up to a total of 4 chips. You always revert back to 1 chip after a loss.
• Geometric betting. This is letting a winning bet ride. So let's go for x3 again. (1,2,4,8) Revert back to 1 unit on a loss.
See the VERY INTERESTING tables.

PS: I want to thank horus for this amazing post in betselection.
« Last Edit: December 28, 2016, 09:30:19 PM by kav »

The following users thanked this post: Reyth, Sheridan44

#### UK-21

• New
• Posts: 115
• Thanked: 39 times
• Gender:
##### Re: Does A Series Of Short Sessions Combine To Make A Long Series?
« Reply #3 on: May 01, 2016, 11:26:37 AM »
On a similar note, I've seen many web sites where the authors will tell you what you'll lose on average playing the wheel of doom (2.7% of the action), but miss the point that the average is unlikely to show. On my site somewhere I've populated a table that shows the worst case scenarios for playing an hour of roulette, at 80 spins per hour and covering various combinations of numbers. You could always do the sums to arrive at a best case scenario for winnings (all based on the Empirical Rule - )

#### dobbelsteen

• Hero Member
• Posts: 1548
• Thanked: 533 times
##### Re: Does A Series Of Short Sessions Combine To Make A Long Series?
« Reply #4 on: May 01, 2016, 12:00:41 PM »
A 600 spins sample of the ECs is a long run sample and for the dozens a large short run.

The following users thanked this post: juice

#### Reyth

• Global Moderator
• Hero Member
• Posts: 4174
• Thanked: 1436 times
##### Re: Does A Series Of Short Sessions Combine To Make A Long Series?
« Reply #5 on: May 01, 2016, 02:27:29 PM »
A 600 spins sample of the ECs is a long run sample and for the dozens a large short run.

Well an EC can be expected to go missing like 25 times whereas a dozen, 41 times...

25/41=.6097560*600=366

So is 366 spins a short run for an EC?

#### Sheridan44

• Mature Member
• Posts: 274
• Thanked: 244 times
• Gender:
##### Re: Does A Series Of Short Sessions Combine To Make A Long Series?
« Reply #6 on: May 01, 2016, 06:15:24 PM »
Thank you Kav for posting the Horus test results! Very useful and informative. Could possibly have saved me hours and hours of work.

The following users thanked this post: kav

#### Trilobite

• Mature Member
• Posts: 370
• Thanked: 142 times
##### Re: Does A Series Of Short Sessions Combine To Make A Long Series?
« Reply #7 on: May 02, 2016, 03:53:38 AM »
Does A Series Of Short Sessions Combine To Make A Long Series?

Of course it does.

#### kav

• www.Roulette30.com
• Hero Member
• Posts: 2086
• Thanked: 1075 times
• Gender:
##### Re: Does A Series Of Short Sessions Combine To Make A Long Series?
« Reply #8 on: May 02, 2016, 07:39:17 AM »
Does A Series Of Short Sessions Combine To Make A Long Series?
Of course it does.
I agree to this.
Theoretically any combination of (really) random spins will produce a really random sequence.
For example the following spin series
• 1000 consecutive spins of the same wheel
• every 3rd spin of the same wheel for 3000 consecutive spins
• if all night we walk withing the casino and record randomly 10 spins from different wheels and then add them together to create a 1000 spin sequence
All the above methods would create an equally random, thus "same for all intends and purposes" sequence of 1000 spins. Any random combination of random short sequences will form a valid, random spin sequence. So theoretically Trilobite is right.

However what I'm saying, is that the player cares for the short term. He can and will play for the short term. And we can come to interesting conclusions by studying the short term. The long term is like eternity. it may very well exist but we only live so long in this earth. Or like Keynes said "in the long term we are all dead".

I don't know what dobbel means by "special small events are together not a a large random row". Maybe he means that if we combine sequences that are not random ("special") we can not create a long random sequence. This is also true. In order to create a long random series we should only combine totally random events.

Anyway the concept of various random events combined form a perfectly random series is very important and has many implications. For example from this concept follows that:
• If you combine all your played spins, no matter where and when you have played them you can create a long totally random series.
• If you plan to bet 7 times on Black, it absolutely the same either if you bet on the same table or on different tables.
« Last Edit: May 02, 2016, 07:47:56 AM by kav »

#### Jesper

• Hero Member
• Posts: 1425
• Thanked: 738 times
• Gender:
##### Re: Does A Series Of Short Sessions Combine To Make A Long Series?
« Reply #9 on: May 02, 2016, 12:23:25 PM »
Yes short runs added to many short runs look the same as a long. Take a few long which is "glued" togeher, and you can not see any difference, and not calculate a diff at all. Personally I don't give a sh... study long runs, it gives statistical information, and that can be achieved using what has been done in the field since dozens of decades.
« Last Edit: May 02, 2016, 12:33:19 PM by Jesper »

#### dobbelsteen

• Hero Member
• Posts: 1548
• Thanked: 533 times
##### Re: Does A Series Of Short Sessions Combine To Make A Long Series?
« Reply #10 on: May 02, 2016, 08:17:33 PM »
Short run samples have always an imbalance. Samples with the same imbalance can not be together random.
The imbalance makes the sample special.

#### Reyth

• Global Moderator
• Hero Member
• Posts: 4174
• Thanked: 1436 times
##### Re: Does A Series Of Short Sessions Combine To Make A Long Series?
« Reply #11 on: May 02, 2016, 08:40:51 PM »
Short run samples have always an imbalance.

Quote
Samples with the same imbalance can not be together random.

Ok, now are you saying that a short session cannot be duplicated by another short session?  But what about ESSENTIALLY the same imbalance; in other words like still 10 misses on two dozen but an entirely different ordering/placement of those 10 misses?

Quote
The imbalance makes the sample special.

Even though the specific ordering is unique to that sample, couldn't a series of short runs combine to make ESSENTIALLY one long run sample (even though it will be unique it will have the same crucial characteristics)?

If I run 16M EC spins, I get 25 consecutive misses as the maximum.  I don't get that in only 100 spins because the odds of getting that result is greater than 1 in 16M and only 100 spins doesn't cut into those rare chances very much.  However, if I run 100 spins 160,000 times, are you saying...

Aha, so this come down to "how many spins will I actually play in my lifetime"?  So does this mean we should only be calculating odds based on our lifetime expected number of spins which represents the true odds for us?
« Last Edit: May 02, 2016, 08:47:41 PM by Reyth »

The following users thanked this post: kav, december

#### dobbelsteen

• Hero Member
• Posts: 1548
• Thanked: 533 times
##### Re: Does A Series Of Short Sessions Combine To Make A Long Series?
« Reply #12 on: May 03, 2016, 08:26:54 AM »
A short run sample of 10 numbers has a probability of once in 512 spins. Very small samples can repeat.

The same samples (sequences) of more than 20 figures are nearly impossible.

There is a large difference between playing roulette and discussing long run sequences.

In the casino I play roulette and my strategy is based on a very large research of the features of the small random sequences. Long run theory can not be a help to become a succesfull player.

The following users thanked this post: kav

#### scepticus

• Hero Member
• Posts: 2401
• Thanked: 521 times
##### Re: Does A Series Of Short Sessions Combine To Make A Long Series?
« Reply #13 on: May 03, 2016, 07:32:08 PM »
Yes, Reyth, you did have a Lightbulb moment. Dobbelsteen is right . We will never be able to bet the number of spins our detractors claim to “prove” that we are certain to lose  our bankroll .We live and bet in the REAL world , not in a maths textbook.
Where I don’t agree with Dobbelsteen - and most others in the forum - is the  need  to analyse lots of numbers to arrive at a reasonable bet selection. I do agree with your view that Bet Selection is the key ingredient in gambling .

The following users thanked this post: Reyth

#### Harryj

• Veteran Member
• Posts: 359
• Thanked: 174 times
• Gender:
##### Re: Does A Series Of Short Sessions Combine To Make A Long Series?
« Reply #14 on: May 03, 2016, 08:14:17 PM »
THE PROBLEM HERE IS THAT WE ARE NOT ALL ON THE SAME PAGE !

Obviously a series of short trials will ultimately equal a long trial, And the "Law of Large Numbers" will apply. That does not mean that each short trial will produce the equality that we expect from the 'Law" This means that eventually the short trial must form some sort of balance, and cancel each other out "in the long run".
This does NOT mean that  The regression will be immediate, but if you join enough short biased trials together to form a long trial a percentage balance will occur. This satisfies the mathematician, but not the gambler, who is locked into short trials.
So the gambler plays each short trial with the expectation that there will normally be a degree of imbalance, or bias. Thus he alters his target based on an "Empirical bet selection". The mathematician on the other hand assumes that the same bet is made ALL THE TIME !  Thus the accumulated HE will deny any chance of profit.

Much of couse depends on the length of short and long trials. Long trials are generally many thousands or even millions of spins. Short sessions on the other hand rarely exceed 100-200 spins.  Even more significant those short sessions are divided into many  bets or games. Each of which could, and usually does, use a different target !

Hence the need for "betting indicators" AKA "triggers" !

Harry

The following users thanked this post: december, Reyth