The Law of the Third

[Bayes]

Next article: The Law of the Third. 

update: there was a chart missing - should be fixed now.
 

[Reyth]

Wow.  Very nice!

[Jesper]

I spend some times on the law of thirds, still I do not know if we can have any advantages. On Dublinbet I use to put a chip as the numbers come, and I did very long play, the sessions gave a small plus, hundreds of times.

If we know the last number delay over 150 spins we could use it and do well, but it is not the case all the time.

I have been trying progressions on many numbers, add as they come, and it is winnings in near 100% of the times, the backside is a bad run will cost a lot, and stop-loss is not working, as that means many winnings spoiled as well.

But if we add all small winning together, and a subtract the loss, it is either a loss or a plus, which is not very large, especially on NOZ.   Using 0.1 it is quite easy to make 100 Euro in a few hours, and repeatedly, we can do it if we not care a loss of 1000 which can happen. I do not know the exactly odds for losing 1000 before winning it, but it must be close to equal.

[Reyth]

Here is a guy using the Law of the Third:

The system consist in playing 37 spins. After 37 spins we should have seen only 25 of the 37 different roulette numbers, so there will be about 12 repetitions.

SPIN1 - 1 NUMBER
SPIN2 - 2 DIFFERENT NUMBERS
SPIN3 - 3 DIFFERENT NUMBERS
SPIN4 - 4 NUMBERS
SPIN5 - 5 NUMBERS
SPIN6 - 6 NUMBERS
SPIN7 - 7 NUMBERS
SPIN8 - 8 DIFFERENT NUMBERS
SPIN9 - 8 DIFFERENT NUMBERS, SO 1 REPETITION
SPIN10 - 9 DIFFERENT NUMBERS
SPIN11 - 10 NUMBERS
SPIN12 - 11 NUMBERS
SPIN13 - 11 NUMBERS
SPIN 14 - 12 NUMBERS
SPIN15 - 13 NUMBERS
SPIN16 - 13 NUMBERS
SPIN17 - 14 NUMBERS
SPIN18 - 15 NUMBERS
SPIN19 - 15 NUMBERS
SPIN20 - 16 NUMBERS
SPIN21 - 16 NUMBERS
SPIN22 - 17 NUMBERS
SPIN23 - 18 NUMBERS
SPIN24 - 18 NUMBERS
SPIN25 - 19 NUMBERS
SPIN26 - 19 NUMBERS
SPIN 27 - 20 NUMBERS
SPIN28 - 20 NUMBERS
SPIN29 - 20 NUMBERS
SPIN30 - 21 NUMBERS
SPIN31 - 21 NUMBERS
SPIN32 - 21 NUMBERS
SPIN33 - 22 NUMBERS
SPIN34 - 23 NUMBERS
SPIN35 - 24 NUMBERS
SPIN36 - 24 NUMBERS
SPIN37 - 25 NUMBERS

He waits 8 spins and bets the last 8 numbers that have come out.  He will continue to bet the numbers that come out UNTIL there are more numbers that have come out than in the above list AND THEN he will bet the OPPOSITE of the numbers that have come out.

[Jesper]

This is played on an AM-Wheel
paying 38 on a straight.

First spin on LOW to get a number. #18.
#23,#8,#18.
  33 in four spins.
---------------------------
One on #18.
#29,#12,#19,#6,#36,#0,#29.
up to   43 in total 11 spins.

We can get a lot sessions hitting
in the first 8 spins (I think most
of them).
Sometimes the first hit comes much later
or the repeaters are just too few.
At the end they must come, but a bad session
can end with a loss of 500 when all numbers are hitted.
The most sleepers have to delay until
we win, and the cost of a losing spin
becomes higher and higher, the payout
decrease.

Run it flat on a fair wheel is a break
even play, not only in theory, it shows
up in the play after a couple of 100
trials.

We win 10 to 100 most of the trials,
we win most of the sessions, but it
needs just a bad one to lose it all.
A stop loss would not help, as losing
sessions becomes more often.

A property of many betting plans.

I have got better result stopping
adding numbers after max ten.
Most sessions have a repeat before
10 numbers, and use a progression up an unit
every spin, and reduce 37 units
on the total on a hit. Even while
reducing chips reducing numbers.

Getting too many numbers in play
is sure losing on a unfair wheel and
do not win on a fair.

[spins]

pick 20 numbers, to keep it simple I am betting on 1- 20    bet for 20 spins or until you have a profit and start again        bet 1 unit on all 20   increasing the bet by 1 every spin      any number that comes up remove it from the 20      example 33, 27, 14, total of 120 units bet, win on 14,-15 units, next is 26, 12, total of 171 units bet, win on 12,+4 units, next is 10, total of 108 units, win on 10, +102,   profit of 91 units and start again     this is working with the law of the third increasing bet amounts and decreasing the amount of bets  any thoughts, ideas would be appreciated

[Jesper]

pick 20 numbers, to keep it simple I am betting on 1- 20    bet for 20 spins or until you have a profit and start again        bet 1 unit on all 20   increasing the bet by 1 every spin      any number that comes up remove it from the 20      example 33, 27, 14, total of 120 units bet, win on 14,-15 units, next is 26, 12, total of 171 units bet, win on 12, 4 units, next is 10, total of 108 units, win on 10,  102,   profit of 91 units and start again     this is working with the law of the third increasing bet amounts and decreasing the amount of bets  any thoughts, ideas would be appreciated

Yes that is a realistic way, if we are very much down due to late hits, we can not come back easy, one way is a hit on a high bet number, and to make it not too expency we have to reduce the numbers in play.

Sometimes I move the bet from a number and add it to an other bet, in order to keep the total bet down and come back again on one hit.

[dobbelsteen]

Bayes thank you for the very clear and understandable paper of the 2/3rule. Some important features of a random sequence are explained with mathematical formulas. I have tryed to understand these sequences by simulations and research.
In my opinion the random sequence must satisfy the statistic rules and formulas, From this point a random sequence is not true random. I always say that a sequence has a memory for statistical features.
For the ECs the triangle of Blaise Pascal can shows us also alot.

[Bayes]

Hi dobblesteen,

Thanks, I'm trying not to include too much heavy maths because I know it's not everyone's favourite subject. 

BTW, I've added an option to the calculator which will hopefully make it more useful. Now you can enter a value X for the number of trials and it will give the probability that at least M coupons will be collected in that number of trials.

E.g. the probability is 72% that at least 23 different numbers will come up in 37 spins.

[scepticus]

We know that Hot numbers occur within 37 spins.
If Hot " numbers"  prevail in 37spins doesn't  this suggest that "Hot" Dozens- Columns Streets. Double Streets etc. may also prevail ?
Consider Dozens and Columns  for example.
If the last 2 - 3 or 4  numbers are considered as their Dozen and column and they show only two dozens and columns would these be " hot " ?
Example;
33
31
15
Their dozens and columns are 2 and 3 D and 1 and 3 Column so we could  bet that the sleeping dozen and column will not happen on the next spin -  and ONLY the next spin. In this instance  we would bet that a number in Dozen 1 and / or column  2 would not happen on the next spin.
The numbers to bet would be
13-16-19-22-25-28-31-34 and 15-18-21-24-27-30-33-36
16 numbers straight up or 8 Splits to win either 20 or 10 chips.
Food for thought ?

[dobbelsteen]

For some systems you can make the same computig
I have the coupon theory programmed for my SSB method.A study of the result can learn you alot about the features of a system. SSB is  nearly a Holy Grail. See the blog.