Law of the 3rd Method - Red only

[Colbster]

This came to me this morning and is testing solidly by hand but it also makes sense from a statistical standpoint, so I figured I would get some feedback.

(Excluding 0/00 for the sake of the math here)

If we cover 30 of the 36 spots, it is essentially a break-even proposition - we expect to win 30 out of 36 spins (30 x +6 = 180) and lose 6 out of the 36 spins (6 x -30 = -180), meaning that we are at 0 at the end of the cycle.

However, in any given set of 36 spins, we expect a third to hit once, a third to hit twice, and a third to not hit at all.  This holds true as we grow or shrink our sample (For example: 1 dozen will hit one, 1 dozen will hit twice, and 1 dozen won't hit in a typical 3-spin cycle).  Similarly, if we watch the 6 we don't have covered over the next 36 spins, 2 should hit twice, 2 should hit once, and 2 should not hit at all.

My method is as follows:

Track until you have 12 unique red numbers.  Keep them in the order that they hit.  As a repeat hits, remove it from earlier in the list and move it to the most recent spot.  Continue this until you have 12 red numbers.

Begin betting 18 units on black and 1 unit each on the 12 red number that have hit.

Every black is a +6 (18 out of 36 = +108 units)

Every time one of our reds hits, we are +6

We lose 30 when we have a hit on one of our 6 sleepers.  However, 2 of those sleepers will hit twice, 2 will hit once, and 2 will not hit.  The two that hit twice have been moved to our rolling list of 12 numbers, so those only hurt us 30 units the first time they hit.  That means we will have 4 x -30 during a course of 36 spins (-120 units)

That leaves us with 14 (instead of the expected 12) hits of +6 (+84).  (+108+84-120=72)

The expectation would be a +72 by relegating the 2 that won't hit to winners instead of losers (Gain 2x6 plus save 2x-30)

Yes, some of our 12 won't hit, but that is fine as long as the sleepers stay sleepers.

[Blood Angel]

Hi Colbster,

I love all things law of the third.

This is a take on it that I haven't seen presented before. I will re-read and give it a try.

When you say "it is testing solidly" what sort of tests is that, spin wise and results wise?

thank you

BA

[Colbster]

I test by hand, as I have not learned the finer arts of programming, regrettably.  I tried a while back, but I just couldn't get my mind around the concept.

I don't have my spins results immediately available, but have been spinning BV until my butt has become numb sitting here, so several hundred spins.  I have yet to fall negative by even a unit, although I am sure it will happen.  I'm thinking a session BR of 90 units, play until up 90 or until a drawdown of 90 from the max.

The math says that you should lose 1 in 6 but eliminating 2 of the 6 not covered as possibilities, there are 4 in 36 or 1 in 9 chances to lose.  That is very consistent with what I have been seeing all morning.  I would say this is at least solid enough for someone that can program to give it a whirl.  I like it well enough and I try to not inundate this board with ideas unless I think they have merit.

[Blood Angel]

  I try to not inundate this board with ideas unless I think they have merit.

No,this is a given from you. One other question if I may? Are you Flat Betting?

[Colbster]

Absolutely flat betting.  I just got a wild thought a few minutes ago about doubling up after a loss to quickly recover and then my imaginary wallet burst into flames, my imaginary wife left me for an imaginary millionaire, and I went right back to flat!

[Blood Angel]

Absolutely flat betting.  I just got a wild thought a few minutes ago about doubling up after a loss to quickly recover and then my imaginary wallet burst into flames, my imaginary wife left me for an imaginary millionaire, and I went right back to flat!

Lol sounds like a good enough reason. Thank you for answering my questions so quickly and fully.

[Blood Angel]

Hi Colbster,

Please find attached a quick 200 spin session.
A very interesting way to bet. I will give it a longer go when time permits.
Thank you for posting this.

BA

[porkeporkeporke]

could you give us an example please?

[Colbster]

Man, BA, what I wouldn't do to be able to work your magic and have info that quickly.  It takes me FOREVER!!!

[Colbster]

Using the stop loss of 90 from the max would improve those results by around 50 units.  Not terrible!

[Blood Angel]

Hi Colbster,

thought I would try a 1000 spin session, and wish I hadn't. 
It never really got going at all. Over a large sample the reds not covered, just hit and hit.

BA

[vundarosa]

Hi Colbster,

thought I would try a 1000 spin session, and wish I hadn't. 
It never really got going at all. Over a large sample the reds not covered, just hit and hit.

BA

------------

BA, were you cycling the red numbers? I mean, picking new set of 12 numbers after each 36 spins?

vundarosa

[Blood Angel]

Hi

Yes. As I finished the 37 spins I looked back and used the last 12.

BA

[Colbster]

For clarification, I was not using a 36 spin cycle.  I was using a running 12 updated with every single red hit.  When one became a repeat, it moved up to the most current spot.  Today, I also considered the benefits of using 12 units on black instead of 18.  That gives a win of 12 and a loss of only 24, meaning only a 2-hit recovery compared to the 5-hit recovery that includes hits on black at +6 units per hit.  I don't know if it would have some impact in the trends, but it is a thought.

I don't see how the sleepers hit like they did.  It seems contrary to what we have all been taught regarding the number of spots hit in a cycle.

[Colbster]

The reason that this may be significant is that if you are keeping the 12 spots bet and the 6 unbet constant, you are missing the two hits that you would expect as repeats by moving up the 4 of the 6 that we would expect to hit.  That would show as 2 losses of 30 instead of 2 more hits of +6.  That is a pretty big swing.