Repeating dozens

[canroul]

Played dozens the other day, I wanted to see how many times would it be single wins before a double showed up.  It went 18 times 1-3-2-3-1-2-1-3-1-3-2-3-2-1-3-2-1-2 and the 2 WIN!! Now when I go to a RNG and crank out 1000 numbers between 1 and 3 I had a hard time even finding 12.  So is it I hit on once in a million or in the next thousand I may find a 20 or 25 without a double win.
Has anyone done some research in this, there is no point looking for info if it is in the library of the Roulette Forum

[FreeRoulette]

Try this tool
nolinks://nolinks.loothog.com/Systems/CheckSpins.php

[guido1]

Played dozens the other day, I wanted to see how many times would it be single wins before a double showed up.   It went 18 times 1-3-2-3-1-2-1-3-1-3-2-3-2-1-3-2-1-2 and the 2 WIN!! Now when I go to a RNG and crank out 1000 numbers between 1 and 3 I had a hard time even finding 12.   So is it I hit on once in a million or in the next thousand I may find a 20 or 25 without a double win. 
Has anyone done some research in this, there is no point looking for info if it is in the library of the Roulette Forum

I say you are close to 1 in 1,000.

You are looking for a multinomial distribution (4 values) of a run of length 2 or the first repeat.
In your example, the first repeat was the 19th trial.

A Good math problem, but I have a small sample of 600 actual 00 wheel spins in Excel and I count 147 repeats or streaks of length 2.
average of 1 every 4. 08 spins

Here is the distribution.
You can see there was 1 time the interval was 12 thru 15, so for you to see one on the 19th trial has a small probability of happening but that will increase as the number of spins increase in a data sample.
I will run some larger samples and post the results later

Code Select Expand

spin freq  prob
2 56 0.380952381
3 31 0.210884354
4 15 0.102040816
5 13 0.088435374
6 10 0.068027211
7 7 0.047619048
8 2 0.013605442
9 4 0.027210884
10 3 0.020408163
11 2 0.013605442
12 1 0.006802721
13 1 0.006802721
14 1 0.006802721
15 1 0.006802721
16 0 0
17 0 0
18 0 0
19 0 0
20 0 0
147 1


[guido1]

 I now say you are close to 1 in 1,000. 

I found another small sample of 1000 actual 00 wheel spins in Excel and I count 236 repeats or streaks of length 2. 
average of 1 every 4. 2 spins

Here is the distribution. 
You can see there was many between interval 11 and 15, and 1 hit on the 22nd spin.

Code Select Expand

spin freq prob
2 81 0.343220339
3 54 0.228813559
4 27 0.11440678
5 20 0.084745763
6 16 0.06779661
7 13 0.055084746
8 2 0.008474576
9 6 0.025423729
10 8 0.033898305
11 2 0.008474576
12 2 0.008474576
13 1 0.004237288
14 2 0.008474576
15 1 0.004237288
16 0 0
17 0 0
18 0 0
19 0 0
20 0 0
21 0 0
22 1 0.004237288
236 1

Enjoy

[Gar Gibbs]

It seems to me that I recently saw an article on the Internet, which said such researches, but unfortunately I can not remember where I found this article.  If I find it again, then tell about it.  I think it would have liked you.

[Kingspin]

I played this method with betting for a dozen to repeat a lot,.mostly it never seems to go past 18 spins without a repeating dozen . I used to play when a dozen has not repeated in 7 spins, a good bet.kp

[pipmenow]

Try this tool
nolinks://nolinks.loothog.com/Systems/CheckSpins.php

Thanks for the link.  It appears most of the progressions at loothog reset after one win. I want to test a progression that sometimes requires two consecutive wins to be in profit.  The progression should not reset until a profit is realized.  Does anybody know of a progression tester that can handle this.  If I overlooked it at loothog, please let me know and send me a link.  I would like to test on even money bets and 2:1 payout bets. Thanks. 

[pipmenow]

-