Author Topic: 3 out of 22 ??  (Read 3239 times)

0 Members and 1 Guest are viewing this topic.

UK-21

• New
• Posts: 115
• Thanked: 39 times
• Gender:
3 out of 22 ??
« on: May 09, 2016, 07:15:20 PM »
How often would you expect to see 22 consecutive results with only three red numbers showing?

The following users thanked this post: Reyth

Reyth

• Global Moderator
• Hero Member
• Posts: 3825
• Thanked: 1207 times
Re: 3 out of 22 ??
« Reply #1 on: May 09, 2016, 10:21:48 PM »
Well its gonna be pretty rare because the wheel can be expected to go 160 spins without a particular set of 3 numbers showing.

Let me draw up a quick app, it should be easy I will just count the occurrences...
« Last Edit: May 09, 2016, 10:35:39 PM by Reyth »

Reyth

• Global Moderator
• Hero Member
• Posts: 3825
• Thanked: 1207 times
Re: 3 out of 22 ??
« Reply #2 on: May 09, 2016, 10:32:11 PM »
Out of 16M sessions, 10416 came up with 3 or less of any EC.  This works out to 0.000651 or 1 in 1536.  Pretty sure this means you could end up spinning 3000 times and not see it...

Code: [Select]
`5 DIM sh(22)10 RANDOMIZE TIMER20 FOR i = 1 TO 22: r = INT(RND * 37): sh(i) = r: NEXT i: se = se + 1: IF se >= 16000000 THEN 10030 'evaluate session40 oc = 0: FOR i = 1 TO 22 50 IF sh(i) > 18 THEN oc = oc + 160 NEXT i70 IF oc <= 3 THEN su = su + 180 GOTO 20100 'output110 PRINT "Out of 16M sessions,"; su; "came up with 3 or less of any EC."`
« Last Edit: May 09, 2016, 11:35:20 PM by Reyth »

UK-21

• New
• Posts: 115
• Thanked: 39 times
• Gender:
Re: 3 out of 22 ??
« Reply #3 on: May 10, 2016, 05:09:17 AM »
I am going to calculate the prob using the binomial coefficient formula, which I found on the web care of Mr Google.

This has happened to me twice in two days, whilst playing online with a major UK household name provider offering RNG roulette from Playtech. Didn't keep a tally of the number of spins I played the first time (less than a thousand I would think), but yesterday it occured within a sample of 192 spins.

I am getting bored of seeing a large proportion of my online gaming results falling between two and three standard deviations south of the EV.

UK-21

• New
• Posts: 115
• Thanked: 39 times
• Gender:
Re: 3 out of 22 ??
« Reply #4 on: May 10, 2016, 05:09:52 AM »
Out of 16M sessions, 10416 came up with 3 or less of any EC.  This works out to 0.000651 or 1 in 1536.  Pretty sure this means you could end up spinning 3000 times and not see it...

Code: [Select] [nofollow]
`5 DIM sh(22)10 RANDOMIZE TIMER20 FOR i = 1 TO 22: r = INT(RND * 37): sh(i) = r: NEXT i: se = se + 1: IF se >= 16000000 THEN 10030 'evaluate session40 oc = 0: FOR i = 1 TO 22 50 IF sh(i) > 18 THEN oc = oc + 160 NEXT i70 IF oc <= 3 THEN su = su + 180 GOTO 20100 'output110 PRINT "Out of 16M sessions,"; su; "came up with 3 or less of any EC."`

What language is this?

Reyth

• Global Moderator
• Hero Member
• Posts: 3825
• Thanked: 1207 times
Re: 3 out of 22 ??
« Reply #5 on: May 10, 2016, 05:19:31 AM »
BASIC.

Wow.  Sorry about your tough luck.  I know what its like though.

UK-21

• New
• Posts: 115
• Thanked: 39 times
• Gender:
Re: 3 out of 22 ??
« Reply #6 on: May 10, 2016, 06:39:16 AM »
Out of 16M sessions, 10416 came up with 3 or less of any EC.  This works out to 0.000651 or 1 in 1536.  Pretty sure this means you could end up spinning 3000 times and not see it...

Your 16m trial simulation is proof of the laws of big numbers - the actual answer when calcuating it is 1540. I'm going to qualify this by doing the calc manually, but if it's of any use to contributors, here's a link to the online calculation tool I found:

http://www.miniwebtool.com/binomial-coefficient-calculator/?n=22&k=3 [nofollow]

Here's another with an explanation of the underlying logic:
http://www.zweigmedia.com/RealWorld/tutstats/bincoeffs.html [nofollow]

+++++

Thought the coding syntax looked familiar. I use VBA from Excel. Not the most efficient for programming, but fairly easy to learn - I managed it.
« Last Edit: May 10, 2016, 06:42:40 AM by UK-21 »

The following users thanked this post: december, Reyth

Reyth

• Global Moderator
• Hero Member
• Posts: 3825
• Thanked: 1207 times
Re: 3 out of 22 ??
« Reply #7 on: May 10, 2016, 07:00:57 AM »

dobbelsteen

• Hero Member
• Posts: 1466
• Thanked: 436 times
Re: 3 out of 22 ??
« Reply #8 on: May 10, 2016, 08:40:30 AM »
If you know the answer, how do you think to use the answer for a roulettesystem.
A 22 EC sequence has 2^22 different sequences. The zero is neglected
SSB uses a 20 sequence. the zero is not neglected. The zero is black when you bet red. With the La Partage rule the zero is an important benefit.

scepticus

• Hero Member
• Posts: 2077
• Thanked: 431 times
Re: 3 out of 22 ??
« Reply #9 on: May 10, 2016, 12:59:10 PM »
I am going to calculate the prob using the binomial coefficient formula, which I found on the web care of Mr Google.

This has happened to me twice in two days, whilst playing online with a major UK household name provider offering RNG roulette from Playtech. Didn't keep a tally of the number of spins I played the first time (less than a thousand I would think), but yesterday it occured within a sample of 192 spins.

I am getting bored of seeing a large proportion of my online gaming results falling between two and three standard deviations south of the EV.

Proof that " variance "is our real enemy ?

UK-21

• New
• Posts: 115
• Thanked: 39 times
• Gender:
Re: 3 out of 22 ??
« Reply #10 on: May 10, 2016, 06:49:04 PM »
I certainly seem to have done something to piss the variance fairy off.

Within my 192 spin session, there was a spate of nine (or was it ten?) reds on the trot - but guess what, they didn't put an appearance in until after I'd started using black as the first step in my progression. Funny that? I'm sure that if you were to compare the frequency of red -v- black over the 192 spins they wouldn't have been two far off overall.

Generally I don't buy into all of the conspiracy theories around online gaming being rigged, but when you hit results like this in such a short span it does leave one thinking "hmmm . . . . ".

Reyth

• Global Moderator
• Hero Member
• Posts: 3825
• Thanked: 1207 times
Re: 3 out of 22 ??
« Reply #11 on: May 10, 2016, 06:52:42 PM »
Many sites are monitored by "fairness auditors".  Also knowing people that regularly win playing at your chosen casino can be very helpful.

Without a doubt it is TOO easy to suspect the casino of cheating; I have found myself being tempted to think so EVEN WHILE I WAS GETTING BETTER ODDS THAN RANDOM!

I mean roulette is so savage a game just on its own, casinos don't need to cheat.

UK-21

• New
• Posts: 115
• Thanked: 39 times
• Gender:
Re: 3 out of 22 ??
« Reply #12 on: May 10, 2016, 07:21:59 PM »
I don't know anyone who regularly wins . . . .

Bayes

• Veteran Member
• Posts: 688
• Thanked: 556 times
• roulettician.com
Re: 3 out of 22 ??
« Reply #13 on: May 10, 2016, 07:34:14 PM »
I certainly seem to have done something to piss the variance fairy off.

Which OC are you playing at?

UK-21

• New
• Posts: 115
• Thanked: 39 times
• Gender:
Re: 3 out of 22 ??
« Reply #14 on: May 10, 2016, 09:05:53 PM »
William Hill. And it was for pennies . . .