Roulette Spin Control

Quote: DJTeddyBear

Crap.

Although that first post was at 4:09pm Vegas time, that's 12:09 in England - where Croupier is from!



Man, if this is an April Fool's joke, ya got us good!

But we (or at least *I*) got sucked in because I saw it before midnight last night!

No disrespect Croupier, I enjoy your posts. If it's not a joke, there's another reason to do away with this silly tradition. I've already had to read 100 fake updates on facebook, deal with text versions of videos on youtube, and I had to Topeka (http://www.google.com) the answer to a co-worker's question.

No offence taken. I wish I had the foresight and imagination to make it up as an April fools joke.

Just a case of unfortunate timing.

OK. You're forgiven.

Now I won't have to come over there and kick your ass.




Yeah, that's right. Payback for an April Fools joke, and I'm on a plane. Exploiting your skill is something that I'll pass on.

I like your priorities.

Am I the only one thinking 20,000 trials could be insufficient? I'm surprised someone hasnt confirmed this, or not. Dickering with wincraps has me increasing dramatically what I used to think was enough trials.

I accept that 20,000 trials is far fewer than the millions or more that could be run with computer simulation. However I am not a computer. I am constrained by 8 hour shifts, the pit bosses decisions on which table I go on, and whether or not people were playing (due to my self imposed rules).

I would love to be able to get this up to the millions, but for me it is just not practical. I freely admit that this was not cuducted under strict scientific rules, and that it is a small sample size. I just wanted to see if I could do it. If I had my own wheel I would happily try and get a higher sample.

I think this could go the same way as the dice control threads, with those who believe and those who dont. I have no problem with this. I just wanted to try this, and see if it could be done. I am not trying to convince, but I did want to put the (small amount) of numbers out there for people.

EDIT - I apologise for my defensive stance here, but I did put a lot of time and effort into this. Having re-read my post, I think I sound like a bit of a tool.

20,000 trials is perfectly sufficient, given the results you obtained, to certify that you didn't just get lucky. The standard deviation for your experiment is approximately 50 "hits". It would be virtually impossible to obtain these results with a fair, random wheel and lucky outcome. The odds would have to be in the millions against it.

Quote: Croupier

EDIT - I apologise for my defensive stance here, but I did put a lot of time and effort into this. Having re-read my post, I think I sound like a bit of a tool.

I don't see anything wrong in what you wrote, or how you stated it. You pretty much wrote what you did, and what the results were. It was left up to the reader to draw their own conclusions. I admire the way you went about the experiment. Doing it with players on the table would not have been fair, to either the players or your employer, regardless of whether or not you are having a true influence on the outcome.
We all recognize that this is a small sample, and this just could be a normal skew of randomness. I'm not sure how this can be simulated on a computer though, since it involves physical dexterity of the dealer.

The only thing that this experience has done for me though is make me wonder on those occasions where you will see one number hit 4 out of 6 times, if the dealer was trying to do that on purpose?
Oh, it has done one other thing for me too. I was not inclined to play roulette before, and I am even less so now.

Quote: Croupier

No offence taken. I wish I had the foresight and imagination to make it up as an April fools joke.

I should have suspected that. You really would have had me. I'll check back tomorrow, just to make sure.

Quote: Croupier

I implemented the following rules.

1. The spins were to be completed on an empty table or "no game".

2. The experiment was to consist of over 10,000 spins with both my left and right hand.

3. A result would be recorded as a hit or a miss.

4.The selected section would always be the same.
This was for ease of recording results, and to give me a consistent target.

5.The selected section would be Zero section, plus 17 and 34.
This gave me just over half the wheel to aim for. (19 numbers). Plus my favourite numbers are in this section.

The following are the results.

Right Hand.
Spins completed: 10,074
Hits: 6482
Misses: 3592
Hit percentage:64.34%

If the spins were truly random, the expected number of hits would be 19/37 x 10,074 = 5,173 for the Right Hand with a standard deviation of 50.17. The probability of Croupier getting 5,360 successes or more is 0.01%. The probability that Croupier gets more than 5,557 successes is about 0.0000000000001%

The probability of Croupier getting 6,482 hits or more is ZERO. Therefore there is bias in his spin.