Roulette On dealer signature

I've been looking at dealer signature recently.

I believe it's measurable, and can provide an edge. I'm currently working on building empirical data but I have some preliminary discoveries I think are interesting.

I'm hoping we can get a discussion started on this topic.

What I've been doing so far is going on to online roulette, tracking the spins by the dealer and separating them by wheel rotation direction.

I started by first checking the number they released over, but switched to just using the last result since dealers are very inconsistent at the time in which they spin the ball. Some will leave you with 5 or 6 seconds to bet, others leave 1 second or less by the time the ball is released. They're supposed to be releasing the ball over the last number anyway in european roulette, which they're doing almost all of the time.

I created a program to track yardage between results then take the yardage by wheel rotation direction to make a prediction. Very rudimentary.

I then had the idea to start checking the standard deviation of the yardages. Very, very helpful.

First thing I needed was to get a baseline of what is actually random. 10.6 std. dev is random yardages confirmed with 100's of thousands of rng numbers. How else would you get a baseline of random / expectation? Obviously, you can't use a live dealer because that goes against the whole point of the experiment. Collecting that many spins from live dealers is also insanity.

I've checked auto roulette airball results for std. dev. and it's about the same.

Now the live dealers on the other hand are very interesting. They tend to have one good direction, probably based on the handedness of the wheel vs their dominant hand. Most dealers have a slight signature. Some have a very strong signature. Many are brutally random.

I've been saving the sessions sorted by winning or losing. I havent saved them by dealer name, which I think is a mistake on my part. This is because I wasn't sure of the consistency with which I'd be able to find them again. I also didn't track them by the wheel / table either - another mistake in my data collection efforts.

Moving forward I'm going to fix this. If anyone knows the best way to collect data for this, I'm all ears.

Here's what I've got so far.

Winning Spins
706
Losing Spins
1124
Total # Spins
1830

The winning ratio has gone up a bit, but being a bit more conservative I think that about 1/3rd of dealers have a strong enough signature that you can profit from.

Random std dev of yardage is 10.6

To begin betting I'd like to see a std dev in one direction of maximum 9. Most of the dealers in the 1/3 will be here. Some will be at 7 and even less will be between 5 and 7. I believe only once I saw a dealer with a std dev of 2.65 on their yardage.

I need to improve my methodology more. There's definitely something here even with just using yardages.

I've thought about dominant drop zones, but I don't know how I'd incorporate that. It seems like more of a dealer or wheel tell than a predictor.

Ball speed seems to be less helpful than wheel speed. It seems the ball will fall off the track at roughly the same speed each time as that's when it's energy will allow it to leave.

Wheel speed seems to be helpful in measuring dealer consistency. Using it to predict future results seems difficult though. I've thought that perhaps applying some sort of average of the wheel speed to calculate potential scatter for yardage predictions would be helpful. Scatter also seems to be directional. Wheels rotating clockwise will knock the ball back in the clockwise direction, and vice versa.

I don't believe measuring the wheel and ball speeds, and then trying to infer the landing location using physics calculations on historical data would be useful in any way

Finally, I think dealer signature is a combination of the wheel and the dealer together. 0.1 degree tilt will cause bias in a wheel. It's impossible to level a wheel that precisely and eliminate that bias.

The idea is to find the dream team.

Hoping people with experience in this area can chime in @Dr. Sir Anyone Anyone i think?

 

So, you're going to trust on-line.
Remember smartlive.

 

What happened to smartlive?

I have no reason to not trust online casinos. Logically, if they were to cheat players, they are opening themselves to even more forms of AP than by not cheating it.

 

Don’t spend your time and money on this. If you find a way to win, they don’t want your business. Ask Nimo!

 

If you win you get this response:
"Having performed a review of your account, they have made a conscious decision that it is no longer economically viable to accept your desired level of business, explaining the reduced staking opportunities you have experienced.The Trading Departments decision was based on many factors, which, unfortunately, we cannot divulge to you.This is done in accordance with our site's terms of use.

 

Yeah, so? That's going to happen if you win anywhere regardless. There's like 50 onine casinos available to me here anyway.

Anyway after a bit more poking around I'm noticing that when yardage std dev is quite high... over 10.6 for instance it appears that it's still a signature, just not a yardage signature.... Let's say a 14 std. dev. for instance.

Unless people with experience want to discuss this I doubt I'll talk about it anymore publically. I'm getting an 80% win rate on 19 numbers when I find a signature, and I am seeing indicators of when to bet and when not to bet.

This is all I'm willing to share here for the time being.

I'm not going to give my analysis away without reciprocation.

 

50 online casino’s owned by 1-3 corporations just like Coca Cola.
When you get tagged you’re done anyway or restricted to 0.10 chips.
It took Nimo 2 years and a lawyer to get his money back.

 

HAL makes a good point. I can't play Online Live Roulette or Baccarat with the Evolution gaming platform. It always comes up that my bet was rejected. Sometimes I can sneak in the first bet and then that's that, no second bet. As an experiment, I made a few bets on some of their live Gameshow products such as 'Monopoly' and the big baller or whatever it's called and I had no problem with them.

Someone once said on these forums that it's not how much you bet but how you bet and I honestly thought they were talking crap. Now I am not so sure. I can still play on the playtech platforms and they have just introduced live online from the MGM and Bellagio in Vegas which has a constant stream with real players at the table and looks legit and so all is not lost if I fancy a quick game online.

 

Hal definitely makes a good point, no argument from me on that. What happened with nimo?

I think they can tag you etc.. but you can make new accounts, VPN etc.. and they don't typically ask for verification of who you are until you want to withdraw. You can use family members etc.. there's ways around it if you've got an edge.

I used to get interruptions a lot but not so much recently. Not sure. Sometimes the video feed cuts out. I've had bets rejected as well. I think it's a user demand thing more than anything, but it could be malicious. Hard to prove either way.

 

With that said I am definitely limited on sports betting on the same casino I'm using now for testing purposes.

 

What do you mean that you have a 10.6 std.?

Can you explain what you mean?

A random sample shouldn't produce a standard deviation of 10.6 for a section arc or for an individual number yardage.

 

I generated a set of random numbers, like 10k or something. Calculated yardage (cw lets say) from the second last number and the last number, then separated by 'direction' and then calculated the standard deviation of the population.

I can post the code I'm using to calculate std Dev for clarity when I get home.

What std dev should I be seeing? Or what problems are you seeing?

What is a section arc?

 

The set of the random numbers in txt or xlsx or csv format would be interesting as well.
So if you like to post it i could do some calculation, i did recently similar calulations.

 

I'll do the calculation again tonight and post my results and data set as well.

 

I have software that will do that as well if you need help.

A section arc is a section bet of the wheel.

Example.. In the case of dealer's signature, relative positions 0,1,2,3,4 (yardage from your observation in pockets) would be considered an arc of five wide.

 

@Herby Here's the data I mentioned earlier.

Here's the 20k numbers I used for my baseline test.

My results:

Std. Dev. CW 10.63
Std. Dev. CCW 10.58

I get the same at 10k, 50k whatever.

My function(Python code) for calculating std dev:

def get_std_dev(data):
# Step 1: Calculate the mean
#data = [list(t) for t in zip(*data)]
#data = data[0]
if len(data) > 0:
mean = sum(data) / len(data)

# Step 2: Calculate the squared deviations
squared_deviations = [(x - mean) ** 2 for x in data]

# Step 3: Calculate the variance
variance = sum(squared_deviations) / len(data)

# Step 4: Calculate the standard deviation
std_deviation = math.sqrt(variance)

return std_deviation

I should probably switch to using the builtin statistics module for the mean calculation instead of doing it manually.

data is the list or population data separated by 'direction'. So yardage of one direction.

Yardage is calculated from the last 2 spins. So CW spin position (second last pocket hit) - CCW spin position(last pocket hit) = yardage

More specifically:
yardage[current_direction].append(get_yardage(wheel, results[-2], results[-1]))

Would the section arc be like betting the neighbours of your predicted number? I.e. the sector bet.

What would the arc be in the counter-clockwise direction below?

Dealer spins/pocket results:
[8, 9, 25, 0, 1, 4, 18, 23, 7, 21, 18, 2, 7, 20, 1, 24, 25, 16, 15, 19, 12, 33, 25]

Yardage:
'counterclockwise': [17, 23, 25, 14, 24, 25, 36, 24, 18, 30, 22]
'clockwise': [11, 30, 18, 25, 11, 14, 30, 34, 14, 1, 26],

My calculations
'STD DEV CCW': 5.79,
'STD DEV CW': 9.78,

If you don't mind sharing the software with me then I'd love to have a look.

 

Regarding the standard deviation. What you want is a graph that will show the standard deviation above normal for each relative position, and in sections of various widths. I now see how you arrived at the value that represented one standard deviation. Originally I thought that you were implying that you had a Z score of 10.63 or a standard deviation above the norm of 10.63 . I still need to check that number though.

I'm going to send you a PM.

 

What to think when you see you had let's Say yardage 10 then 11 then 10 again on one direction but with different number of ball s lap, not exactly same rotor speed, different ball jump, different diamond hit?
I already Saw almost exact same spins with finally same yardage but it happens barely to really take advantage rationally. Most of the Time it will be due to randomness but our confirmation cognitive bias will tend to Say if the yardage IS similar it's because the dealer IS regular...

 

From a guy that doesn’t believe in my strategies which is maths laws based to a guy that believes dealer signature is a sure fire thing is insane lol

all you need is luck to win for a dealer signature and betting 10 neighbours each number you go for haha !

Over engineering strategies is a thing bro

get real haha