odiousgambit
odiousgambit
Joined: Nov 9, 2009
  • Threads: 314
  • Posts: 8827
April 2nd, 2010 at 4:54:02 AM permalink
Quote: Croupier

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.



no offense taken, maybe I should apologize for bluntly stepping on your thread, but I knew we could flush out some answers on what I thought needed to be asked.

Something I read some time ago about human relations pointed out that "asking a question" unavoidably creates a tension and often unintended friction. I can see I should have softened my tone a bit, sorry, this is a fascinating thread.
the next time Dame Fortune toys with your heart, your soul and your wallet, raise your glass and praise her thus: “Thanks for nothing, you cold-hearted, evil, damnable, nefarious, low-life, malicious monster from Hell!” She is, after all, stone deaf. ... Arnold Snyder
FleaStiff
FleaStiff
Joined: Oct 19, 2009
  • Threads: 265
  • Posts: 14484
April 2nd, 2010 at 5:30:19 AM permalink
>fascinating thread...
yes indeed.
I think it was established decades ago, as recounted in the Eudaemonic Pie, that roulette wheels were often slightly biased and that janitorial vacuuming of the casino carpeting or other administrative actions often involved moving the roulette wheel without any subsequent testing for its being level. The physicists who were about to publish their results were contacted by academic types whose interests were more financial and who claimed to have been making rather good money at roulette wheels.

Now ofcourse we have more modern and reliable wheels.

I do know that the wheel and the ball travel in opposite directions and I do agree that the most vital time for the croupier to keep his eye on the chips rather than the wheel is when a player will attempt a past-posting to add chips on a winning bet or move chips to a winning number. However, I believe that casino procedure manuals do in fact require the croupier to look at a specific spot when releasing the ball rather than permitting him to look at the wheel.

I also know that most people who speculate or experiment about roulette wheels generally settle on an octet as the target area. All prior experiments or claims of control over the wheel's results will focus on octets. Therefore I would have hoped that our croupier would have done that.

In many American casinos a croupier at an empty table is indeed encouraged to spin the wheel so as to appear inviting to passersby who may stop out of simple curiosity and then later decide to play. Its also the reason that the electronic annunciators that give the recent results are so highly visible to passersby rather than just those actually playing the game.

One croupier joked about his maiden spin having resulted in the ball landing in an attractive young lady's cleavage and the croupier's friends having immediately used the referee's signal for "field goal". So we must bear in mind that a lot of strange things can routinely happen at a roulette wheel. I would like to inquire of our poster if in the UK a roulette wheel's little white ball actually comes in two sizes and is often switched in and out of the game? This might have an effect on play in the USA where I understand that it is common for the croupier to switch between a smaller ball and a larger one. I do not recall the wheel being spun in alternating directions however.

The one obvious problem that I wish to point out is experimenter's bias. Observing the outcome and recording the outcome generally involve inescapable errors and a motivated experimenter can become subject to the biases imposed by his expectations and desires. That is why it is often best to have one person spinning the wheel and another observing the outcome. Unfortunately this is not always possible. When Persi Diaconis of Stanford was testing the notion that a coin toss was fair, he found that his graduate student assistants made many errors because the task was boring to them. The result of his experiments were that a tossed coin was biased 51 to 49 percent rather than being unbiased. Yet we still utilize coin tosses for sporting events, elections and bar room games. Perhaps it will be the same with roulette wheels. They may be biased but are sufficiently unbiased that the casino doesn't care.
FleaStiff
FleaStiff
Joined: Oct 19, 2009
  • Threads: 265
  • Posts: 14484
April 2nd, 2010 at 5:46:29 AM permalink
As to octet betting: it is a recognized "system" used by roulette players and therefore anyone with knowledge of a specific target number could easily make octet bets to cover his actions and delay any intense scrutiny by surveillance and floor personnel since casinos usually love system players.

Persons who write down roulette wheel results and see trends or biases often leave casinos broke but if Croupier conducted his experiments on a dead game when there were no players present then no one would have been concluding that there was any target number.
midwestgb
midwestgb
Joined: Dec 8, 2009
  • Threads: 12
  • Posts: 465
April 11th, 2010 at 8:17:29 PM permalink
I have a question about this study, which btw I find very consistent with my own observations:

Would it be possible to take your data to the level of 4 sectors? If so, have you done that as of yet?

In my own Roulette experiences, I've found there are 'days' and 'dealers' for whom a 9-number sector becomes predictable, and betting much easier as a result. Thoughts?
Croupier
Croupier
Joined: Nov 15, 2009
  • Threads: 58
  • Posts: 1258
April 11th, 2010 at 9:21:33 PM permalink
I am currently attempting a second study into this, where I am narrowing down the target field to 9 numbers. I completed the first few hundred spins of this trial today. I will post the further results here when I have them.
[This space is intentionally left blank]
odiousgambit
odiousgambit
Joined: Nov 9, 2009
  • Threads: 314
  • Posts: 8827
April 12th, 2010 at 4:03:15 AM permalink
Quote: Croupier

I am currently attempting a second study into this...



I tip my hat to your employers for not getting on your back about this. They could have accused you of planning collusion with players unknown, couldnt they have? Oh, and btw, when I am in England how do I find you [g]?

Actually, I found myself fantasizing about it, and had to snap out of it. It would be quite tempting to think of this as legal, as in card counting, however unwelcome. Further thought has me thinking nay, and shame on me or anyone else for pondering collusion. Nonetheless, there are going to be people out there who would jump at it I think.

If you were able to find a dealer who unintentionally was favoring sectors, I would say that player should be held blameless for betting accordingly.
the next time Dame Fortune toys with your heart, your soul and your wallet, raise your glass and praise her thus: “Thanks for nothing, you cold-hearted, evil, damnable, nefarious, low-life, malicious monster from Hell!” She is, after all, stone deaf. ... Arnold Snyder
DJTeddyBear
DJTeddyBear
Joined: Nov 2, 2009
  • Threads: 194
  • Posts: 10735
April 12th, 2010 at 4:32:09 AM permalink
As I stated in a past post, you might have gotten lucky with your first test. Your choice of zero as the middle of your half wheel zone seemed arbitrary.

Assume that you DO tend to hit the same zone. If zero had, by chance, been on the divider between the good and bad halves of the wheel, your first test would have been absolutely inconclusive with near even results.

So instead, assume that the zero was NEAR the middle of your zone. When you go for quadrants, zero may be on the line and you'd get bad/incorrect results. Therefore, don't shoot for quadrants.

Just shoot for the zero as you had before, and record every result. You might even want to keep FOUR sets of results: Each hand, and each direction. Then plot the results, creating a bar chart in the sequence of the numbers on the wheel. Your results will look like a bell curve, with zero near the top of the bell.
I invented a few casino games. Info: http://www.DaveMillerGaming.com/ ————————————————————————————————————— Superstitions are silly, childish, irrational rituals, born out of fear of the unknown. But how much does it cost to knock on wood? 😁
Nareed
Nareed
Joined: Nov 11, 2009
  • Threads: 373
  • Posts: 11413
April 12th, 2010 at 7:34:03 AM permalink
Quote: odiousgambit

Actually, I found myself fantasizing about it, and had to snap out of it. It would be quite tempting to think of this as legal, as in card counting, however unwelcome. Further thought has me thinking nay, and shame on me or anyone else for pondering collusion. Nonetheless, there are going to be people out there who would jump at it I think.

If you were able to find a dealer who unintentionally was favoring sectors, I would say that player should be held blameless for betting accordingly.



Right. Collusion is a kind of cheat, which is illegal and immoral. Noticing a bias, either a "natural" wheel bias or a dealer bias, is making use of available information.

Either way, it would be relatively easy for the casinos to end both. The first alternative that comes to mind is an automated roulette wheel, where the dealer pushes a button and the wheel does the rest. The weheel speed, ball release point, ball direction and ball speed would vary randomly. Then the dealer collects and pays bets.

Second,a nd much simpler, would be to semi-automate the wheel. dealer pushes a button and the wheel spins with a random direction and speed. Anotehr RNG chimes or turns on a light which tells the dealer to release the ball in a given direction.

good for the casinos, but players would then begin to suspect the wheel of being fixed not to pay off frequently. Oh, well, win a few lose a lot ;)
Donald Trump is a fucking criminal
midwestgb
midwestgb
Joined: Dec 8, 2009
  • Threads: 12
  • Posts: 465
April 12th, 2010 at 6:35:33 PM permalink
Quote: Croupier

I am currently attempting a second study into this, where I am narrowing down the target field to 9 numbers. I completed the first few hundred spins of this trial today. I will post the further results here when I have them.




Awesome news. I've been around Roulette in a semi-serious way for the past couple years, and I'm convinced that SOME dealers have SOME degree of wheel control available to them, down to the sector level. I'm not suggesting they can hit a sector of their choosing every time, but that such dealers can overcome true Randomness when they opt to do so.

And then, the observant player may take potential advantage.
Keyser
Keyser
Joined: Apr 16, 2010
  • Threads: 35
  • Posts: 2089
April 16th, 2010 at 3:55:21 PM permalink
This is a simplified explanation as to how dealers can do effect the outcome via intentional or unintentional sector shooting.

DS does not work because the dealer is super human or extremely skilled. It happens because the wheel enables it to happen and the dealer is basically lazy.


There is actually quite a bit of confusion about Dealers signature. It's not so much that a dealer even really has a particular skill, but that the wheel is ideal for it. The wheel speed and a dominant drop is what makes it look as though the dealer is "super" skilled. It has more to do with matching up the "frames" or sections of the wheel that are most likely to pass under the dominant drop when the ball is present.


For example: Let's say the dealer "shoots" the ball and that the ball makes 17,18,19,20,21 revolutions over a series of different spins. At a certain wheel speed, the dealer may have the same chance of hitting a specific "frame" or section of the wheel on more than one ball pass. In some instances, the dealer may have as many as three chances at a section of the wheel, if it is traveling at the ideal speed. (Of course ball track fitness and ball health play a big roll since these can lead to irratic ball decay chatter and slop on the track.)

(This is a very rough explanation)

At wheel speed X the dealer may have the opportunity to hit the same section of the wheel from the ball release number on ball rotation 17,18, and 21

At wheel speed Y the dealer may have the opportunity to hit the same section of the wheel only on ball rotation 17, and 19.

At wheel speed Z the dealer may have only one opportunity to hit the same section as the previous spin, if the ball drops on ball rotation 19 only. Understand?


That's why I use to call it "framing"- (looking for the best wheel speeds for the largest number of opportunities of a sectional hit). The above example is ONLY an example. It leaves out so much information. There are actually more variations to it. If you want to really grasp the concept, you need to film a wheel. The above example is really to illustrate how a dealer may hit the same section of the wheel, even if the ball makes an additional rotation.

I have already tested this method to several thousands of spins. Would you like to see some of the real test results?-See below. I did not segregate by wheel speed. I did not want to curve fit the trials. I also wanted to measure the raw effect over more than one dealer. The biggest difference the dealers made, were spinning their wheels at a less than optimal speeds. The advantage of the method still emerged even in the raw trials.

When I tested this method on wheels that had completely random drops the edge evaporated.

The two steps that you MUST include in your test for it to work is:

1. Compare only the change in travel yardage between every two spins of the dealer. Compare only spin 1 travel yardage to spin 2 travel yardage etc.,... The reason is that the dealers will continually fine tune each spin and the wheel speed continually drifts throughout a session. The signature is therefore perishable.
This also removes any doubt that the test results were some how "peak picked" or "curve fit".

2. A wheel with dominant drops.



If I would have cut and pasted the Excel program it would have looked all screwed up, so I have posted only the totals and the test results. I tested this years ago on Mark4,5 Huxleys, and Paul-sons. The chi square and the standard deviation results were impressive.

These numbers represent the change in the dealers travel yardage between consecutive spins only.

For example: spin 1 the dealer releases the ball from a specific number and the ball lands 10 pockets from the release number. The travel yardage for spin one is 10 pockets.

Spin 2 the dealer releases the ball from a specific number and the ball lands 12 pockets from the ball release number. The travel yardage for spin two is 12 pockets.

Now this is how you determine how accurate the dealer is:

Measure the change in travel yardage between every two sets of spins. In the above example the change in travel yardage is Spin 2 - Spin 1 = change in yardage of +2 pockets. Understand?
Spin 2 is 12 pockets. - Spin 1 is 10 pockets = a change of yardage of +2 pockets.

Here is how the plot looked on Wheel 1 (only 623 spins). I actually have tracked and recorded just over 7800 spins on a few different Mark4's and 5's to test dealers on these wheel designs with dominant drops. I will post those as I find them. This plot is actually 4 different dealers over the course of a few days. The relative positions that we are most interested in are of course, for relative positions -1,0,+1.
This three number sector was already over 4.28 st dev. at only 623 spins.
The chance of randomness for the twelve pocket sector was quite impressive.
While the standard deviation could be considered random given the small trial, what makes it significant is that it is where we would predict it to be. Understand?
The scatter out to the left and right of relative position 0 is also interesting considering the location of the ball deflectors.

When the sum of the neighbor 5 and 10 are examined the signature is quite obvious. There is also other frequencies that should be tested. I have software that does the work for me.



relative position

-18-7 hits
-17-15
-16-15
-15-11
-14-16
-13-17
-12-12
-11-21
-10-23
-9-12
-8-11
-7-20
-6-13
-5-23
-4-16
-3-13
-2-15
-1-28
0-26
+1-24
+2-13
+3-13
+4-22
+5-15
+6-19
+7-25
+8-15
+9-20
+10-12
+11-14
+12-16
+13-12
+14-17
+15-13
+16-16
+17-11
+18-17
+19-15

(Sorry, I don't know how to post my program graphs on this website.) On some of the wheels we tested, the standard deviation for the five pocket section were well above four and five standard deviations after just a few thousand trials. Again, we are only interested in the results for one specific location on the graph, the zero change in yardage region.)


While this method is interesting and does provide you with a real edge, there are better ways to play using real VB.
Please note, I'm not selling anything and I'm not the international spokesman for anything that is for sale.

When testing your results, collect statistically relevant sample sizes. Start by collecting at least 1000 spins for your tests. (Measuring just 10 or 20 spins is meaningless).

  • Jump to: