Roulette prediction computers

[ian]

Computers are getting ever smaller and powerful.

There are now devices that can be used to calculate the speed of the wheel & velocity of the ball and come to an almost instant conclusion as to a section where the ball may land.

Is this cheating? Or is it just using all available information to come to an educated opinion as to the outcome of the game?

What do you think?


A few links that may be interesting.

Ed Thorp and Claude Shannon reveal their invention of the first wearable computer - Way back in 1966!
http://www.media.mit.edu/wearables/l...ine.html#1966a

The Eudaemonic Pie
http://www.thomasbass.com/work13.htm
Actually has a picture of the shoe fitted with computer and communicating devices - 1970's

Ian

[anin]

Ian
In the case of many of these devices one is forced to ask why those who produce them do not use them to make their fortunes rather than go through the exercise of manufacture, marketing and sales to do so.
After my years in this industry my cynical mind tells me that some how all is not as clear cut as is often claimed. Any punter who believes he has an edge on the house does not willingly share the secret.
I recently was shown a device manufactured to look like a motor car remote that contained the hardware and software for card counting. Only through much practise would this be of any use and then any customer sitting pressing the buttons on his car remote in the casino would soon have questions asked!
In the 60s and 70s there were many books written on systems that helped beat the casino, apart from the clasics written on card counting most were an attempt to make up the authors losses in the casinos.
To measure the speed of the wheel and the velocity of the ball some form of device will have to be fitted to the wheel or pointed at the wheel. I can not see an operation allowing this to be done.

anin

[Brian]

Ian,

Yes. If you use anything to defeat the chance element built into casino games you are cheating.

Card-counting is not cheating because the chance element in BJ is flawed and not the fault of the counter.

You wrote: "There are now devices that can be used to calculate the speed of the wheel & velocity of the ball and come to an almost instant conclusion as to a section where the ball may land."

My question would be: HOW?

It appears that your device would require two separate velocity meters (radar guns?) in order to take a snapshot of the velocities of both moving parts at a certain point in time. Either those meters or some other device would also be required to capture where the ball was in relation to the wheel at that point in time. These pieces of information would then have to be fed into a computer that had been calibrated for the friction produced by THAT particular wheel and THAT particular ball (this would involve hours of data collection and processing and an ability to differentiate between white plastic balls). If you had such a device, your only other problem would be in getting the "result" from the computer/radar gun operator to the bettor. Then there's the random effect of the wheel obstacles and ball bounce. The device doesn't sound like it would work in a real-life situation.

The first rule of casino cheating is:

Distract attention away from the cheating, not towards it.

There are fixed sets of behaviours and movements that belong to each game. Surveillance should be consciously aware of what those normal behaviours comprise. Dealers and other employees are usually aware of them also, albeit unconsciously. If a player or spectator does something that doesn't belong to those sets, the dealer's gaze will automatically be attracted toward the movement. The act of capturing the velocities of ball and wheel would necessitate non-normal behaviour on the part of the computer user and would therefore be easily noticed.

In order to cheat successfully, all relevant game employees (and some players) must be distracted. The people doing the distracting will exhibit unnatural behaviours (usually exaggerated, to force an extra moment while the employee tries to rationalize the event) in order to capture the attention of an employee. The only member of the cheat team who will be moving and acting normally (except for the split second of the actual move) will be the cheater himself.

The bottom line is that if you have to distract all these people for the excruitiating long period of one second in any case, you might as well just stick 10 chips on the winner before the dolly gets there. It's easier to plead ignorance in court when you don't have $200,000 of hardware in your pocket.

"I'm just pleased to see you, your honour!"

[eyeinthesky]

Agreed,.......... I can't see, (but would love to know how) how such a computer can work. I have read a bit about the roulette computer but doubt it very much as no two roulette wheels are identical, making the outcome even more "variant".
The best wheel cheating method I have seen is on the Big Wheel, (Big 6) where a lumpy delaer was calling no more bets just as the wheel was about to stop!! Making it possible to at least eliminate what NOT would come up as the winning section. All 20 of us appreciated his help.
Cheers

[ian]

There perhaps was a time in the antiquities of casino gaming when allowing punters to write down the results from previous spins or coups may have been looked upon with suspicion.

Pen's and paper are now freely available as the result of a previous decision in AR or PB has been shown not to effect a future outcome.

That same pen and paper cannot be taken onto the BJ table as the composition of the deck remaining will effect the outcome of hands to come. If a player started to write down how many 10's or Aces had gone they would be politely asked to stop.

I also have to agree that card counting is not cheating but could be defined as "Advantage play". The card counter after all has not interfered with the shuffle or marked the cards.

Punters have always been allowed betting "in Running" on Roulette. If a device is in play that does not effect the ball, wheel or involve late betting or past posting etc.. IMHO I struggle to define it as cheating.

Having however a magnetised Roulette ball is of course a different story.

Further views are very welcome

Ian

[Brian]

Cheating in casinos doesn't have to involve intentionally affecting the outcome of the coup, e.g. false shuffle, picking the ball up and placing it into a number, etc.

Cheating is the act of defeating the built in house edge of the games.

For example, if you see the dealer's hole card by accident and act on that information, you're cheating. You probably won't be caught or even challenged, but on moral grounds you will have broken the tacit contract that you made with the casino when you made a bet, i.e to abide by the rules and to accept the house edge. It's like finding a wallet and keeping the money in it. Not a crime that you'll probably be busted for but a crime and an immoral act just the same.

Obviously, marking cards, using a computer or any other manner of defeating the house edge must be seen for what it is: cheating.

Eyeinthesky is on the right track regarding roulette computers; recognizing "biased" wheels is the key to many tracking methods used, be it a computer, a stop watch, or an individual with a remarkable talent for counting the number of rotor spins in his head. And, while no two wheels are exactly alike, many wheels do exhibit a bias, i.e. the ball frequently drops off at the same spot on the wheel.

Once the "drop-off spot" is identified, the advantage player calculates which segment of the rotor will be under the ball at the time its orbit decays and it drops. "Knowing" which segment the ball is likely to land in is certainly not a sure thing, but it gives the bettor a chance to wager on all of the numbers appearing in that segment, lowering the house P.C. substantially. Finding biased wheels is not easy, but neither is it that hard to do, according to reliable sources. The more casinos a city has, the better chance that one can be found and exploited.

I've read several posts on the Internet about roulette tracking possibilities (one indicating that mathematicians were formulating roulette principles by timing the ball and rotor speeds as early as 1984). I have no personal knowledge that computers capable of tracking the ball and rotor speeds are effective, but nothing would surprise me in this business anymore.

[mikebird]

I can confirm that roulette computer do work, having trid one out for a casino operator some years ago.
Basically, you enter a signal to the computer each time the zero passes you in one direction while also entering a different signal each time the ball passes you in the opposite direction.
The computer then 'learns' the wheel and, after about 10 or so spins, is able to predict the portion of the wheel that the ball will drop into. Obviously it can't predict the effect of the wheel studs may have so it can't be desperately accurate. Nevertheless, it can predict up to a one third segment of the wheel with surprising accuracy.
Remember that, on a single zero wheel, you only need to eliminate 3 numbers to be on the winning side and we are talking about eliminating up to 20 numbers or more.
Obviously, a computer user will regularly bet late and also use lots of neighbours bets to cover a section of the wheel.
Hope this helps
Regards
Michael Bird

[Brian]

Thanks Mike,

Again, not very practical for real life gaming situations.

This line took my notice, though:

"The computer then 'learns' the wheel..."

Given that the rotation velocity of both wheel and ball would each be different on every spin during the learning phase, what would the computer "learn"? Since the wheel does not slow down materially over the length of the average spin, the only thing that the computer would need to "learn" would be the ball resistance factor. However, even if it could "learn" such a factor, the computer would require the input of two occasions of when the ball passed (for example) the zero, in order to have an "initial" velocity to work with. It would also require two inputs of when the zero passed a certain place on the wheel rim in order to have a wheel velocity for its computations. It would also have to be programmed to handle alternate wheel direction.

When all is said and done, the probable result of attempting such an scam would be the manager pointing at the computer guy and telling security,
"Throw that dizzy (fellow?) out!"

[ian]

The thread theme idea came about after finding the website
http://www.newtonroulette.com

Mike - It's only 1 number advantage to make it an even game in single 0 roulette.

At least 7 variables as I see it.

Wheel slowdown over time
Wheel speed for individual spin
Ball slowdown over time
Ball speed
Deflector effect on ball
Ability to take measurements undetected
Ablity to make calculations and have time to place or call bets.

Looks to be a very tall order.
But would probably work in a laboratory trial.

Just like the apple - the ball has to come down sometime!

Ian

[xraytrog]

Device use is "alive and well". I have seen devices in use and yes they do rely on training and skill {as do we} and yes as has been mentioned a players standing by a wheel "clocking" it may arise suspicion - we hope!

They are not accurate to a number, but sections or neighbours as mentioned are viable.

Where these devices are starting to come into their own are with the use of "Touch Bet" or "Rapid Roulette" games. The user has a good view of the wheel {we even put it on a nice screen for him/her}. He/she doesn't need to be near the wheel. He/she can place bets at the touch of a button, any combination of bets for that matter in a split second.

Question - "Is our technology playing right into their hands?"

[ian]

Yes I can see where the Touchbet terminals & Rapid roulettes tucked away in the corners with a player tapping away on a PDA will not attract to much attention and could be very vulnerable.

"Oh I'm just putting in some business data in-between spins"

Especially when a Voisions or Tier can be placed at just the touch of another button.

Ian

[Brian]

Ian,

I believe you missed one item from your list:

The ability for the computer operator to receive the "answer" from the computer and to pass that answer to the bettor.

Mike's comments seem to suggest that a roulette wheel has three possible thirds. A single zero wheel has 37 possible sections, no matter the size of the sections from one to thirty-six numbers. In order to have a fair chance of hitting the winning third after passing the obstacles, the computer would have to return the winning number, i.e. the number that would hit if there were no obstacles on the wheel. The person making the bets would have to bet that number and its twelve (or whatever) neighbours and would, therefore, have to memorize 37 different section bets of 13 numbers each. He wouldn't have time to consult a roulette card and memorizing the wheel would probably also prove too slow a method if the "answer" was the middle number of the 13 number section. I suppose there could be more than one person doing the betting.

Arbitrarily dividing the wheel into three fixed thirds would not make efficient use of such a computer's potential. Efficiency would lessen as the drop point predicted was nearer the ends of a particular third. In any event, the computer could not be expected to return one out of three fixed wheel thirds as an answer without first knowing the exact drop point. Why would you want the less valuable information of correct third?

Just in case though, have you eliminated all roulette call bets from your casino yet?

[ian]

Yes Brian,

The information from the device to the bettors is obviously the most important link in the chain. No use the answer coming back when the dolly is on the number or the croup's on French are anouncing and indicating the winner.

But in Germany and many European countries a neighbour bet can be by 2, 3, 4 or even 5 either side of the target.

Covering a section to give an Even chance in almost half the wheel can be done with 3 streets.

I presume that the players using the system will know the wheel backwards, forwards and the layout inside out.

In most casinos the average player knows the wheel better than the average dealer.

Ian

[Brian]

How long does the longest, hardest ball spin take to end, at a guess? Do you believe that a computer could accurately predict anything from four input signals that were entered upon the ball passing the same point twice and the wheel passing the same outer rim point twice, with the ball whizzing round in a blur? Or would you have to wait until the ball slowed a little? If so, would you have time to get the answer and pass it to the bettor?

Again, has your European casino outlawed all (by 3, 4, 5 or whatever) call bets yet?

[ian]

Hello Brian,

The neighbour bets are standard fare and are in use in Germany each day on both French and American Roulette. Can't see them ever been turned away as all business is desperately required.

If its a hard spin then the person with the gadget clicking away cannot possibly be accurate. Only when the ball slows should a click at a specified point be reasonably accurate.

But as you say - slow ball. Only a second or two away from No more bets. With alert staff it's probably not possible.

But who's to say that they are alert.

Ian

[Brian]

The best way to protect yourself from loss by theft in a casino, no matter the sophistication level of the means, is to employ alert staff who follow adequate procedures.

You don't need to invest in any "electronic computer-theft protection".

[Moike]

Thorndyke

Not being in Germany since the dry dock crossing this seems to be a distant memory but do I recall a game where a train wheel and ping pong sized ball were used? 24 numbers also I think. What I do remember was no hazards on the wheel and a spin that lasted about three minutes. This would be the game to hit.

Of course it could all have been a drunken dream

[ian]

Hello Moike,

Still in play in various locations and was and still maybe classed as a game of skill. I saw it at the Femina in Amsterdam last year and although not as popular as it once was there was some big action when I was watching.

Large turned metal wheel, a ball that looks like the white from a kids shooker table. 24 number set in the middle with a 0 and X. Numbers going from 1 to 12 and 12 to 24.

Problem is that the ball rotates in a slow oval that moves with time. No more bets being called about 15 seconds before the ball finally drops into a pocket.

But I bet that some punters have certainly had a try.

Ian

PS. Will start a new threat to get some more input as to where the game is still played

[Craig]

Hi Ian,

Originally posted by ian
PS. Will start a new threat to get some more input as to where the game is still played

'a new threat to get more input...' hope you don't mean it :-) we're scared enough of you already!

Craig

[mikebird]

This game was called 'Observation Roulette' and was regularly offered by the 'illegal' casinos that used to operate in Holland.
The central numbers section contained 24 numbers and was stationary. The outer rim was very large and shallow, with no studs. The ball, when spun into the outer rim would go round and round for a long time before, finally, falling into the numbers section. Playing Roulette in Holland was illegal because it was considered 'Gambling' so this game was invented because a judge had ruled that this game was a game of skill (the skill was observation and ability to predict) and therefore not gambling. They also has a version of Black Jack although I cannot remeber much about ti.
Regards
Michael Bird

[Newview]

Just a couple of points ....

Firstly, I believe the roulette prediction machine can work, and I would point out that if you have a heightened chance of, say a twelve number section being hit, you don't need to cover all twelve numbers. You just pick a five number neighbour section, play that, and over time you will use your edge to get ahead.

Secondly as to whether it is cheating or not, my standard answer (re card counting) as a casino manager used to be that taken to its logical conclusion, if we decided it was legal and ok and everyone started doing it skillfully, all that would happen wouldbe that roulette (or BJ in this case) would stop being offered by the casino, and everyone loses out.

On automated roulette, a major UK operator started getting very worried about the fuly automated version on the basis that all the timings on the TCS product were predictable, thus theoretically you could predict section (ie wheel starts from stationary, spins at x revs per minute, ball fired after y seconds at z speed). After some good analysis (presumably the theory was right except for unmeasurable fluctuations and wheel studs) and some absolutely laughable comments from people who didn't really understand the probabilities involved (it's spun the same number three times in four spins !! We're going to be bankrupt !!) the conclusions were mixed, but they added a couple of random elements anyway. It always annoyed me, because you only need to fully randomise one part of the cycle, but they were determined to randomise three. Waste ....

[Brian]

Newview,

Your creating an edge and winning over time is okay in theory, as is the entire computer prediction issue, but how long do you suppose you could get away with it?

My view is that it's possible in theory but you'd easily get caught in practice. Again, the dealer, albeit subconciously, develops a ideal of acceptable punter movements and actions (not to mention hairstyles, wigs, spectacle prescription strength, length of nose, boobs: quality, credibility and gravitational factors, height, weight, companion, apparent game knowledge, size of ears, parallel quality of gaze, apparel, gait, limb count, digit count, race, religion, sexual orientation, etc.) Anything outside of the dealer's normal range attracts immediate and critical attention.

The clumsiness of inputing the data, extracting the answer and either passing that information to a confederate or calling weird bets wouldn't afford you a long enough run to make use of your edge.

[Newview]

Brian,

True enough, there are a million and one things that could give you away, but I think the whole principle of this is that it's not guaranteed, it just turns a 2.7% edge in your favour by a few points, so you better find a way of making it last, because these things aren't accurate enough to give you a definite landing area for the ball.

Also, you'd want to accumulate winnings over time, rather than all at once, or your win level would get suspicious.

I don't know how you'd be best trying to run it, maybe with the "machine" operator standing away, but where they can see the wheel, and the bettor receiving paged signals ? You could make it work somehow, especially if you havea dealer on the inside, in which case you're laughing.

I'd be much more afraid of this than of card counters - by the way, as a surv / security guy Brian, do you agree with the popularly held theory that casinos make much more out of poor card counters than they ever lose to the good ones ?

[Brian]

Newview,

The poor card-counter theory may be true (who knows?) but I can't see how anyone ever came up with it. You could never accurately predict how much your casino loses to undiscovered theft, errors or counting, nor how much you win from poor counting. You couldn't even make an informed guess, having zero information to go on (what's poor counting but systematic non-counting). Anything though, books, movies, systems or whatever, that encourages the money-for-nothing casino idea are all good for business.

Again, I believe that card-counting is possible in theory but not very practical. The concentration and card clocking, etc., would easily give you away. I mean, if the guy next to you or the supervisor says something to you during the hand, do you just ignore him?

Playing perfect strategy, then deviating or upping your bets suddenly are other things that will get easily noticed. We'll assume that you'll be using these skills on a decent limit game having adequate supervision.

The worst part about card counting is that there is no guarantee that you'll ever win a hand. If you intend making a living out of playing blackjack, you'd better have a healthy bankroll to cover the inevitable bad times. You'll also require cojones to ride those times out.

If you play perfectly and everything goes your way physically, you can expect to win the equivalent of one average bet unit every hundred hands over the long run. That's a grind item that doesn't get much publicity amongst the hype. No other expectation may be guaranteed because even if you may gain the advantage of knowing when its favourable to raise your bet, there is no guarantee that such an occasion will ever arise. So you get 1%, if you're good and the rest is a crap shoot. Get a job! Better still, write a book.

The strangest part of the whole card counting myth (not that it's not possible but the legends that surround it) is that people with above-average brains who studied at MIT could be expected to spend years sitting playing a blackjack game that has had its chance aspect "guts" ripped out for the most part. It must be nearly as bad as playing BJ for hours with nothing wagered. For MIT alumni, that's got to be hell.

Concentration and utter boredom seldom go hand in hand.