Hemisphere Roulette

[ausguy]

Prior to retiring my background was in the field of Engineering Surveying. I was involved in many construction projects including road tunnels. As an intergral part of  surveying all optical equipment (theodolites, construction levels etc.) have leveling bubbles of various types to suit their design operation.

Most people would have seen surveyors with their camera man like yellow legged tripods working around Highways & construction projects ?

One particlar levelling instrument is a vernier calibrated precision optical level (a high tech mini telescope for ease of understanding). The optical vernier allows levelling accuracy to 1/100 of a millimetre. Engineers/geologists use the readings, taken on a regular basis from fixed points in the tunnel, to measure any movement in the rock.

Movement usually indicated instability & the possibility of rock falls or tunnel collapse.  Sometimes Engineering machinist companies get surveyors in with their vernier precision levels to accurately level certain equipment - lathe beds, suface grinders, CNC computerised milling machines.

Skilled tradesmen/women then work from that with the aid of engineering grade precision spirit levels with a bubble accuracy of up to one 50th of a mm per metre. These far exceeds the accuracy of "cheaper" builders spirit levels you can buy in hardware stores.

As mention of CNC machining has been made, CNC woodworking machines ( I was told by a wood joinery owner) works to an accuracy of 1/100mm. That's why when you buy a flat pack piece of furniture everything fits together so precisely. In fact, according to the joinery guy, you could buy 2 identical flat packs lay them out & mix up the panels & they would still all fit together precisely.

Roulette wheel makers use CNC machinery to manufacture the Timber/wood roulette body & the ball track support rebate. The ball track itself is a seperately manufactured synthetic item that is fitted (glued ?) into the support rebate.

With precision manufacturing technology wheel makers do accurately replicate products. This ensures that if a casino installs all the same wheels the product will be identical & therefore give equal performance at every table. Provided, naturally, that the wheels are levelled to the manufactures specifications. I'm sure in most cases they are.  Cammergh wheel maker mentions their software that monitors the number outcomes checking for bias. Ball track wear over time is entirely another issue.
It would surely be a certainty the once ball track wear on any pariticular wheel became an issue they would take that wheel out of sevice ? My guess is that they can renew the worn ball track, fit new spindle bearings (if needed ?) and restore the wheel to "as new' performance.

As to the leveling bubbles - The red triangular one in Steves pic. has a circular "bullseye" bubble. The general accuracy of them is no better than 2mm per metre.
They are on many survey instruments to get "Ball park" accuracy. The tubular spirit level type bubbles, also on the instrument, are then used to accurately level the instument via levelling screws. Automatic telescopic builders/construction levels also have a "bullseye" bubble. Once inside the circle line an internally, finely balanced array of blade like caged wires containing some magnification lenses swings under gravity to point to the centre of the Earth (called a compensator), the same as a plumb bob on a string does.
Accuracy can be in the order of +/- 1mm in 50 metres. For most people "el perfecto".

I'm suprised that some wheel maker hasn't developed a pin point laser level reader that operates off the centre spindle that along with a wheel track shaped reader. That way they could accurately check the full 360 deg. "trueness" of the ball track.

I watched some tech heads at my local casino some time ago. They had engineering quality spirit levels for the wheel levelling. They also had some electronic readers, akin to a small multi meter. They were spinning the wheel at times then discussed (I couldn't hear them) values on the meters. I was guessing they may have been cheeking for bearing "rumble" as a bias problem area ? All secret casino business.

The other thing with the red triangle level is that it is just placed on the ball track. Not very "scientific" from an engineering/physics view point. As the 3 touch points can be in slightly different track positions ?

A better levelling triangle would be one that had 3 tubular bubbles, 1 on each arm, that was precalibrated on a purpose built "SUPER LEVEL" rotating test rig.

When placed on the ball track it should have concentricity to the vertical centreline of the spindle, some form of firm location would be needed. A parallel is the wheel of a car that usually has a circular formed centre that accurately locates the wheel centrally prior to tightening your wheel nuts. Without being centralised the wheel may run eccentric = unbalanced.

AMK - Very informative link to wheel bias.   

[Steve]

Ausguy, thanks for this information. The link AMK posted isnt about "wheel bias" as such, it is about roulette computer testing and wheel tilt. The tilt is basically is a cause of dominant diamonds, and dominant diamonds help roulette computers and visual ballistics beat roulette. The study from the NWML was requested by mike Barnett who teaches casinos to catch professional players. He wanted the report to show casinos, basically to make them poo themselves so he can sell his services to them. His device uses the typical computer algorithm as explained in my earlier post.

Regarding the equipment, the ball tracks are coated in an epoxy resin. Although durable, they are basically hard plastic and eventually wear, bubble and crack. So if you measured 3 points of a ball track to make them level accurate to 0mm, it wouldnt matter because you can test 3 different sections of the ball track and get a different result. Roulette wheels will probably always have significant imperfections. This can work for or against the casino, depending on the approach the professional player takes.

My level is digital and accurate to ± 0.2° but it has the same problem explained above.

Every time my wheel (newest wheel) is moved, even if I carefully calibrate it, the dominant diamonds are different almost every time. This is because the ball track is much tougher than the average wheel (Velstone), like in the image in my post and from the Huxley site. This means "dominant diamonds" are less likely to occur from ball track imperfections, and a very slight difference in calibration makes a big difference to ball fall points.

There is a lot of physics in roulette but even the wheel manufacturers dont fully understand it. I spoke to one of the main designers from Huxley, and he made a few critically incorrect statements. This is a good thing for players, but certainly designs are becoming less predictable.

[ausguy]

Steve - All good information. One thing I don't quite follow is your mention of you recalibrating your newest wheel & the dominant diamonds are different every time. The only reason I see for that is that the physical levels are imperceptibly different from the old position to the new ?

I'd guarantee with 99.9999% certainty that with a vernier precision optical level that upon marking your premove wheel level checking points that after relocating your wheel that those premove marks could be set in the same horizontal planes as the original position. I'm not talking about height off the floor, in this instance that's irrelevant. You could move from L1 to L33 in an office tower & still level your wheel in the same plain as it was at L1.

Also with the vernier level any 3 equally spaced points around the ball track would enable those points to be perfectly levelled, within reason subject to manufacturing tolerences. This may reveal that your wheel is/never was truly level ?

Also steve how do they "finalise" the ball track ready for use ? Do they CNC machine it in situ to ensure a perfect circle ? or CNC machine it prior to installation ? Is it then a snug fit in the track wooden rebate or glued/fixed in some other way ? 

You may remember my mention that level relates to a plumb line towards the centre of the earth. This all comes about by gravity which allows the enclosed bubble & the spirit to react to any tilt. It shows level because the gravitational forces in the spirit tube are in equalibrium because the tube is level & exactly 90 deg to that plumb line to our global centre.

I don't see how what the wheel is made of influences the path of the ball from when it leaves the ball track, hitting or missing a diamond(s) prior to arriving at the rotating wheel/pocket ?

Just on your digital level, 0.20deg (how did you get the little deg symbol in there?) that's not all that good Steve. That represents 12 minutes of circular arc.  Calculating that for 1 metre represents +/- about 3.5 mm & 1.7mm for 1/2 a metre.  Engineering spirit levels have a much bigger/longer spirit tube & bubble & so offer far higher levelling capability. I just did a check on some eng. level makers & one that caught my eye was a WYLER Co. of Switzerland model that has calibration accuracy to +/- 0.02mm, that's x 175 tighter tolerences than your digital level Steve. Your digital level is only really good for casual home projects, like I've seen them in those home improvement shows levelling picture frames -"WHACKO".    Something of higher accuracy would give better outcomes,  Steve.

How do digital levels work in regard to calibrating to the plumb core line ? If they use some type of pendulum then that's where their accuracy suffers ?

The standard way, like most apprentice carpenters are taught, on a flat level surface, is to very carefully check the spirit level bubble on one "face"  then rotate the level 180deg and compare the 2nd bubble check with the 1st observation. If there's a detectable difference then 1/2 way between each "bubble faces" is true level. Some levels have adjustment, which allows for a calibrated level. Others are sealed in place at the factory & you have what you've got.

Stabila spirit levels are a well known brand sold in most hardware stores (the yellow ones). Brick layers & builders use & abuse them frequently, often throwing them in the back of the truck with all the hammers, timber & scaffold. It's a miracle that some walls go up anywhere near plumb ? I suppose it's a tribute to the toughness of many spirit levels.

How's this for a challenge Steve - Casino cruise ships. I was talking to a dealer at Star City (now rebadged as The Star) & he was telling me about his stint as a dealer on a cruise ship.  I laughed & said what about rough weather ? He said with the stabilizers out mostly it's ok. If the swells build then they stop roulette. I then said to the dealer take a rather calm day with a regular mild swell what about the tilt of the wheel ? He said they just wear it as most of the dumb cruise "CLIENTS" lose their money regardless. 

So with that regular mild swell in mind how would it work out with a roulette computer working out the tilt with an inbuilt digital level plus co-ordinating the wheel speed & the ball speed ? Now the rise & fall of the ship would really put some bias on the spins. Also what would you say when questioned by the dealer about your mini spirit level on the table ? How about something like, "It's my seasick meter when the red light flashes I'm off to recycle todays dinner". 

[amk]

Steve,

From my understanding this NVML report does not discuss the roulette computer device, there are two articles on the net. This one only deals with the wheel being tilted to a certain degree by hand and seeing which sections have a bias after 176 spins and what the chance of this bias is.


Thanks ausguy for the indepth posts, really interesting insights. I had a few questions.

Will all roulette wheels show a slight tilt angle around the .1 degree or do they read 0?

Would you regularly see a .12 degree tilt or closer to .1?

[Steve]

Ausguy:

One thing I don't quite follow is your mention of you recalibrating your newest wheel & the dominant diamonds are different every time. The only reason I see for that is that the physical levels are imperceptibly different from the old position to the new ?

Yes thats correct. Its near impossible to recalibrate a wheel so that it is has precisely the same proportion of hits between diamonds.

I'd guarantee with 99.9999% certainty that with a vernier precision optical level that upon marking your premove wheel level checking points that after relocating your wheel that those premove marks could be set in the same horizontal planes as the original position. I'm not talking about height off the floor, in this instance that's irrelevant. You could move from L1 to L33 in an office tower & still level your wheel in the same plain as it was at L1.

I'm sure it can be done, but not with what the casinos themselves use either. The wheels are extremely sensitive to changes to dominant diamond changes if the wheel is in good condition. If they are in bad condition, then the dominant diamonds tent to stay the same. I can get the diamond hit ratios to be about 90% correct (so they are the same every time).

Also steve how do they "finalise" the ball track ready for use ? Do they CNC machine it in situ to ensure a perfect circle ? or CNC machine it prior to installation ? Is it then a snug fit in the track wooden rebate or glued/fixed in some other way ? 

I'm not sure about this. But I know how they are often repaired. They use a very thin epoxy and rotate the entire wheel on a rotating table, so when it hardens the ball track is more symmetrical.

I don't see how what the wheel is made of influences the path of the ball from when it leaves the ball track, hitting or missing a diamond(s) prior to arriving at the rotating wheel/pocket ?

The ball track is far from a perfect surface. It only looks like it in a loud casino from a few metres away. But get up very close to the ball track and you'll see loads of tiny cracks and imperfections.

Cut and paste gets the symbols in.

And 0.2 degrees is not that different from what NWML used for their levelling. They used one with the vial too though, as do I but to level my wheel I do not measure it from the ball track. I measure from the top of the wheel, then use custom-made screw supports on the wheel base then very marginally adjust the supports until I get even hits on all diamonds, or at least as close as possible. Then I mark the points where the level sits, and get digital readings on both axis. The accuracy may be only 0.2 degrees, but the repeatability is 0.1 degree which is more important in my case than actual degree accuracy.

I have several different levels and find the digital easiest. Even with my perceived precise placement of the actual level, there is still variation with any level because the actual surface is not flawless from 1mm to the next. Its the same case when measuring from the ball track with the triangle used by casinos. Its not as simple as placing it on top and adjusting. There are variables that make it different every time, at least to some degree.

Regarding cruise ships, certainly they influence the dominant drop areas. But not as much as I first thought because consider the variation on an individual spin due to waves. Consider this variation to be random. But the core cause of the dominant diamonds to begin with doesnt change from the waves. So while you cant know exactly what the dominant diamond will be on an individual spin (due to variation of variables), the averages remain the same. So it doesnt really change anything.

Consider if the wheel was vibrating at a high rate. There would be no changes. The ship swaying is no different except the vibrations are at a very slow rate. So while you may think it randomizes spins, it doesnt. To back this up, many players focus on boat casinos and they find it no different.

AMK:

Yes there are two parts to the report. See if yu can find their conclusion which as I recall is basically "yes the device works, even with minor tilt, but on different days, the ball deceleration rate was different and this made it impractical". This is one of the reasons why most computers arent useful in real conditions, and will only work with heavily tilted wheels where the wheel defects ensure air pressure changes dont have such an impact.

The report focuses so much on tilt when the real cause of dominant diamonds is just as much, if not more, ball track defects.

[Crapshooter]

The question of hemispheres is interesting and really worth giving some thought to. I am sure there is something to the balance of the earth which can cause differences.