Monday 7 August 2017

Why Roulette Does Not Need to Cheat

I had one of those highly enjoyable, loud, blistering arguments in the pub Sunday evening about roulette and its television presentation. I noted with them that ITV's presentation has an automatic wheel and one presenter often looking silly, and Channel 5's commercial presentation of SuperCasino used to have a presenter and a "trained" spinner of the precision craft-built wheel. [In fact it varies between one and two.] I have never gambled on any and I never will, and for the reason why see my website.

But what my two friends insisted upon, and I know another of our drinking party on Tuesday agrees with them, is that it is all a fiddle. I insist it is not, because it does not need to be.

Their contention was that there is an algorithmic computer that instantly gets the biases of the bets on a throw to come, and picks the video to produce a number outcome to maximise profits. That this suggested fraud would finish their broadcasts, end their credibility, and lead to a hefty fine matters not, they said, because the television history of game shows is that these phone in and similar contests have been fraudulent, they were found out, they paid the fine, and simply carried on.

My argument is that there is enough profit in the structure of the game guaranteed, and so little to be gained by the elaborate set-up needed to cheat that it isn't worth the additional effort. I was told that I am not cynical enough, and that there is never enough profit for capitalism. But, against that, I said if the profit is guaranteed then planning can take place. All of these 'offers' are for new players, and there is every reason to have new players: not simply more profit, but more stability; these offers come with restrictions (to come within the margin of profit - 'free money', so called, is somewhat like the old Truck money in shops, at least for long enough).

One friend mentioned the button trick in the real casino to bias the result. But that's in a game where there may be say 50 players around the table. The bets will vary from game to game to a visible bias. And even then they don't. We agreed on a playing number of 100,000 per random number generation (I cannot discover any statistic): at that level the numbers are so great that everyone ends up betting on all the numbers more or less evenly. And if they just about don't on one number generation, they do over a small number of throws of the ball over the wheel.

That's the point, and the only point. The more players, the more certainty for the casino providers, and it is simply a means to create money. With huge numbers on just one table, fiddling a result is ridiculous. That was and is my essential argument.

There are all sorts of gambling fallacies that people come up with to suggest cheating, and my friends are not stupid enough to have mentioned any of them. One is that a string of say reds (or any other characteristic) makes it more likely that the black (other characteristic) comes up next time. The likes of SuperCasino do participate in this fallacy, calling them "hot numbers" and "cold numbers". If they were really hot and cold, then the wheel really would be wonky. Probabilities have no history: they are all future based.

Another fallacy is the strategy to win, which wasn't suggested, but which I volunteered in order to make a point. In a celebrated evening in a church hall raising funds many years ago, I took my roulette set. Children were losing money and going away. So I said, to a few, I could improve matters: bet on evens only, or reds only (all 50-50) and be moderate. When one wins a penny or twopence, next time play just the penny or tuppence, but if one loses then double up, and soon there's bound to be a win, and the money is restored and it will continue to build up. Soon children started to gather round because they could see others winning. Parents became interested too. But one by one, the children (their parents, watching with interest) ran out of coins, ran out of money. And when they did, I told them to learn a lesson: "You cannot win over the long run. Never gamble."

Something not mentioned at the pub. There is a book that I think has a million of five number sets, randomly generated. It is a very boring read, apparently, but does have some highlights. One is a sequence 12345 and another is 00000. They appear a percentage number of times. The book shows that when people complain that a number sequence is not random, it usually is. I also did not mention the fallacy that the universe is so critically and necessarily specific that it must be designed, otherwise it would collapse in on itself or fail to function. That's easy. I have a rule for this. Suppose a pack of cards has to be in numerical order for hearts, diamonds, clubs, spades and jokers in that order for everything to take place; if not all collapses and vanishes. But here we are, and we see that order of cards. We are bound to say, it MUST have been arranged: but no. The cards were dealt, from random arrangements, over and over and over again, collapsing any future every time (other than dealing cards). Suddenly, once, after billions of card orders, the magic arrangement appears and everything can go ahead and be stable. Stability is also evidence that chance happened. Indeed the longer something goes arranging and trying again, the more chances exist for the infinitesimal outcome to happen - and once it does, that's enough.

My friends attacked "economics" in all this (i.e. capitalism). No no, economics has this covered. What I did say at the pub was that roulette on screen and online generates profit through negative utility. Profit ought to come about from increased utility, cost but meeting need, but there is only a small 'entertainment pastime' utility in a gambling pastime (and it has a very steep marginal utility curve). Thus creating a gambling supply, e.g. a casino, as a form of economic regeneration is false, because profit is based on misery. Indeed, a want that is addiction is a false utility -  a negative demand. It deprives people. For all the winning names that the likes of SuperCasino display (something probably random or first in), the same sort of list can be shown of losers. And the probabilities and mass numbers mean that there are always more losers than winners, and each throw generates profit from losers who could know better.

Oh, SuperCasino says that it tests each precision craft-built wheel for random number generation. Number outcomes are shown being tested. Well, that is misleading, because a sequence of 11111 is as likely as 49318. And so on.

No comments: