|
|
Post by mrsonde on Oct 20, 2018 0:45:48 GMT 1
For the past three weeks I've been experimenting with a variation of a Martingale system at roulette. This goes back to an old debate I had at college, but that doesn't matter. The interesting point is that by using this system I've won just over three thousand pounds - not quite a thousand a week, as I upped my stakes from a tenner to a hundred in the third week, but you get the idea.
The "system" is very simple. You cover your losses. The variation is that you don't start to do so until a run of at least five in a row. Mathematically, this is guaranteed to make you a winner, in the end. The drawback is that if in the event of a catastrophic long run you'll exceed either the table limit or your available funds. To avoid this drawback, you only start betting on a run of five. Or six, or seven, depending on how timid you are. And - this is crucial - you drop out, or at least drop down to your initial bet, after a losing run of five (making, if you enter on five, a loss of 5 times your initial stake.) This ensures you will never lose more than you've won.
The question is - will you win more than you lose? As I say, I'm up about three grand so far, after three weeks. I can't see how I can lose enough to ever cancel that win out - statistically, such a run of runs is astronomically unlikely.
I'm not really interested in the money-earning side of it, attractive though that is, despite myself (I wish I'd thought of it when I was a kid, and needed the money.) What's interesting is the logico-mathematical side of it. Is the Gambler's Fallacy really a "fallacy" under these conditions? And if so, why?
|
|
|
|
Post by mrsonde on Oct 20, 2018 2:36:46 GMT 1
You can try this yourselves, without any risk. Every night Supercasino run an online roulette on C5 - or you can download their app and do it all day. The system is simple: After a run of red/black, even/odd or low/high, place your bet. If you lose, double your bet on the same placement - and so forth. Until you come up against your limit - I use a five times run. Then you drop back to your initial stake (a hundred quid in my case, though I started with ten.) Try it for a few hours, see how you do - you don't need to follow it constantly, just check in every 15 mins or so, as they show the last dozen or so results.
You'll find you win - or you would have done, if you'd bet money. I mean - give it a fair roll. Once I was down a hundred - but that was against being up over thousand.
If anyone's interested in the mathematical analysis of all this, I've done some work. Let me know.
|
|
|
|
Post by aquacultured on Oct 21, 2018 0:09:12 GMT 1
And they pay you for this ad, do they?
|
|
|
|
Post by alancalverd on Oct 21, 2018 9:45:26 GMT 1
"All aeroplanes bite fools." Roulette wheels likewise. A casino is a business, not a charity.
|
|
|
|
Post by jonjel on Oct 22, 2018 16:21:57 GMT 1
You have reminded me of a tale I told only a month or so ago.
As a young man I worked with a bunch of people who did the pools, looking for the massive pay-out that always eluded them. The money we 'gambled' was peanuts, but it was a bit of a game really.
So, I chatted to my pal and we decided it would be a lot more fun to do a four or five horse accumulator every day. And it was, sneaking off to the loo to hear the results. We won a little, but as always, lost more than we won, but it was not by any means serious gambling with money we could not afford to lose. In todays money maybe a tenner a week between us.
This went on for a while and then I came up with a plan.....
Why do gamblers always lose? Because they are greedy so the idea was that we would set out to win a certain sum every week, and if we won it, we stopped until the Monday. I remember we set out to win £20 a week, a lot of money then, roughly a skilled mans wage. The plan was simplicity; put money on the first horse, then if it loses put money on the second horse, enough to win the 20 as well as what you had lost on the first one and so on, until you had won your twenty quid, then stop.
We decided to use the racing journalists in the paper 'tips for the day' kind of thing working on the theory that if they were not right some of the time they would not have a job. And we thought we had better try this out on paper for a few weeks first. And we had always won our £20 by the latest on a Wednesday, and more often than not by Monday night. My mate was dead keen for us to do it for real, but I persuaded him - one more week.
That week it rained, hailed and snowed I think and did everything. I am sure a lot f favourites went to the dog meat factory. We would have won our £20, but not before Saturday by which time we would have invested several hundred pounds, several hundred more than we had. And also the amount we would have been putting on would have shortened the odds, making the amount we had to gamble even more.
And the other flaw of course is, once you start winning regularly the bookies stop taking your bets. Remember the old expression, you don't often see a bookie on a bike.
|
|
|
|
Post by alancalverd on Oct 22, 2018 17:08:41 GMT 1
Why do gamblers lose? Either you are playing against other gamblers, and there are more of them than you, and/or the odds are rigged either by the flow of bets or by the structure of the machine you are betting against.
Who becomes a bookie or a casino agent? Remember the kid at school who offered to hold your coat whenever you got into a fight? Same kid grew up to be a bookie, estate agent, politician or casino owner - whatever profession, he never staked his own money on an uncertain outcome.
|
|
|
|
Post by mrsonde on Oct 22, 2018 17:43:44 GMT 1
Why do gamblers lose? Either you are playing against other gamblers, and there are more of them than you, and/or the odds are rigged either by the flow of bets or by the structure of the machine you are betting against. That wasn;t what I asked, Alan. I asked why is the Gamblers' Fallacy a fallacy, when in fact under the circumstances I laid out they win? You can verify that you will win under these conditions, as I said. You can do it with a pack of cards, or by rolling a dice, if you wish. It would take less than an hour, I assure you. Now, the original debate I had in college I referred to was what this implied - my argument was, funnily enough, that it was a definitive argument against the Solipsism that you and Aqua has so absurdly accused me of. Proper solipsism, that is, rather than your sloppy misuse of the term. The Gambler in this case will always win - and it is denied by no one that they will always win, and must always win (which is why every casino has a table limit, to insure themselves against an unmodified Martingale system) - because the world exists, and runs according to the laws of probability. These are invented by nobody - they exist. |
|
|
|
Post by mrsonde on Oct 22, 2018 17:44:05 GMT 1
And they pay you for this ad, do they? An ad? For what? |
|
|
|
Post by mrsonde on Oct 22, 2018 17:45:23 GMT 1
"All aeroplanes bite fools." Roulette wheels likewise. A casino is a business, not a charity. Try it. Supercasino have now closed my account, as jj intimated. After less than four weeks. I repeat - I'm not asking you to believe me. Try it yourselves, at no risk whatsoever. I'm not interested in the money-winning side of it, either. My account closed with a gain of £4,350 in a month - I've donated it to the charity my missus runs. What's interesting about it is the philosophical implications: mathematical proof that an external world exists, and runs to an impersonal order. This is - unless one takes Wittgenstein;s language argument seriously; which I do, but others such as Jean do not - the only such proof that we can formulate. |
|
|
|
Post by alancalverd on Oct 24, 2018 0:49:08 GMT 1
BeforeI go into the mathematics, please explain what you mean by "a run of five"
|
|
|
|
Post by jonjel on Oct 24, 2018 16:44:48 GMT 1
I think we have the reincarnation of Horace Bachelor on this board.
Dept 3, Keynsham, Bristol.
|
|
|
|
Post by mrsonde on Oct 24, 2018 22:33:25 GMT 1
I think we have the reincarnation of Horace Bachelor on this board. Dept 3, Keynsham, Bristol. I'm not advocating you gamble. On the contrary. You can't lose. Try it and see if you don't believe me. As I said, the interesting question is what this necessarily implies about the world. As I also said, you don;lt have to actually go to any sort of casino to do this experiment. Toss a coin, use a couple of packs of cards, roll some dice - download a Monte Carlo simulator if you trust that sort of thing more than the real world. Wait until you deal five red or black cards in a row - place your bet on the next turn. If you lose - the sequence continues - then double your bet on the next turn: and so forth. Eventually the seqeucne will break, and you'll cover your losses and win your initial stake. Casinos guard against this mathematical certainty by imposing a table limit - both minimum and maximum. Every now and again a catastrophic unlikelihood will occur and a sequence will occur which you won't be able to cover. Thirty years ago when I debated this in college the longest run ever recorded in any casino was 13 - that may have gone up in the iterim, I don't know. It doesn;t matter - because under this modified Martingale you've defended yourself by dropping out after the limit you've set yourself. I dropped out at 16 times the inital stake - I dropped back to a single bet, and took the loss on the chin. The point is - these runs are mathematically unlikely: that's what you're "betting" on. Over a number of such occurrences, you'll nevertheless be up - because the world is regular. I'll make you a no-lose bet, if you like. Try this for an hour with a pack of cards, or a coin. I bet you will be up by the end of that hour by two units - or, after two hours, three, which is closer to the mathematical prediction. You can call that unit a penny or a quid or, what I quickly used in the real world, a hundred. If it turns out, highly improbable though it is, that you're down instead, by however many imaginary units, I'll owe you a tenner. Agreed? |
|
|
|
Post by mrsonde on Oct 24, 2018 22:36:33 GMT 1
BeforeI go into the mathematics, please explain what you mean by "a run of five" Toss a coin. If you get five heads in a row, that's what I mean by a "run of five". It's like a run, five times. A sequence of events, with equal chances of occurring or not, five times in a row. In roulette you have three such possible sequences. You have others, which are two to one chances, which you can also use if you're mathematically inclined, but I've kept things simple. |
|
|
|
Post by alancalverd on Oct 25, 2018 9:34:38 GMT 1
The gambler's fallacy is the belief that future events are determined by or predictable from past events. A run of five in roulette is no indicator of the next result, though it might suggest that the wheel is biased.
Thus it doesn't matter when you start (assuming the wheel is fair): Martingale works in theory because the sum of 1 + 2 + 4 ...2n is 4n -1, so if you continue to bet ad infinitum on a fair wheel, you will win at least 1. But anyone with a degree in philosophy would have noticed the terms "ad infinitum" and "fair", neither of which applies in a casino. There is a house limit, and a house zero. Full marks, however, for using the entrepreneur strategy: cash in when you are winning, and start again - venture capitalism at its best.
A run in a card game is obviously different. If you deal 13 consecutive blacks from a full pack, the probability of the next card being black is 1/3, not 1/2. Thus card counting can improve your winnings in the short term. In the long term, you will be banned or shot.
|
|
