-
This topic is writen by Bayes at VLS.
-
This is a short review of a book called "The Fortune Maker Tackles Roulette" (available on Amazon @ $13.99).
You'll need to buy the book to see the details fleshed out, but the basic idea, and the heart of the method, is the following:
Collect some spins, the author recommends at least 1000 from each wheel, and for each number:
calculate how much you would have won (for all spins) betting the previous number and its neighbours - for 2 neighbours, then another 2, then another 2 etc, so you are doing the calculations for betting on 3,5,7,11,13,15,17,19,21 numbers. e.g. if your spin sequence from a wheel started - 23,4,13,29... you would compute your winnings betting 3,5,7 etc numbers centred around 23. Number 4 came up next, which would have been a loss for all, because even betting 21 numbers this would encompass a sector from 25 through to 9, which doesn't include number 4.
Having done all this, the next step is to make a comparison between the amount of numbers you would have bet, in terms of how much each would have won. The amount of numbers which gave the best performance will be the amount of numbers to bet (the last number spun and its neighbours). You then play that wheel using those numbers.
Obviously, to make a direct comparison, you need to "equalize" the results, because if you're betting 15 numbers then this is going to cost you more than if only betting 5. ie; 5 numbers @ $1 per number = $5, and 15 numbers @ $1 per number = $15. So in order to compare the two, you need to raise the bet level of the 5 number strategy to $3 per number ($3 × 5 = $15).
The multiplier is actually referenced to the maximum number of number you would play ie; 21.
So,
21 numbers @ $1 = $21
19 numbers @ $1.1053 = $21
17 numbers @ $1.2353 = $21
15 numbers @ $1.4000 = $21
etc.
Do this for each number you've collected, and record the results. You should then be able to determine, on the basis of a direct comparison, which (if any) number strategy is best for that wheel - whether 3,5,7 etc numbers should be played.
The idea is that IF any kind of bias is present, this analysis will pick it up. It is recommended that even if you don't collect 1000 numbers, you should always do what he calls a "dry run" before playing any wheel. This means collecting 37 numbers (one "cycle") and doing the above analysis. If any number strategy stands out as being particularly profitable, then by all means play it.
The author emphasises that it may not be the case that you have found any bias even if a strategy is suggested by the data generated by the dry run (37 spins is too small a sample) - it may just be "luck". However, there are also detailed explanations of how to closely monitor your results while playing, so you know when to quit.
It's not necessary, the author maintains, to know the cause of the bias, and states that:
He defines the effect of a bias as:
And this provides the rationale for the method described.
So that's basically it. Obviously, to do all these calculations by hand would be incredibly time-consuming and tedious, so you'll need to set up a spreadsheet or write a computer program. The author gives a url where you can download tools, but this seems to no longer exist.
The author also recommends studying the data to try to find other kinds of bias which are not necessarily keyed to the last number played. For example, you might notice that the number 5 has hit twice in a cycle (37 spins), and it's conceivable that the ball tried to get into this pocket on other occasions but didn't quite make it. Examining the data might reveal this to be so, in which case you might decide to target 5 and surrounding numbers. The above scenario might have occurred for more than one number, so you could end up playing more than one sector.
It would have been better if there were some guidelines on how to find statistical significance in the data, instead of merely telling you to pick the sector which won the most, which isn't very rigorous. Also, if you can't determine the cause of the bias, you're on shaky ground, but this isn't easy. Presumably, if you can do it, so can the casinos, that's why they have regular maintenance schedules.
However, in spite of these limitations, the book is worth reading. There are chapters on money management and detailed charts showing, for each number strategy (whether playing 3,5,7 etc) whether you are in the "win zone" and if so what is the strength of the bias.
The book now seems to be out of print.
But thanks for reminding me, I was going to write a tracker for it. :)
Quote from: Bayes on Sep 23, 08:40 AM 2012
The book now seems to be out of print.
But thanks for reminding me, I was going to write a tracker for it. :)
Hello Jules,
Look forward to see and use this tool.
Cheers
John Rotterdam