Testing Roulette Strategies/Systems

Xander · November 24, 2015, 08:28:14 PM

Note to strategy testers:

When you test your methods, you must carefully define the testing parameters. You can't hop back and forth and change the rules as you move along with your demos. You can't say, "If I would have done this then I would have won that, or if I would have had a stop loss here, then I would have won that." Such testing formats are annoying, pointless, and it's nothing more than curve fitting. Below is some advice that will make the process more enjoyable, valuable, and less annoying to forum members.

1. List the rules in advance.

2. If you're going to have a stop loss, then define exactly what it is.

3. Carefully explain how you're selecting the numbers on which to bet. Provide a link to a specific page that describes the rules of the system, not several pages of rambling posts.

4. Test statistically relevant sample sizes. Testing just 50,100 or even 200 spins at a time is silly. You can't look back and say that such a small test is proof that you should have made any changes whatsoever!

5. Get to the point. Less is more. Don't ramble on and on and on. People are annoyed by long rambling posts, and they will often ignore them. Also, don't make up bullshist terms in a vein attempt to make your system sound more scientific than it really is. If you must use jargon, then define the jargon terms at the bottom of each post.

6. Run out of sample tests. One positive result isn't enough.

7. If you change the rules, then you must restart the testing, and restate the new parameters.

8. If you're going to put your system out on a public forum, then you should be willing to accept some criticism. Don't block comments once you've presented the method...it's weird.

Best of luck.

-Xander

Mike · November 24, 2015, 08:46:50 PM

Good advice. Not addressed to anyone in particular, of course...

Xander · November 24, 2015, 08:51:33 PM

QuoteGood advice. Not addressed to anyone in particular, of course...

Correct. I will try as hard as possible to stick to my own advice. I'm probably a terrible offender.

zelx77 · November 24, 2015, 09:32:35 PM

Amen!

Xander · November 25, 2015, 01:15:19 AM

Great avatar zelx77.

greenguy · November 25, 2015, 04:23:04 AM

That's not a great avatar, this is a great avatar.

Xander · November 25, 2015, 04:25:23 AM

Yes, yes it is indeed.

Gizmotron · November 25, 2015, 07:29:17 PM

What you call "curve fitting" in research experiments in known as confirmation bias.

So if I create an algorithm that stands at least a million spins, in profit that is, I must disclose my secrets in order for my trade secret to be true. Is that what you are demanding? Creating an AI algorithm, even one focused so tightly on such a small stream of data is no easy accomplishment. I don't feel like giving away that much work and that much research, more than twenty years and a lot of pay checks. With all due respect.

greenguy · November 25, 2015, 08:45:17 PM

Good for you Gizmo,

But what about your legacy?

Jimske · November 25, 2015, 10:01:53 PM

Quote from: greenguy on November 25, 2015, 08:45:17 PM
Good for you Gizmo,

But what about your legacy?

Hahhahaha. That's right. Few people want to go to the grave without recognition. Heck, few can keep such a secret!

Xander · November 25, 2015, 10:15:57 PM

QuoteSo if I create an algorithm that stands at least a million spins, in profit that is, I must disclose my secrets in order for my trade secret to be true. Is that what you are demanding? Creating an AI algorithm, even one focused so tightly on such a small stream of data is no easy accomplishment. I don't feel like giving away that much work and that much research, more than twenty years and a lot of pay checks. With all due respect.-Gizmotron

Gizmo,

I'm guessing you're playing the "I've beaten a millions spins game" even though you're not really betting on a million spins. Reverse engineering to a million spins is meaningless.

Bragging that you have a winning system, and then playing the secret squirrel game is also weird.

Furthermore, I promise you that nobody is interested in it.

Gizmotron · November 25, 2015, 11:56:51 PM

Quote from: Xander on November 25, 2015, 10:15:57 PM
Gizmo,

I'm guessing you're playing the "I've beaten a millions spins game" even though you're not really betting on a million spins. Reverse engineering to a million spins is meaningless.

Bragging that you have a winning system, and then playing the secret squirrel game is also weird.

Furthermore, I promise you that nobody is interested in it.

Two can play that chicken before the egg game. I've shared more than I should have already. You just don't get it yet. I did it because I knew people here have to go through eliminating things they are more interested in first. That gives me plenty of time to hold back.

Ever heard of Edward O. Thorp & "Beat the Dealer?" He had a different agenda than I do. He had simple arithmetic on his side too. Boy, they listed once he published his proof. I'm betting on it being too late for you experts. It's my revenge for putting up with flat-earthers. I control when people take it serious, not you.

Xander · November 26, 2015, 02:08:28 AM

Gizmo,

Beating a million spins means that you actually have to bet on a million spins. If you're just using triggers and watching the spins go by without betting, then it really doesn't mean anything.

QuoteTwo can play that chicken before the egg game. I've shared more than I should have already. You just don't get it yet. I did it because I knew people here have to go through eliminating things they are more interested in first. That gives me plenty of time to hold back.-Gizmotron

Gizmo,

We simply don't care. Your claim is weird, and nobody wants to send you any money for the system's secrets.

Xander · November 26, 2015, 06:44:27 AM

QuoteI notice Xander and his ilk have recently re-emerged on the Forum. X is certainly not welcome here as his personal ill informed abusive tirade against the late and eminent Dr Martin Blakey from Melbourne was a disgusting performance of vitriol for which he never apologised to friends or family of the brilliant roulette professional who was foremost a true gentleman who remained totally secure in his self belief and focus after 40 years of professional play that was his primary income to support his family. How many of us could take on that responsibility?

So to Xander and similarly closed minds, do not bother coming here as your posts will be deleted.

XXVV,

Please take the time to read the list above. Oh, and by the way, pay special attention to numbers 5 and 8 on the list.

Cheers,

-Xander

Mike · November 26, 2015, 04:09:37 PM

Quote from: Xander on November 24, 2015, 08:28:14 PM
8. If you're going to put your system out on a public forum, then you should be willing to accept some criticism. Don't block comments once you've presented the method...it's weird.

Far better not to actually post the details because then nobody can show that it doesn't work. Just stick to no. (5) and make grand claims. Someone's bound to pm you and make an offer.