Pattern recognition, sequences in random numbers.

[Chance]

The above example is very exagerated.(except for the online rng yesterday) but i wanted the math to be understood.

[Still]

Most rngs use methods of things like time between key strokes x 1000 to come up with a number added to some seed. But, what about craps,

Reminds me of a time, 15 years ago, when i programmed a craps method into VBA, the basic in Excel.  I came up with some variation that did indeed appear to have an edge, and was confident enough to try it out in the casinos around Lake Tahoe, only to be disappointing that i was losing at the expected rate.  So i went back to the drawing board and looked at my random number generator code, and added a time delay between rolls of the "dice".   At that point, it broke the system i thought was working.  The charts, from before and after this tweak looked nothing alike.  It appears the system had exploited some non-random pattern, and when the stream became more truly random, it broke. 

I would add, the original craps system was purchased from a classified ad in USA Today for about $200, promising a 2.17 (if i recall) edge in craps.  I didn't trust it when i read it, so tested it first in VBA. That's what happened. Neither his method, nor any tweaks i had made worked, either in the first place, and especially after the random generator was tweaked.

[Chance]

Sounds like you have been at this a while. VBA was still new 15 years ago. You have to be cautious. I was skimming numbers off a mechanical roulette wheel, the sectors were predictable, but it was in its low level mode without players, when someone sat down it was another story.

After charting 5 games after a long high random sequence (a run of random numbers above normal for that wheel) start betting on sectors to repeat. Buy in with 140 chips (a full flower) one chip on each of the 9 numbers of the sector that has just repeated, 1 chip on 0 the other on 00. Keep placing a chip on the sectors that have just hit. You are betting on the bias of the wheel will take over and a sector is going to repeat. Like the card wheel I was playing in Oklahoma it was a no brainer playing, guessing sectors to repeat. When you have to make a decision of which way it will go, the bias says repeat. When in doubt repeat….much more later.

If you are playing a wheel that has a lower bias (higher entropy) you have to shift from placing bets on sectors that have hit to review the math of numbers followed by what number. Practice zooming out. For instance, if you have 1,3,4,1,1,2,4,1,3,2,2,3,4,1,3,4,3,3,3,4,? I could go into the math of this stream but if we are looking at a highly random wheel what should we bet next You. I do not want to go into detail until later but you could run the same alg once again on the conditioned stream, or another…more later (remember on a random stream no matter how many times you run an alg it should stay random. If it does not and you know the wheel is random make bets it will repair the anomaly and become random by corrections.) or just bet that after the 4 it will hit a 2. Why not a 1….it has done that the last 3 times. Why not a 2……..high number of 2s have hit. Why not a 4…….high number of repeats have already hit. Once again we are playing the math and biases. If it hits a 1 stop playing and wait for the next time it hits a 4 and continue with the logic of the math of the last 20 or less numbers like we just did. If it hits a 3, bet sector 1(most probable)…. Much more later

[Taotie]

Actually progression can be a good thing...

Even flat betting a percentage of bankroll is a progression so I suppose that's right.

[Taotie]

1 chip on 0 the other on 00.

Hedge betting on the zero's is another red flag.

If you skim your numbers then run your algorithm then bet a section of 9 numbers, that means not counting the zeros there are 27 numbers that can cause the bet to lose. Why would the zero's be any more likely than any of the other numbers to save the bet?

[Chance]

An old timer told me that and i have always done it. The math says don't do it......agreed. with a single 0 wheel it give you it gives you 37 a prime can't divide the wheel into sectors,  with 38 its cumbersome, so either always or never bet on green.......that is just my opinion.....

[Herby]

same alg once again on the conditioned stream, or another…more later

Will be interesting to see your other algs.
You are a very humorous person so you wont be sad to read that some users here already found out that your alg was already discussed here, different appearing but quite the same essence.

I'll show the thread later for not to interfer your ideas.

[Chance]

Great, how did it do and if it has already been discussed i don't need to explain further, there are some very smart people on this board. The bottom line in all this discussion is, can you make a random stream using fractal geometry laws.  All else is preferences on how to use the inverse for prediction, to find sequences, biases in random devices and patterns. I have not searched threads to see if the person i learned this from has posted here before.

[Drazen]

What if we have perfectly "pure" randomness and we discover some bias there? Hope that makes sense.

Can any of your algorithms be used for that?

Thanks

[Herby]

The topic was almost only about what you call alg but for other chances.
Nothing about fractals or geometric laws of fractals.
The results appear like a mirror to your results, so maybe it's not easy to see the connection.

You are not suspected to know this thread.

[Steve]

What if we have perfectly "pure" randomness and we discover some bias there? Hope that makes sense.
Can any of your algorithms be used for that?

In any game you only need a slight accuracy increase. What we should be looking for is proof of concept, like significant testing that shows the odds are changed. I haven't seen that yet but it doesn't mean the principles don't work. We just don't know yet. At least it doesn't appear to be the same old known losing approaches.

[Taotie]

An old timer told me that and i have always done it. The math says don't do it......agreed. with a single 0 wheel it give you it gives you 37 a prime can't divide the wheel into sectors,  with 38 its cumbersome, so either always or never bet on green.......that is just my opinion.....

To me it seems strange to go to great lengths of random reckoning, and then just start plonking chips on two green pockets for no other reason than they are there, or some old timer told you to.

I'm not saying it's wrong to hedge bets on the greens, just that if you are going to hedge bets at all, then you probably should have another whole level random reckoning for the hedge bets. Or find a way to cost effectively hedge every number on the wheel.   

[Chance]

Taotie, agreed. You could run the same fractal  geometric logic individually with 0 and 00 but that is zooming in more than i comfortable with. I have done extreme odds advantage play with 0s before, it was successful but mathematically i don't thinknit is reliable.

Drazen, using one of my alg i (close your eyes programmers you may not like what i write next) can make a perfectly random stream based on geometric fractal laws, like what we have been discussing here. Doing that i can hide encrypted bytes that are impossible to find.....like an electronic one time pad hidden in a purely random stream. That should also encompass your question.....almost off topic sorry mod!

[Herby]

Looks like this was the last Chance.

[Chance]

Ohhhh no. I will be around for a long time, i have a project i was working on. Will post  today. Just wanted to clarify the math on a stream  of any length.

Singles or 1s happen .56% of the time
                   2s               .14%
                   3s               .35%
                   4s               .00875%
                    Etc. N/4