One of the rarest numbers events on the board might be 1 number appearing 4-5 times in a row, but since this isn't a good example for describing my method, let's describe another equally rare event: let's say that a group of 18 numbers could appear 37 times in a row over a billion spins (I don't know the exact amount; it's also possible that there could be a maximum for any amount of spins). Now if that was to happen when you first arrived at the table the expected fluctuation for any number should be no more than 1 if the law of large numbers had its way (all 37 numbers should have appeared exactly once in an average run of 37 spins); in fact, this is what the forces of nature are always trying to do - keep the fluctuation to a minimum in our chaotic universe.
So we now have the average expected fluctuation for every 37 spins = 1 or 2.7%
18 numbers are each -37 compared to the average fluctuation (percentage?) = maximum fluctuation
We would bet the table limits of 2,000 on each of the 18 numbers = maximum number of chips
So by keeping a record of number totals, average fluctuation each spin, how each number compares to the average fluctuation we can determine (based on the above example) how much to bet individually on each of the 37 numbers each and every spin!
A few corrections to my post:
*The number fluctuation should be based not on the average, but on the most hot number
*Figuring out a rare event for which to bet 2000 (and base all other calculations on) cannot be based on numbers coming in a row, but must be based on the difference between the hottest and coldest numbers.
Attached is the spread sheet for entering number totals (in the red-black-dark green boxes) and calculating chips to bet on each number, based on fluctuation.
You can also change the bankroll figure should you never have enough money to place up to 35 x 2000 chip bets.
I am going to test over 1 million spins to figure out the rarest event in terms of number fluctuation and then I'll let you know what value to enter into the yellow box (right now it's set to 0.973684211 but it needs to be somewhere nearer to 100% methinks)
Oh, and the numbers boxes can't be set to 0, so reset them all to 1 if you start a new session (a total of 5 appearances would really be 4 but set it to 5)
Some fluctuation tests on red and black over near 1 mill spins - this system could easily be adapted to include any sections of the board besides numbers!
Number Fluctuation results incoming:
Spin Min Max Gap
14 1 2 1
18 1 3 2
49 1 4 3
Edit: this test was done wrong. Not looking for maximum gaps; looking for max ratio. Will re-run...
U are cray cray
Doing this on red-black was enough to transform most losing sessions into winning sessions! If it doesn't work on numbers then I guess I am crazy or there is something fundamentally wrong! ;D
Away we go...! :thumbsup:
Spin Min Max Ratio
14 1 2 0.5
18 1 3 0.666666666666667
49 1 4 0.75
77 1 5 0.8
113 1 6 0.833333333333333
150 1 9 0.888888888888889
Damn, my ratio calculation is wrong. It's currently based on the maximum, but may need to be based on total or a more complicated formula involving both.
Edit: OK I think it's solved - the formula should be slightly more simpler than it was - won't upload the new spread sheet till I finish the current test.
Tell a lie, we are actually looking for max gaps ratio - still expected to close even in the long run. Original spread sheet seems valid for now - though I've added some formatting and rounded down the wagers.
Spin Min Max Gap
18 1 3 2
49 1 4 3
77 1 5 4
113 1 6 5
135 2 8 6
150 1 9 8
259 4 13 9
300 5 16 11
306 2 16 14
383 4 20 16
514 6 23 17
654 10 28 18
741 11 30 19
Closing in on the maximum gap now... will it get to 700 even?
400199 10432 11045 613
401559 10469 11083 614
403264 10517 11132 615
403313 10518 11136 618
406458 10590 11210 620
406560 10593 11214 621
406612 10594 11216 622
406723 10598 11221 623
450546 11748 12372 624
450828 11754 12379 625
450934 11756 12382 626
451077 11759 12387 628
Damn, it's started going back up again! ::)
655403 17307 18058 751
655456 17308 18061 753
655522 17310 18068 758
655595 17311 18070 759
655643 17312 18073 761
655753 17315 18077 762
655803 17317 18081 764
655922 17319 18085 766
656270 17323 18093 770
656320 17324 18098 774
656535 17329 18104 775
658911 17386 18162 776
660958 17439 18217 778
It stopped breaking records after that for at least 300,000 spins, so that's our rarest event, folks! :girl_to:
660958 17439 18217 778
Will continue tomorrow! :yawn:
Does anybody understand this?
sorry no :twisted:
Spread sheet is finished now except for one more feature to be added - deciding when not to bet based on too small a fluctuation! (spread sheet seems to work by resetting all numbers to 0 instead of 1 now)
Here's the rarest event entered into the spread sheet to determine all other figures: (this on sheet 2)
(link:://s22.postimg.org/rmj8awzdt/rare.jpg)
Here's what happens when Number 1 goes ahead by 2:
(link:://s10.postimg.org/59f10kgop/image.jpg)
Here's what happens when Number 1 goes ahead by 3: (most numbers stay the same bet based on current average)
(link:://s22.postimg.org/f0dertmap/image.jpg)
Here's what happens when Number 1 goes ahead by 3 and Numbers 2 is at 2 (rest at 1):
(link:://s10.postimg.org/4gpe4ntkp/image.jpg)
Make sure you change the bankroll before you start playing via sheet 1! Bet selection doesn't get better than this!!
:o
If any of these 16 most fluctuated numbers were to appear (highlighted in blue) then the profit is HUGE! Even if the 17th most fluctuated number was to appear then the loss is only a couple of hundred - BUT since we've reached the maximum gap, the theory is that the ball has to land on the 2000 chip number (most fluctuated) in order for the laws of physics not to break down (or so the theory goes)! And the profit on that will equate to THIRTY THREE GRAND!
(link:://s28.postimg.org/h5e1fgpst/PROFIT.jpg)
BTW, I attached the wrong results on page 1 for Red-Black fluctuation analysis; attached is the correct files.
Long time lurker here but your post has urged me to reply.
I'm not quite sure what your spreadsheet is showing. Are you able to explain what is going on please?
Do we enter spun numbers on sheet 1 down column A? Is the bet selection all the ones then highlighted in green?
Quote from: sean43 on Nov 02, 04:11 PM 2014
Long time lurker here but your post has urged me to reply.
I'm not quite sure what your spreadsheet is showing. Are you able to explain what is going on please?
Do we enter spun numbers on sheet 1 down column A? Is the bet selection all the ones then highlighted in green?
Yep - you enter the total number of appearances in column A thereby keeping a tally. You then bet everything green (but am not sure whether to round up or round down or whether it makes any difference).
It's like an advanced labby on Numbers and reminds me of torrent ratios - after you download so much you have a new slightly higher ratio that you should maintain! Likewise, I think not betting some spins until the ratio has changed will yield even more wins. But I need to code it first to take it to the next level.
Quote from: falkor on Nov 02, 05:55 PM 2014
Yep - you enter the total number of appearances in column A thereby keeping a tally. You then bet everything green (but am not sure whether to round up or round down or whether it makes any difference).
It's like an advanced labby on Numbers and reminds me of torrent ratios - after you download so much you have a new slightly higher ratio that you should maintain! Likewise, I think not betting some spins until the ratio has changed will yield even more wins. But I need to code it first to take it to the next level.
Ahh right so total number of appearances of that number? So column A1 would be the number of times number '1' has appeared?
With a smaller bankroll would we simply not bet the green highlighted that are '0'?
Thanks for the reply
This system is based on a very simple method I discovered when testing red/black. I found that waiting for 10 reds in a row before betting did not affect win/lose ratio, but helped with cutting down progressions. The action that significantly transforms losing sessions into winning sessions (or more wins over losses; what every labby player is trying to achieve) is waiting for a fluctuation (or gap) in the totals.
I've attached 2 files on EC bets (3 Double streets vs. the rest of the board). Both are the same session of 1,000 spins.
Test 1 ended in more losses and a negative for all kinds of betting (though here everything is flat-betted for 1 step for the sake of this demo:
Flat-bet Grand Total: -25, Progression Grand Total: -25, Riding Progression Grand Total: -25
Total Wins: 487, Total Losses: 512
Test 2 ended with more wins and a positive - simply by betting when the fluctuation is 1 or more.
Flat-bet Grand Total: 12, Progression Grand Total: 12, Riding Progression Grand Total: 12
Total Wins: 172, Total Losses: 160
Exactly the same spin data is used for both! Now can you imagine how much more power one has when betting numbers using the same method?
Here's the final note with regards to developing this system further and beating roulette (if the system cannot already do that in the current state):
(link:://s22.postimg.org/rx1wq3icx/avg.jpg)
I dont quite understand what you are talking about as the majority of us, are there any examples you can show?
Quote from: sean43 on Nov 02, 05:59 PM 2014
Ahh right so total number of appearances of that number? So column A1 would be the number of times number '1' has appeared?
With a smaller bankroll would we simply not bet the green highlighted that are '0'?
Thanks for the reply
Exactly! If the green is less than 1 it means the fluctuation isn't enough to bet anything of value on (unless for some cases if you want to use 50p chips or we could use smaller amounts). And not all the greens will make profit depending on the levels of fluctuation. So the system determines not only how much to bet, but how many numbers with potential profit you'll be playing - all determined by the individual fluctuation.
In the above example, 16 numbers make profit in the best/worst case scenario (rarest event). But after only a couple of spins at the beginning of a session, there would be about 35 numbers making a profit and only 2 losing.
So the system/spreadsheet takes cares of everything except when to miss out a spin entirely, i.e. wait for more virtual losses than wins before continuing to bet (or waiting for a new higher ratio following a new high).
(link:://upload.wikimedia.org/wikipedia/commons/thumb/f/f9/Largenumbers.svg/960px-Largenumbers.svg.png)
Where that downward spike is at the beginning is our maximum gap or fluctuation = 778! For Numbers this occurred in the 600,000 range. For reds/black it happens earlier and isn't as far out (probably less than 200).
It seems reds and blacks (ECs) are fluctuating to a similar gap as the Numbers (in the 700s gap range), but this occurs earlier for ECs in the 250,000 spin range instead of 660,000 range. Therefore, with the numbers we should have to wait for less virtual losses before betting. Red-Black/ECs are more of a steep climb up to the highest ratios, so they require waiting for larger gaps before betting.
Spin Min Max Gap
256775 124591 125337 746
Nope, that is wrong. The Reds-Blacks had not reached the maximum at 250,000 range:
750503 364617 365855 1238
So they fluctuate way more than the Numbers as I originally predicted and the steepness of numbers may therefore be comparable.
EDIT: 995544 483681 485210 1529
1 million spins are not enough to determine the wild behaviour of red-black!
I'm not sure what you doing, but I'm curious... Need to look at the spreadsheet when I'm at home, can't do it here...
We can turn 36 numbers into 144 numbers by monitoring 36 numbers over a 4 spin cycle. In other words we could monitor 144+ outcomes over a multiple spin cycle in order to try to bring the max gap/fluctuation (778 for 36 numbers) closer to the starting line. However, because we are waiting 4+ spins before recording a result that event may not happen any sooner than what it does for 36 numbers (660,000 spins). The upward climb is harder than the downward climb, so maybe we would need a bot to beat roulette.
EDIT: 36=37! :lol:
Nah, you still get the same number of outcomes per spins regardless, but it has got me thinking about another system like Pattern Breaker.
You monitor 37 numbers over a 2 spin cycle = 74 Numbers/outcomes. You then somehow bet on the other 72 outcomes happening next, but not the pattern/outcome that just happened.
If you get 36,1
Bet 1: 1-35
Bet 2: 2-36
If you lose the first bet then you have to multiple each bet by about 35 or something. Anyway, I digress.