I have been testing lately with a concept I thought. As the title indicated, it has to do with averages. After few numbers spun in Roulette (10-20 numbers or sometimes even less) there is a moment where the average reaches 18. This is the average that we expect anyway to reach in the long run. So having this info on my mind and calculating the average after every new spin I tried based on my bets to get closer to the average. I most of the times just wait for the first 2 spins to pass since the fluctuations are still very big.
Example: I'll give a sequence of numbers I had on my testings.
16 34 24 19 35 26 3 12 16 15 7 17 23 5
Waiting the first 2 spins to pass we have average of 25 ((16+34)/2 = 25)
So now for the next spin we bet numbers BELOW 25 so as to get closer the the average we know it is going to reach sooner or later. So we bet numbers 0-24.
24 hits and we win. Now average is ((16+34+24)/3 = 24.667)
So now for the next spin we bet numbers BELOW 24.667 which means numbers 0-24 again.
19 hits and we win. Now average is ((16+34+24+19)/4 = 23.25)
Next spin bets will be from 0-23.
etc etc.
On the other side of the average if for example we get many lows we could have an average of 16. Next bet we will bet numbers 17-36 etc etc.
I am thinking and working on a more complex system based on this that will have proportional bets on the numbers depending how close or far are from the average and how much they will affect it. For example with average on 16 the next bets will be smaller bets on numbers 17 18 19 20 etc and bigger ones as we come closer to 36 since they will bring the balance of 18 faster, which means less spins will be available for us.
Not sure how exactly to proceed on this proportional idea yet so any ideas or changes/discussions are welcomed.
A much simpler system that I have tested so far is flat betting the numbers based on the current average as explained in the beginning.
I stop the session whenever my current average has reached from 17.5 to 18.5 range or if I am -100 units at any moment.
The results so far are very good. I know it is very small sample but it is interesting and fun to play. Small sessions also which is good for live play. My testings are so far online on RNG fun mode.
Let me know what you think guys! Happy holidays and Merry Christmas! :)
Sessions so far:
Balance Spins
70 7
41 19
11 7
42 12
-60 19
29 4
52 38
11 15
-107 15
80 10
49 3
51 12
Total: +269 units in 161 spins
Average: +1.67 units/spin
Average session: +22 units in average 13 spins-length sessions
Try these spins actual played rng. today is pp, paddypower and the others are last nights, the red is looking at devil street when came home last night
I'm not sure what you mean notto and how to try these spins. Tbh I cant understand what numbers you say or mean on your paper picture. So I prefer online testings which are faster and automated.
Anyway. I'll keep making some more tests tomorrow and post. I already had 4 more succssive sessions.
(link:://:.pichost.org/images/2016/12/21/temp_453130.png) (link:://:.pichost.org/image/UPJ7) |
First set
(link:://:.pichost.org/images/2016/12/21/temp_673317.png) (link:://:.pichost.org/image/Uj5Q) |
Smaller set
Heres the paddy power set, all rng
(link:://:.pichost.org/images/2016/12/21/temp_814802.png) (link:://:.pichost.org/image/Ucxy) |
Bell,
How many spun numbers do you use to arrive at the average?
Very interesting idea!
(26+1)/2 = 13.5 are we to bet #'s 13 to 36, win
Quote from: nottophammer on Dec 21, 07:20 PM 2016
Heres the paddy power set, all rng
(link:://:.pichost.org/images/2016/12/21/temp_814802.png) (link:://:.pichost.org/image/Ucxy) |
Quote from: nottophammer on Dec 22, 04:40 AM 2016
(26+1)/2 = 13.5 are we to bet #'s 13 to 36, win
That's right Notto,
Not sure if it'll work out flat, as you end up backing a lot of numbers and mainly high or low, as the average does not alter very quickly.
Quote from: nottophammer on Dec 22, 04:40 AM 2016(26+1)/2 = 13.5 are we to bet #'s 13 to 36, win
Actually to bet 14-36 since we want HIGHER than 13.5 average but yes. You got the style. I agree that flat bet may not be the best option but proportional depending how far or close to the average the numbers are. But it is much more complex so I started flat. Which btw still yields good results
Nice concept. Definately has some logic potentials in it. Always betting a slightly larger than 50% bet expecting the lower/higher number that should statistically show .. should hit more than average but with smaller wins. This should definately catch some of the annoing moments you when playing your usual high/low EC bet and it just hit 19 or 20 while you where betting on low to finally show. Perhaps a bit like betting on dozens but with the benefit of betting towards a goal that is pretty much garanteed longterm.
However i guess its both bit of a blessing & a curse as betting towards the average of 18 quickly turns into a normal high/low bet when the "swings" get smaller & smaller. Then there is the usual "death" pattern of like 13x high numbers when you are expecting a low number. Thankfully the death pattern is pretty rare on a real wheel but still happens every other handfull of hours. Still even a small streak of the wrong numbers can but a larger dent in your bankroll.
Perhaps sticking to just adding the last 6-8 numbers could continue the positive flow of a expected average? It would also hurt less when hitting a "death" pattern. Well at least in theory, im just guessing.
Isnt too hard to play this on a real wheel as well, even if its on a timer, as say you need to put a unit on 1-20, then you can simply put 18 units on "low " aka 1-18 (and save it as a favorite bet) and then put 1 unit on 19, and 20 manually. However doing the math still takes time too.
Quote from: Thunder Pants on Dec 23, 06:46 PM 2016Nice concept. Definately has some logic potentials in it.
Thank you. It is supposed to work. Of course there is always the 0 unfair payout factor. Maybe even better on NZ Roulette.
Quote from: Thunder Pants on Dec 23, 06:46 PM 2016Always betting a slightly larger than 50% bet
It's not always more than 50% Most of the times it's close there indeed but sometimes it can yield +20 units which means 16/37 numbers bet. but many times yes around 18-19 numbers bet flat.
Quote from: Thunder Pants on Dec 23, 06:46 PM 2016Thankfully the death pattern is pretty rare on a real wheel but still happens every other handfull of hours. Still even a small streak of the wrong numbers can but a larger dent in your bankroll.
Agree betting like 16-19 units per spin on almost high/low looks pretty dangerous. That's why I am playing git with 100 units stop loss. I know that this doesn't make it too much safer wince small stoploss can be translated into reaching it more frequently. So it can add up like a big stop loss or not any. But the good news is that with the 100 units stoploss the strategy is still after many more new sessions now still in positive territories :)
Quote from: Thunder Pants on Dec 23, 06:46 PM 2016However i guess its both bit of a blessing & a curse as betting towards the average of 18 quickly turns into a normal high/low bet when the "swings" get smaller & smaller.
I don't really understand what you mean here and it looks important. Feel free to explain. If you mean that after many spins the average sticks to 18 so it is mainly high/low, I must clarify that once average reaches 17.5-18.5 (experimenting with 17-19 for smaller need of stoploss) I reset and start again over until the new spins reach again the desired average.
Quote from: Thunder Pants on Dec 23, 06:46 PM 2016Perhaps sticking to just adding the last 6-8 numbers could continue the positive flow of a expected average? It would also hurt less when hitting a "death" pattern. Well at least in theory, im just guessing.
Sounds interesting. You mean ignoring the older spins though? Not sure I follow you :)
Quote from: Thunder Pants on Dec 23, 06:46 PM 2016However doing the math still takes time too.
I do the same to save some betting time. I bet 12 units on completed dozens or 6 units on completed lines. It doesn't take much time the average. If anyone is interested I have make on excel. I just type the number value spun and it shows the new average of all the spins together
Quote from: BellagioOwner on Dec 23, 07:35 PM 2016
...
I don't really understand what you mean here and it looks important. Feel free to explain. If you mean that after many spins the average sticks to 18 so it is mainly high/low, I must clarify that once average reaches 17.5-18.5 (experimenting with 17-19 for smaller need of stoploss) I reset and start again over until the new spins reach again the desired average.
Sounds interesting. You mean ignoring the older spins though? Not sure I follow you :)
I do the same to save some betting time. I bet 12 units on completed dozens or 6 units on completed lines. It doesn't take much time the average. If anyone is interested I have make on excel. I just type the number value spun and it shows the new average of all the spins together
Ah yes, what i was thinking was in the beginning of the session when only a few numbers has hit, the winrate is high as we bet a lot numbers again & again but after like 100 numbers the "advantage" is gone as its basicly a 50% straight win/loss with the average 18. Thus also being very vulnerable to streaking high numbers like 24, 19, 36, 29, 27 or low streaks like 8, 14, 13, 2, 4 as yes perhaps we would eventually get the average up/down to catch the numbers in the streak, but because of the many previous numbers it'll move slow like 18,5 to 19,5 to 21 etc and the many losses eats a large chunk of our bankroll.
So the random idea was try to keep the initial postive momentum where we only use the average of few numbers in the hope to "catch" more of those scary streaks. Like the most simplistic one is just to average the last 2 known numbers. The swings/bets would then be high aka one moment betting 1-25, the next 12-36 and so on. Best is if we get like 34, 29, 35, 19, 24 we wont win on the first 1-3 numbers but we should at least catch the last 2-3 numbers because the average gets so high that it manage to "catch up" with the momentum of the numbers. At the same time this also narrows down the "death pattern" something like continually lower/higher number streaks .. aka 16, 11, 4, 2, 1 or 20, 27, 31, 34, 35 that i consider an extremely rare event.
All in theory of cause & totally not tested. Not entirely sure where the "sweet spot" is either. Like just the average of 2 is clearly the most expensive bet as well & you could argue that the average of the last 4, 6 or 8 might be cheaper especially if the wheel you play at never seem to do high/low streaks etc.
This is an older post/approach of mine that I had time to test a variation these days.
The results are pretty good so far. I don't claim it to be HG etc etc but if anyone is interested in further testing it by coding it let me know so that I can tell him how it works. It has some changes from the original thread opening.
Results so far are 17/17 ALL won sessions and balance: +551 units @ 369 spins. In detail:
Balance Spins
90 21
19 2
21 9
36 13
25 44
45 5
2 25
12 40
41 6
31 10
6 3
24 2
52 74
39 7
18 3
43 5
47 100
Bellagio,
This is an interesting concept and I would like to test your ideas on the subject.
Can you say more about your new approach?