AI - Machine Learning - NeuralNetworks - SERIOUSPrev Topic Next Topic

New Topic New Poll

luckyguyxxx

Italy
Member #184,204
August 30, 2017
11 Posts
Offline

Dec 21, 2018, 11:56 am

Hey guys, i'm have a good knowledge in the AI/NN (machine learning) field, and have access to huge computational powers servers and sophisticated machine-learning framework.

I'm also a passionate lottery player, always looking to find ways to find the winning numbers.

So let's talk about AI, as i said above, what i have is the opportunity to build and train huge neural networks with high computational power (using GPU chips that are much much faster than normal CPUs), i've tried to train the networks in many ways, but so far nothing seems to prove that the lottery numbers are not totally random (and in fact, technically, they are totally random).

However, if somebody have an idea about what to train the NN with, just write it down here in this thread, i'm open to test any idea on my resources, for free!

Thanks :)
oate

United States
Member #193,161
October 22, 2018
491 Posts
Offline

Dec 21, 2018, 2:33 pm

It sounds like an interesting project. What about training it with the precise time each drawing starts and ends, outside temperature and pressure around the facility where drawings are held, data from radiation and seismic activity in the area. Are your lotteries with balls or RNGs, or both?

Best of luck to you!

Only my hairdresser knows for sure.
GuesssWork

I-95
United States
Member #178,799
December 17, 2016
2,232 Posts
Offline

Dec 21, 2018, 2:41 pm

Quote: Originally posted by oate on Dec 21, 2018
It sounds like an interesting project. What about training it with the precise time each drawing starts and ends, outside temperature and pressure around the facility where drawings are held, data from radiation and seismic activity in the area. Are your lotteries with balls or RNGs, or both?

Best of luck to you!

Hey, do you think he should throw a horseshoe at the mainframe hit it in the main cpu while it is doing it's calculations.. Just for luck you know, cause even AI/NN needs luck when it comes to random

Strategies Timing Luck

It's all Guess-Work
luckyguyxxx

Italy
Member #184,204
August 30, 2017
11 Posts
Offline

Dec 21, 2018, 2:56 pm

since now i tried training it with past drawings (both in form of numbers of the balls, and in permutations numbers), as well as with retards (from how long each number come out), and binary data (permutation>binary sequence), i guess i tried anything possible but never thought about outside temperature and pressure around the facility where drawings are held, i doubt it would even work though (yes it use balls).

i will keep my mind open for new ideas
oate

United States
Member #193,161
October 22, 2018
491 Posts
Offline

Dec 21, 2018, 3:12 pm

Quote: Originally posted by luckyguyxxx on Dec 21, 2018

since now i tried training it with past drawings (both in form of numbers of the balls, and in permutations numbers), as well as with retards (from how long each number come out), and binary data (permutation>binary sequence), i guess i tried anything possible but never thought about outside temperature and pressure around the facility where drawings are held, i doubt it would even work though (yes it use balls).

i will keep my mind open for new ideas

i tried training it with past drawings (both in form of numbers of the balls, and in permutations numbers), as well as with retards

You're in the right place.

Only my hairdresser knows for sure.
ALX

Maine
United States
Member #99
January 27, 2002
6,190 Posts
Online

Dec 21, 2018, 4:14 pm

Are you interested in EuroMillions, EuroJackpot, Pick4 or Pick3?

If it wasn't impossible I wouldn't even consider it! As the Lottery Guru Gail Howard used to say, "What happens most is what is most likely to happen".
Todd

50

Chief Bottle Washer

New Jersey
United States
Member #1
May 31, 2000
28,846 Posts
Online

Dec 21, 2018, 5:19 pm

<Moved to Lottery Discussion forum>
Please post in the appropriate forum ... thank you.
KY Floyd

NY
United States
Member #23,834
October 16, 2005
5,128 Posts
Offline

Dec 21, 2018, 6:14 pm

"so far nothing seems to prove that the lottery numbers are not totally random"

Imagine that.

"outside temperature and pressure around the facility where drawings are held, data from radiation and seismic activity in the area"

Seismic activity would be largely random, and I'm skeptical that radiation (other than something like the sun's radiation affecting temperature) would have any effect, so I'm curious how you'd account for them. Pressure and temperature would affect air density, and mechanical properties of the balls and mixing machine, so are certainly part of the variables that affect the chaos that results in randomness. I'd also add in exact time of the drawing along with gravitational factors. Relative position of the sun changes with time of day and from day to day, and the same is true of the moon. Then there's humidity and static electricity. Those are just some of the factors that will have some effect.

In theory accounting for every variable should result in accurate predictions, but even if you could be positive which variables have an effect that needs to be considered you'll never be able to accurately account for all of them ahead of time. One of the variables you won't be able to account for is the most important, and that's the exact timing of the drawing process. A tiny fraction of a second will result in the mixing apparatus being in a different position when the balls start to be impacted, and tiny differences in the timing of ball selection will result in the balls and machinery being in different positions. Even minor differences in the position of the machine could affect how level it is, and that would affect the speed at which the balls enter the chamber and how soon they fall to the spot where they contact the mixing apparatus.
Tucker Black

Colorado
United States
Member #173,294
February 25, 2016
1,135 Posts
Offline

Dec 21, 2018, 7:28 pm

If you could find a bias in the numbers from humidity or the tides gravitional pull, it would be a tiny fraction of one percent. This is not going to help you overcome the gigantic house edge.

It's like counting cards in a 100 deck blackjack game where you have to pay a 50% commission on all wins (a game that doesn't exist because no one would play). No matter how good you are at counting, you will never be able to overcome a 50% house edge. You'll just lose slightly less.
oate

United States
Member #193,161
October 22, 2018
491 Posts
Offline

Dec 21, 2018, 7:48 pm

Quote: Originally posted by GuesssWork on Dec 21, 2018
Hey, do you think he should throw a horseshoe at the mainframe hit it in the main cpu while it is doing it's calculations.. Just for luck you know, cause even AI/NN needs luck when it comes to random

What about having an herbal medicine man bless some neodymium magnets to put on the CPU? In California you can't throw a deodorant rock without hitting a few hippies up for the job. Best of luck and Merry Christmass to you and your octagons, Guessswork.

Only my hairdresser knows for sure.
hearsetrax

United States
Member #52,343
May 21, 2007
4,034 Posts
Offline

Dec 22, 2018, 5:39 am

Greetings and welcome Luckyguyxxx

sounds like extremely fascinating work indeed

good luck and I hope you win more then you lose
BobP

Dump Water Florida
United States
Member #380
June 5, 2002
3,678 Posts
Offline

Dec 22, 2018, 6:29 am

Quote: Originally posted by oate on Dec 21, 2018
i tried training it with past drawings (both in form of numbers of the balls, and in permutations numbers), as well as with retards

You're in the right place.

What you're looking for is Chaos Theory, the way a stream may be totally random, yet perfect whirlpools form break down and form again, i.e.: order within chaos. The pattern we're always looking for is one like some previous ones and what the next draw was like back then.

Of course the actual draws don't repeat, so you need to enlarge the pattern say to a 12 to 18 number or more cluster from multiple draws to produce a matching projection that can be wheeled.

BobP
AllenB

California
United States
Member #167,120
June 27, 2015
1,777 Posts
Offline

Dec 22, 2018, 1:25 pm

It seams to me that you should attack the RNG Games. Ignore the Physical Stuff and Just Go Computer vs Computer. Ask you AI Program to come up with it's own Random Generator that can somehow Hit all of the Past Games and When it does, Let them go at it.
KY Floyd

NY
United States
Member #23,834
October 16, 2005
5,128 Posts
Offline

Dec 23, 2018, 3:09 am

"If you could find a bias in the numbers from humidity or the tides gravitional pull, it would be a tiny fraction of one percent."

I doubt that any of those would cause a bias. They'd simply be factors that introduce chaos, and therefore randomness, in a mixing process that uses the consistent and regular motion of the mixing apparatus, but accounting for them introduces a degree of predictability. Does anyone think that it's not theoretically possible to make accurate predictions if you could run simulations that properly account for all of the variables that introduce chaos?

I think it's all somewhat interesting in terms of the theoretical calculations, but the magnitude of effects caused by things that can be accounted for doesn't matter when there are clearly factors that can't be accurately accounted for. A one minute variation in the timing of the drawing would have almost zero effect in terms of otherwise significant gravitational and meteorological factors, but a tiny fraction of a second can change which part of which ball first contacts moving parts and where that contact happens. Even with just a few balls there's a nearly infinite number of possibilities for how that first contact starts the randomization.

Perhaps we should think of lottery results (for ball drawings, anyway) as being chaotic rather than random. The results are certainly random in terms of practical abilities to predict them, but in some sense the mixing process isn't really random. The balls fall based on gravity, are deflected by contact, and then selected by timing. That's a simple function of known physics. It's how all the various factors will effect all of the collisions that keeps it from being a straightforward, if complex, calculation.
GuesssWork

I-95
United States
Member #178,799
December 17, 2016
2,232 Posts
Offline

Dec 24, 2018, 7:53 am

Quote: Originally posted by oate on Dec 21, 2018
What about having an herbal medicine man bless some neodymium magnets to put on the CPU? In California you can't throw a deodorant rock without hitting a few hippies up for the job. Best of luck and Merry Christmass to you and your octagons, Guessswork.

I tried that with a hippie/Herbal Pc Programmer guy. Hey told me to forget about that we come equipped with mega computers built in with no limits. He told me to take 3 N52 flat magnets shave the top part of my head to the scalp.

Tape magnet 1 negative side down positive side facing the cosmos. Magnets 2 and 3 goes positive sides in below each side of the head below each cheek bone on the carotid arteries. I ask him really? He says sure youngster i'm of the old school your from the new era you don't know jack snip of what i know. If you want to see what no others see do as i say.

I told him sure thing i will try it out as soon as i get the magnets. So i say thanks for the info and proceed to leave. He says youngster there's one thing i left out. I ask what could that be, he said listen to "Simple Man" while your magnets are in place in a quiet area and you will open the doorway to oneness being in tune with the past and present. When i finally got far enough away from the crazy i went to youtube to search the title up he told me to listen to and i was like, that old man is off his rocker.

So for all of you 35 and older old timers you should try this out its not of my era not my cup of tea it may work for you but i will never know. Merry xmas to the old Hippies out there

https://www.youtube.com/watch?v=sMmTkKz60W8&list=LLAIMpu2wcO9tu5enOiaA9ug

I will stick with the octagons.. lol

Strategies Timing Luck

It's all Guess-Work

New Topic New Poll

Subscribe to this topic

AI - Machine Learning - NeuralNetworks - SERIOUSPrev TopicNext Topic

AI - Machine Learning - NeuralNetworks - SERIOUSPrev Topic Next Topic