18 {{ upvoteCount | shortNum }}
18 {{ upvoteCount | shortNum }}

My Experiment...

by max5ty
17 replies
In another section, Gordon was talking about using a crystal ball to help predict the next lottery numbers. I'll be interested to hear how his experiment works out.

I thought maybe, just maybe, I could use AI to help me with a more statistical approach...

as you can see, I live an exciting life at times.

I fed Gemini all the Ohio pick 3 draws for the last 180 days.

Of course, it started out by telling me it couldn't predict winners...

which I agreed, but then explained I wasn't necessarily looking for winners, but a statistical analysis based on the past drawings for the future drawings.

After a couple of back-and-forths, it decided to comply with my request.

I then asked it about 12 different questions that related to statistics for the last 180 days to narrow down and refine what I was after. I had sat down and devised the questions beforehand over about a weeks time to not only narrow the results down but also give me the best outcome.

Eventually, it gave me these numbers as potential draws in the near future:

140, 041, 410, and 104.

Now, will it be right?

I don't know. I'll be watching though.

I'll see if AI can be used to pick lottery numbers using a statistical approach.

Who knows? It may be totally off, or, it could have some potential with the lottery since it seems to solve a ton of other issues from medical to writing.

I may need to refine my list of 12 questions. I don't know.

It will be interesting to see if the experiment has any value using AI.
#experiment
Avatar of Unregistered
  • Profile picture of the author Kay King
    After a couple of back-and-forths, it decided to comply with my request.

    In my mind I see an A.I. message saying "Welllll, OK, I'll try"....




    140, 041, 410, and 104.

    Obviously, A.I. really likes 4's, 0's, and 1's.... If this works you have a real money maker of a method on your hands.
    Signature
    Saving one dog will not change the world - but the world changes forever for that one dog
    ***
    Be the reason someone believes there are good people.
    {{ DiscussionBoard.errors[11797240].message }}
    • Profile picture of the author max5ty
      Originally Posted by Kay King View Post


      Obviously, A.I. really likes 4's, 0's, and 1's.... If this works you have a real money maker of a method on your hands.
      It picked those numbers based on the statistics of the last 180 days. I guess they are 'hot' numbers.

      Yeah, I don't really have high hopes for this being a money maker haha.

      Just thought it would be a fun experiment.

      I figured AI is able to diagnose medical conditions and plan for rockets to go to space...so hey, let's see how it works for the lottery.

      Maybe, just maybe, if I can refine my prompts enough it will work? But again, I don't have very high hopes.
      {{ DiscussionBoard.errors[11797242].message }}
      • Profile picture of the author Frank Donovan
        Originally Posted by max5ty View Post

        It picked those numbers based on the statistics of the last 180 days. I guess they are 'hot' numbers.
        Yeah, I don't really have high hopes for this being a money maker haha.
        The money would probably be in selling the system to those unaware of how chance works.

        At least it seems AI is developing a sense of humor.

        Signature


        {{ DiscussionBoard.errors[11797245].message }}
        • Profile picture of the author max5ty
          Originally Posted by Frank Donovan View Post


          The money would probably be in selling the system to those unaware of how chance works.
          Shhhh...let's not talk about details here hahaha.

          But seriously, if you look at the pick 3, the overwhelming majority of the time 2 numbers will be from the high side or low side and 1 number will be from the opposite side.

          Low side 0 - 4.

          High side 5 - 9.

          Statistically, over 80% of the time the numbers will fall into this grouping.

          There are other statistics of probability that I entered into AI.

          So, if we use statistics and base our picks off these statistics, we can maybe use AI to pick our numbers?

          This is all stuff I entered into my grand scheme to make a million dollars off the lottery
          {{ DiscussionBoard.errors[11797246].message }}
  • Profile picture of the author Kay King
    I think you are aiming too low - make it one of the biggies - the hundred million sort of win. Then send my percentage to me for this valuable advice.




    I had a co-worker who won a 9 million court settlement - and collected on it from the insurance company.....and was totally broke again 3 years later. Apparently there is no common sense attached to winning money.
    Signature
    Saving one dog will not change the world - but the world changes forever for that one dog
    ***
    Be the reason someone believes there are good people.
    {{ DiscussionBoard.errors[11797247].message }}
    • Profile picture of the author max5ty
      Originally Posted by Kay King View Post

      I think you are aiming too low - make it one of the biggies - the hundred million sort of win. Then send my percentage to me for this valuable advice.




      I had a co-worker who won a 9 million court settlement - and collected on it from the insurance company.....and was totally broke again 3 years later. Apparently there is no common sense attached to winning money.
      I actually thought someone would point out that the 4 numbers it gave me didn't fit into the high/low statistic.

      This is why I really don't have high hopes for this system.

      But we'll see.

      I'm trusting AI to make me a millionaire eventually.

      The biggie is good...but, statistically, you're more likely to win on games that have higher winning percentages.

      Truthfully, I seldom play the lottery. But, I'm going to the store and playing the 4 numbers it gave me
      {{ DiscussionBoard.errors[11797248].message }}
  • Profile picture of the author GordonJ
    Maybe we should have a contest, my balls against your AI?

    Actually, here is what I would want the stats to tell me, so, along with the magic of the scrying, I have some data to back it up.

    How long between the same number being drawn in the same position? So, with 140, how many times a month would 1xx follow 140? Or if 401, 4xx?

    What is the longest time (draws) between the numbe1 in the first digit until it comes up again (63 has been the longest in Ohio), it gets tripped at 60 (one month) and becomes a must play.

    Double digits are 27% of all possible draws, yet they avg. closer to 24%, and over a decade are mostly even, but in a year, and then in a given month, there is a definite bell curve, and the mistake most make is to play what is due. I would want to know if AI wants to play what is DUE, vs, what is HOT.

    Can you ask it that? Hot v due. My whole theory could be wrong. (I play hot only).

    Exciting times for all of us, using AI in simplistic ways, until the robots kill us all ... a good new show on Apple is the Rashida Jones show, SUNNY. Tune in if you can.

    The problem with statistical analysis and lottery is, the meet place is just about break even...in other words, to win 10k in big draws, one must spend 10 k to get there, and that just isn't very attractive to most investors, is it?

    But on the little, "money ball" side of things, a ROI of 22% is doable, just when greed wants one to SCALE IT UP, maths knock it on its arse.

    Keep us posted, I'm ready to buy your system...but I will hold on to my balls while doing so, eh? he he.
    GordonJ


    Originally Posted by max5ty View Post

    In another section, Gordon was talking about using a crystal ball to help predict the next lottery numbers. I'll be interested to hear how his experiment works out.

    I thought maybe, just maybe, I could use AI to help me with a more statistical approach...

    as you can see, I live an exciting life at times.

    I fed Gemini all the Ohio pick 3 draws for the last 180 days.

    Of course, it started out by telling me it couldn't predict winners...

    which I agreed, but then explained I wasn't necessarily looking for winners, but a statistical analysis based on the past drawings for the future drawings.

    After a couple of back-and-forths, it decided to comply with my request.

    I then asked it about 12 different questions that related to statistics for the last 180 days to narrow down and refine what I was after. I had sat down and devised the questions beforehand over about a weeks time to not only narrow the results down but also give me the best outcome.

    Eventually, it gave me these numbers as potential draws in the near future:

    140, 041, 410, and 104.

    Now, will it be right?

    I don't know. I'll be watching though.

    I'll see if AI can be used to pick lottery numbers using a statistical approach.

    Who knows? It may be totally off, or, it could have some potential with the lottery since it seems to solve a ton of other issues from medical to writing.

    I may need to refine my list of 12 questions. I don't know.

    It will be interesting to see if the experiment has any value using AI.
    Signature
    Remote Influence, Your Key to IM Success
    {{ DiscussionBoard.errors[11797262].message }}
  • Profile picture of the author DWolfe
    I wonder how that would work with horse racing. I used to see certain Jockeys win at the Meadowlands race track every night. The programs would list the names and how many wins each jockey has during the season. The top 2 jockeys won a couple of races every night normally. Could Ai take the average and predict how many wins they would have on an average night? On top of predicting the percentage chance, they finish 1-2 in the daily and late double?
    Signature

    Learn to increase revenue & marketing skills with ai - https://www.coursera.org/explore/genai-for-marketing

    {{ DiscussionBoard.errors[11797270].message }}
    • Profile picture of the author max5ty
      Originally Posted by DWolfe View Post

      I wonder how that would work with horse racing. I used to see certain Jockeys win at the Meadowlands race track every night. The programs would list the names and how many wins each jockey has during the season. The top 2 jockeys won a couple of races every night normally. Could Ai take the average and predict how many wins they would have on an average night? On top of predicting the percentage chance, they finish 1-2 in the daily and late double?
      During the early 90s, I was a member of the handicappers club at Scioto Downs.

      I didn't get wealthy. I did ok though. Probably about broke even overall.

      The whole idea of handicapping a race was to not look at the fastest horse, but at the horse that was the fastest overcoming the halfway mark.

      There was a formula to determine that. If a horse was the fastest overcoming the halfway mark they had the best chance of winning the race.

      Anyway, the pick 3 numbers for the mid-day drawing in Ohio were 025. Once again 2 were from the low side and 1 from the high side.

      I'll see what the evening numbers are. I may need to modify my prompts a little better.
      {{ DiscussionBoard.errors[11797281].message }}
      • Profile picture of the author DWolfe
        Originally Posted by max5ty View Post

        During the early 90s, I was a member of the handicappers club at Scioto Downs.
        I didn't get wealthy. I did ok though. Probably about broke even overall.
        I honestly do not know that much about handicapping. If I go to a horse track and break even that's a good night. I'm not much of a gambler and would rather make place bets or exacta boxes. Instead of looking for a big jacket pot on a horse that is a long shot.

        Speaking of sports mentioned in one of the posts. The old Newark News (newspaper) at one time had pages of each NFL game breakdowns, with all the statistics. If you took the scores, rushing yards, receiving yards, and one or two other statistics. You could figure out the winners in each game. Unfortunately, that was when the NFL only played 14 games, and the Newark News was brought out by a larger paper company.

        Let's see how you make out with your lottery picks.
        Signature

        Learn to increase revenue & marketing skills with ai - https://www.coursera.org/explore/genai-for-marketing

        {{ DiscussionBoard.errors[11797452].message }}
        • Profile picture of the author max5ty
          Originally Posted by DWolfe View Post

          I honestly do not know that much about handicapping. If I go to a horse track and break even that's a good night. I'm not much of a gambler and would rather make place bets or exacta boxes. Instead of looking for a big jacket pot on a horse that is a long shot.

          Speaking of sports mentioned in one of the posts. The old Newark News (newspaper) at one time had pages of each NFL game breakdowns, with all the statistics. If you took the scores, rushing yards, receiving yards, and one or two other statistics. You could figure out the winners in each game. Unfortunately, that was when the NFL only played 14 games, and the Newark News was brought out by a larger paper company.

          Let's see how you make out with your lottery picks.
          I kind of think like most betting, it's all a matter of money out and money in.

          There are those that spend a lot of money gambling and when they win big others are amazed...but, in reality, most of the time they've spent a lot of money to win big.

          There's a betting thing at a local carryout where you can buy some kind of tickets out of a machine and win up to like a 1000 dollars. I was in there one day when a guy won the big jackpot. After the guy left the owner told me the guy probably broke even because he'd go in there and spend like $500 at a time some days trying to win.

          It's like the lottery...you can win big if you're lucky, but most of the time some of the winners will spend a lot of money every day just to win...

          once they do win, do they really win? Or, are they really just breaking close to even from all they've spent?

          Of course, if someone wins the jackpot that's another story. But the odds are against winning the big jackpot and the jackpot is based on all the losers that didn't win.

          Millions are spent on lottery systems to increase their chances of winning...

          but, in reality, can anyone really predict what balls will pop up?
          {{ DiscussionBoard.errors[11797457].message }}
  • Profile picture of the author Kay King
    The top jockeys won more because they were given the best horses to ride...because they were the top jockeys.

    I wonder how A.I. would match up against professional handicappers.
    Signature
    Saving one dog will not change the world - but the world changes forever for that one dog
    ***
    Be the reason someone believes there are good people.
    {{ DiscussionBoard.errors[11797271].message }}
    • Profile picture of the author GordonJ
      Originally Posted by Kay King View Post

      The top jockeys won more because they were given the best horses to ride...because they were the top jockeys.
      I wonder how A.I. would match up against professional handicappers.
      Horse racing, dog racing, has too much potential for human interference; drugs, jockeys, etc. In other sports, like basketball, a recent life ban was put in place on a player who bet on his own team...and although Charlie Hustle was certainly a hall of fame player, (Pete Rose), he too got the lifetime ban.

      Now the biggest sport bet on in USA is NFL football and college football. And there is a lot of handicapping taking place, and injuries play a huge roll.

      Could one feed the data of a player coming "home", to his HS/College town, and being a key player, say wide receiver, would AI predict he would have a great game, or not? How would AI handle the adrenaline of player getting his chance to shine? Would it be a factor?

      I think that is going to be the AI stumbling block on handicapping, because past performance may predict accurate behavior, on any given Sunday, a Brock Purdy can emerge. It is interesting to see how AI factors in.

      Lotteries actually shut down numbers if too many are played, no more bets can be made when it reaches a point...would sports books shutter certain wagers if AI started winning?

      GordonJ
      Signature
      Remote Influence, Your Key to IM Success
      {{ DiscussionBoard.errors[11797274].message }}
  • Profile picture of the author max5ty
    So, I'm working on refining my prompts.

    I'm working on my 12 prompts to something I think will lead to better results.

    Yeah, I know this is a little crazy...

    but then again, a lot of lottery sites always say past statistics are useful. So I'm using past statistics more in my prompts...and asking for future picks based on the past statistics.

    I still don't have high hopes for this since it is a game of chance.

    But, it's like I'm now obsessed with seeing how close I can come.

    I'll post the results of the AI picks later when I get the prompts refined.
    {{ DiscussionBoard.errors[11797446].message }}
  • Profile picture of the author Cards
    text me when you win big, I'll congratulate you in person, even though I don't really believe in it
    {{ DiscussionBoard.errors[11797541].message }}
    • Profile picture of the author max5ty
      Originally Posted by Cards View Post

      text me when you win big, I'll congratulate you in person, even though I don't really believe in it
      I'm not holding my breath that I'll hit it big.

      I'm thinking maybe Gordon's idea of using a crystal ball might be just as effective

      But, if I do win big...
      {{ DiscussionBoard.errors[11797558].message }}
  • Profile picture of the author up2lead
    [DELETED]
    {{ DiscussionBoard.errors[11797663].message }}
    • Profile picture of the author max5ty
      Originally Posted by up2lead View Post

      have you tried this with just the Gemini?
      Yes.

      I downloaded the past 180 days from the ohio lottery site and uploaded the file into gemini so it could read the file and find patterns.
      {{ DiscussionBoard.errors[11797668].message }}
Avatar of Unregistered

Trending Topics