Welcome Guest
You last visited December 10, 2016, 11:02 pm
All times shown are
Eastern Time (GMT-5:00)

# Best way to predict numbers

Topic closed. 16 replies. Last post 13 years ago by plnwebguy.

 Page 1 of 2

United States
Member #3757
February 18, 2004
15 Posts
Offline
 Posted: April 8, 2004, 8:23 am - IP Logged

I'm trying to gauge how people play jackpot lottery games.  Let me explain a little.

I'm currently researching what methods people who play jackpot games use to pick their numbers for jackpot games like a state jackpot lotto, powerball, mega millions, etc.

For example, do people use patterns, like 45 degree patterns, based on previous results?  Do they use sums, or a variant on sums?  Do they use birthdays or lucky numbers consistently?  Hot versus cold numbers, frequencies, astrology, etc?

Any thoughts?  I would like to get a better idea of what people use.

Thanks everyone!

New Jersey
United States
Member #3457
January 22, 2004
248 Posts
Offline
 Posted: April 8, 2004, 8:53 am - IP Logged

I wrote a program many years ago that maximizes, for the number of tickets purchased, the percentage of (x,y) pairs represented in the universe of such possible pairs.

I go forward in the realization that  people who claim to "predict" numbers succeed only through the same dumb luck as do pick quick buyers.

United States
Member #3757
February 18, 2004
15 Posts
Offline
 Posted: April 8, 2004, 10:01 am - IP Logged

Quote: Originally posted by prob987 on April 08, 2004

I wrote a program many years ago that maximizes, for the number of tickets purchased, the percentage of (x,y) pairs represented in the universe of such possible pairs.

I go forward in the realization that  people who claim to "predict" numbers succeed only through the same dumb luck as do pick quick buyers.

Hmm, good point.  Since any lotto draw is based on pure random draws, it would seem that a quick pick would be just as accurate.  Has this been your experience?  I'm doing that right now actually.  I'm using Lotto Pro on one ticket and a quick pick on another.  So far, they are about equal as far as the numbers go.

New Jersey
United States
Member #3457
January 22, 2004
248 Posts
Offline
 Posted: April 8, 2004, 11:59 am - IP Logged

My "system" works reasonably well.  I win at about the expected rate for the number of tickets I buy.

Quick picks can actually fall somewhat below the expected rate under certain circumstances.  There is, however small, some probability, equal to the odds of winning, that if you buy two quickpicks, you will actually get exactly the same numbers, for instance, the computer might give you 7-10-13-16-19 with a megaball of 22 twice.  In this case you would have beat extraordinary odds, but your chance of actually winning the lottery with these two quick picks would be exactly the same as if you bought one ticket and not two.  There are intermediate gradations of this.  Actually people who "wheel" are (under most circumstances) doing this sort of thing to themselves.

Sometimes I buy two quick picks just for the thrill of beating extraordinary odds.  If you buy two megamillions quickpicks, the odds of getting the ticket with exactly those two numbers printed on them is roughly 1.8 X 10^16, or about 3% of the age of the universe measured in seconds.  Beating such odds, as you do everytime you buy such a ticket, doesn't get you any money, but it's thrilling nonetheless.

Member #2673
November 2, 2003
497 Posts
Offline
 Posted: April 8, 2004, 12:43 pm - IP Logged
Quote: Originally posted by prob987 on April 08, 2004

Sometimes I buy two quick picks just for the thrill of beating extraordinary odds.  If you buy two megamillions quickpicks, the odds of getting the ticket with exactly those two numbers printed on them is roughly 1.8 X 10^16, or about 3% of the age of the universe measured in seconds.  Beating such odds, as you do everytime you buy such a ticket, doesn't get you any money, but it's thrilling nonetheless.

prob987,

I do not agree with your calculation. The odds of getting two quick picks to match exactly is roughly 1 in 135,000,000. The odds of the first ticket is 1 in 1 (that is, any ticket will do). The second ticket now needs to match the first. And therefore the odds of 1 in 135,000,000 times the odds of 1 in 1 yields the combined odds of 1 in 135,000,000.

Good luck,
Jake

New Jersey
United States
Member #3457
January 22, 2004
248 Posts
Offline
 Posted: April 8, 2004, 1:04 pm - IP Logged

No, not really Jake, the odds of getting two tickets of any combination is 135,000,000 squared, no matter whether the two tickets are the same or different.  The first ticket has odds of 1 in 135,000,000 odds of being generated.  At that point, when the first ticket is generated (against extraordinary odds) the second ticket will then have odds of 135,000,000 to one against it.  The second is independent of the first ticket though.  The nature of the first ticket has no bearing on the nature of the second.  Overall, before the generation of either number, the odds are 1 to 1.8 X 10^16 against the two numbers combined being on the printed ticket.

Member #2673
November 2, 2003
497 Posts
Offline
 Posted: April 8, 2004, 2:21 pm - IP Logged

prob987,

There are two situations.

1. What are the odds that the following ticket (1, 12, 23, 32, 41, MB13) is selected twice in a row using quick picks? Answer: (1 in 135,000,000) squared.

2. What are the odds that any two quick picks match exactly. Answer: First you buy one quick pick. The odds of getting any combination is 1 in 1 because any combination will do. Then you buy a second ticket where the numbers must match the first ticket, odds of 1 in 135,000,000. The factor of the two odds is 1 in 135,000,000.

See the difference?

Good luck,
Jake

New Member
Battle Creek
United States
Member #4262
April 6, 2004
5 Posts
Offline
 Posted: April 8, 2004, 2:52 pm - IP Logged

Picking numbers.    It is not necessarily so, that its the luck of the draw,  some numbers are selected more than others on a regular basis,  in fact in some studies i have made,  one number came up every other day during a certain month,  the numbers 11,14,17,19,22 for example won twice, in the Michigan games in the same month,  Rolldown and Winfall,  Mar 03. Mar 25  other numbers appeared with them,  in some cases,  some of those same numbers won in Ohio or Ill.  not sure,  which,    I know its all luck but there are certain phenomenon that makes some numbers appear more frequently,  so why not bet on them,,  also,  there is the case of Cold numbers,    for example the number 1 and 4 came up 8 times each in Mi. draws,  but No. 13,23,26, came up only twice,  so how would you bet,  lol

I bet two plays one on the frequent numbers,  and one on the Cold numbers, sometimes a mix.

I know its still just luck,  for you never know when such numbers will come up, but I do think its worth a study, and bet on the frequent numbers,

When I lived in Texas I won the 3 pick boxed,  7 days in a row,  small pay out but at least I was winning,  lol,    My method,  Telephone area Codes.

Pennsylvania
United States
Member #1340
April 6, 2003
2450 Posts
Offline
 Posted: April 8, 2004, 3:30 pm - IP Logged

The thing about number frequency trends is I don't usually spot them until it's over...

In doing some research on Neural Networks and their underlying formulae, a very interesting concept was presented... (this book is based on actually modelling the process of learning, not NNs for lottery use...) Short Term Memory and Long Term Memory... I think something here can possibly be applied to the biases present in drawing histories...

People (myself included) who use entire game histories are using something analogous to long-term memory, people who try to spot trends from a few draws or weeks of draw data are more in the realm of short-term memory... perhaps we can replace the term "memory" with trends...

Long-Term Trends(LTT's)... Use of long term trends allows you to profile the average draw effectively. Frequency data based on LTT's yields most popular numbers overall, this does not directly translate into winning numbers in any single draw, however... Example... the number 27 has been absent for all 157 drawings as the powerball

Short-Term Trends(STT's)... Looking at the last few draws presents the player with wilder variances when taken out of context of the big picture, but it does put the microscope on short-lived indicators such as repeating numbers. Unfortunately, false trends abound in this narrow area of data... Example... Between 3/6/04 and 3/20/04, all 5 drawings had powerball numbers under 10

What would be great is if we could somehow take short-term trends and possibly add or subtract weight to them in order of relevance and importance based on Long Term Trends... But how?

Playing more than one ticket per game is betting against yourself.

mid-Ohio
United States
Member #9
March 24, 2001
19831 Posts
Offline
 Posted: April 8, 2004, 5:44 pm - IP Logged

When looking at the last 150 draws of MegaMillions 198 drawings, 75 of these drawings have had all 5 main numbers hit in the previous 21 drawings or less, so I've look at all the previous 21 drawings of the 150 drawings and compared them as short term trends to pick the most common trend amoug them to apply to my next picks, so I have a set of parameters to cover both trends when picking my numbers.  Both sets of parameters cover 50%+ of the history.

RJOh

* you don't need to buy more tickets, just buy a winning ticket *

New Jersey
United States
Member #3457
January 22, 2004
248 Posts
Offline
 Posted: April 8, 2004, 8:31 pm - IP Logged
Quote: Originally posted by Jake649 on April 08, 2004

prob987,

There are two situations.

1. What are the odds that the following ticket (1, 12, 23, 32, 41, MB13) is selected twice in a row using quick picks? Answer: (1 in 135,000,000) squared.

2. What are the odds that any two quick picks match exactly. Answer: First you buy one quick pick. The odds of getting any combination is 1 in 1 because any combination will do. Then you buy a second ticket where the numbers must match the first ticket, odds of 1 in 135,000,000. The factor of the two odds is 1 in 135,000,000.

See the difference?

Good luck,
Jake

Yeah, I do.  But I think our problem lies in that we are talking about two different situations.  I am buying two quickpicks, as described in my original post, to demonstrate that one can in fact routinely experience "beating the odds" of 1 in 1.8 X 10^16.  I am not doing it to see if I can produce two equivalent tickets.  The odds of producing a quick pick of arbitrary sequence is indeed exactly equal to winning the lottery.  The odds of producing two identical quickpicks of a specific sequence however is the probability squared.

As incredible as these odds seem, they are nowhere near the odds against many other ordinary situations that occur in ordinary life.  Thirty two grams of oxygen gas, the amount of gas present in about 20 liters (5 gallons) of oxygen gas at ordinary pressure, contains about 6.02 X 10^23 molecules of oxygen.  The laws of statistical mechanics that define the properties of these molecules notes that each molecule has an independent (random) speed and position.  When you consider all the possible speeds and all the possible locations in a container of such a gas and multiply by the number of molecules, the odds against the situation that actually exists are so vanishingly small as to be absurd.  Still that is the actual state of the oxygen!  Next to that probability, the winning the lottery seems like a comparitive certainty.  (I hope I've made sense to you with what I've said.)

Keep that in mind the next time someone compares your lottery odds with some other highly improbable dvent.

United States
Member #3757
February 18, 2004
15 Posts
Offline
 Posted: April 9, 2004, 6:41 am - IP Logged
Quote: O

United States
Member #3757
February 18, 2004
15 Posts
Offline
 Posted: April 9, 2004, 6:46 am - IP Logged
Quote: Originally posted by prob987 on April 08, 2004

My "system" works reasonably well.  I win at about the expected rate for the number of tickets I buy.

Quick picks can actually fall somewhat below the expected rate under certain circumstances.  There is, however small, some probability, equal to the odds of winning, that if you buy two quickpicks, you will actually get exactly the same numbers, for instance, the computer might give you 7-10-13-16-19 with a megaball of 22 twice.  In this case you would have beat extraordinary odds, but your chance of actually winning the lottery with these two quick picks would be exactly the same as if you bought one ticket and not two.  There are intermediate gradations of this.  Actually people who "wheel" are (under most circumstances) doing this sort of thing to themselves.

Sometimes I buy two quick picks just for the thrill of beating extraordinary odds.  If you buy two megamillions quickpicks, the odds of getting the ticket with exactly those two numbers printed on them is roughly 1.8 X 10^16, or about 3% of the age of the universe measured in seconds.  Beating such odds, as you do everytime you buy such a ticket, doesn't get you any money, but it's thrilling nonetheless.

Let me ask you this - when you check the results online, do you see a lot of winners on the lotto home page who won via a quick pick?  I know for the NY lottery it seems everytime I go to http://nylottery.org, and see the winner's profiles, they seem to be quickpicks.  If what you're saying is true, these people are beating extraordinary odds, even by getting 5/6 numbers, or something like that.

Member #2673
November 2, 2003
497 Posts
Offline
 Posted: April 9, 2004, 8:13 am - IP Logged
Quote: Originally posted by prob987 on April 08, 2004

Yeah, I do.  But I think our problem lies in that we are talking about two different situations.  I am buying two quickpicks, as described in my original post, to demonstrate that one can in fact routinely experience "beating the odds" of 1 in 1.8 X 10^16.  I am not doing it to see if I can produce two equivalent tickets.  The odds of producing a quick pick of arbitrary sequence is indeed exactly equal to winning the lottery.  The odds of producing two identical quickpicks of a specific sequence however is the probability squared.

prob987,

That was exactly my two example situtations. Therefore we agree.

Good luck,
Jake

New Jersey
United States
Member #3457
January 22, 2004
248 Posts
Offline
 Posted: April 9, 2004, 8:19 am - IP Logged

Anyone who wins the lottery beats extraordinary odds.

The ratio of numbers of people who win by quick picks and number of people who win by picking their own numbers is probably (I have no data) slightly higher than you would expect from the ratio of quick pick tickets bought to the number of people who pick their own numbers.  This is because many people who pick their own numbers do a poor job of expanding the covered field size.

This is because people tend to play their birthdates and they wheel subsets of the numbers in the field.  If even one number appears that is outside their wheel, they are prdvented from winning the jackpot.  (They are however, under certain circumstances, deriving a certain powerplay like advantage if they win smaller prizes:  There once was a winner in California, if memory serves me well who won the 1st, 2nd, a few 3rd prizes, a few 4th prizes and a few 5th prizes.  Lottery officials wer astounded since they never saw such a thing, and it is unlikely they will see such an dvent again.  This person was probably a wheeler of a relatively small subset of numbers.  In spite of his reduced odds, his ship came in:  His odds were still not zero.)

Since people who play quick picks have a poisson distribution in terms of the coverage of the field, as opposed to the other type of player, whose choices cover a narrower range, quick pick winners are over represented.

Buying quick picks, as described above, however is not the best way of maximizing your odds if you buy more than one ticket.

 Page 1 of 2