Welcome Guest
You last visited December 11, 2016, 6:48 am
All times shown are
Eastern Time (GMT-5:00)

# Statisticians search for how to beat the lottery

Topic closed. 79 replies. Last post 4 years ago by jamella724.

 Page 5 of 6
New Hampshire
United States
Member #136492
December 12, 2012
322 Posts
Offline
 Posted: February 23, 2013, 5:30 am - IP Logged

I'm guessing that you're not serious, but QP's are exactly like any other characteristic of lottery tickets.

The game is random, so no characteristic gives an advantage to a particular ticket, but if a characteristic is common we can expect it to appear in proportion to it's popularity. Since QP's account for 70 to 80% of tickets sold we would expect them to account for 70 to 80% of winning tickets if each one has the same chances as tickets that aren't QP. Since QP's do account for winners in direct proportion the rational conclusion is that there are no adavantages or disadvantages to playing QP's interms of chances of winning.

If the lottery tracks other characteristics they don't publicize it much, but you can be sure that if a game has drawings twice a week the drawings that produce winners will be proportional to sales for that day.  The percentage of winners named Bob will be proportional to the percentage of tickets sold to guys named bob. Tickets picked as a pattern from the play slip can be expected to be proportional to the percentage of tickets chosen from patterns. Tickets chosen based on "systems," "strategies" or the opinion of phychics can be expected to be proportional to the percentage of tickets chosen based on "systems," "strategies" or the opinion of phychics. And so on.

When you can find a characteristic that is represented significantly more in tickets that win than in tickets that are sold you'll have found a method for choosing tickets that's not just logical, but actually beneficial. Don't hold your breath.

We all know that quick picks win jackpots in a directly proportional relationship to the overall percentage of quick picks purchased, but has there ever been a study of the success of self picks vs overall odds to win ANY prize level?

For example if someone wins any level prize at a rate of 1:10.4 for a game with 1:16.1 overall odds, or 1:5 for a game with 1:6.6 odds with a sample size of several hundred tickets per game would you then say they have a viable system, or would you say the results fall within expected standard deviations or that it's just plain luck?

Oklahoma
United States
Member #82391
November 12, 2009
6290 Posts
Offline
 Posted: February 23, 2013, 9:47 am - IP Logged

We are all searching for ways to beat the lottery, I wrote the numbers down

I'm with you sista

I Love Pink & Green 1908

Oklahoma
United States
Member #82391
November 12, 2009
6290 Posts
Offline
 Posted: February 23, 2013, 9:53 am - IP Logged

I was thinking the same this xaab...........Sounds like pumpi...Just  little more medicated.

I miss pumpi......

I Love Pink & Green 1908

United States
Member #93947
July 10, 2010
2180 Posts
Offline
 Posted: February 23, 2013, 8:00 pm - IP Logged

We all know that quick picks win jackpots in a directly proportional relationship to the overall percentage of quick picks purchased, but has there ever been a study of the success of self picks vs overall odds to win ANY prize level?

For example if someone wins any level prize at a rate of 1:10.4 for a game with 1:16.1 overall odds, or 1:5 for a game with 1:6.6 odds with a sample size of several hundred tickets per game would you then say they have a viable system, or would you say the results fall within expected standard deviations or that it's just plain luck?

"...the results fall within expected standard deviations..."

...and luck didn't hurt.

NY
United States
Member #23835
October 16, 2005
3475 Posts
Offline
 Posted: February 23, 2013, 11:50 pm - IP Logged

We all know that quick picks win jackpots in a directly proportional relationship to the overall percentage of quick picks purchased, but has there ever been a study of the success of self picks vs overall odds to win ANY prize level?

For example if someone wins any level prize at a rate of 1:10.4 for a game with 1:16.1 overall odds, or 1:5 for a game with 1:6.6 odds with a sample size of several hundred tickets per game would you then say they have a viable system, or would you say the results fall within expected standard deviations or that it's just plain luck?

When the lotteries tell us that sales and wins by QP's have adirect correlation I assume they're talking about all prize levels. I expect that they pay attention to such things, but don't bother publicizing the results. Giving people enough information to convince them that they don't really have an advantage is probably  bad for business.

Anything that has random results produces what's known as a normal distribution, and a normal distribution isn't an equal distribution. If each number in a lottery game was drawn the same number of times as every other number it would be almost certain proof that the results were not natural. A normal distribution results in most results falling in a narrow range and a few results falling outside that range.  For  the numbers in alottery game that might mean that  most numbers are drawn between 12 and 15 times over a set period with a few being drawn 9 to 11 times or 16 to 18 times and a smaller number drawn less than 9 or more than 18 times.

The same thing happens to winners of a random game. Most people who play often enough will have results that are fairly close to  the probable result suggested by the design of the game. A few will win a bit more often and few will lose a bit more often. A smaller few will win or lose  a good bit more often than the norm. Those few who win much more often may think they've hit on somehting that gives them an advantage, but probability demands that they exist as a result of random results.

New Hampshire
United States
Member #136492
December 12, 2012
322 Posts
Offline
 Posted: February 24, 2013, 7:26 am - IP Logged

Thanks for the replies KY and Jimmy. The examples I used in my post are my actual results from the smaller multi-state games Hot Lotto and Lucky for Life. I started tracking my win per play ratio a year ago to measure the success of my number picking methodology. While I'm still looking for an upper tier hit I've been able to win back enough money to greatly reduce my cost of playing. I would still play regardless as enjoy it and have the disposable income to do so. Of course I'll keep picking numbers the same way I have been as it's worked well for me and I seem to be in the upper quartile for successful lottery players.

You guys are probably correct in saying my results can be expected and fall within a normal probability range or as Mr. Spock used to say, "Random chance seems to have operated in my favor."

Michigan
United States
Member #81740
October 28, 2009
40613 Posts
Offline
 Posted: February 24, 2013, 9:09 am - IP Logged

I miss pumpi......

so do I

Did you exchange a walk on part in the war ?

For a lead role in a cage?

From Pink Floyd's " Wish you were here"

New Hampshire
United States
Member #136492
December 12, 2012
322 Posts
Offline
 Posted: February 24, 2013, 9:59 am - IP Logged

When the lotteries tell us that sales and wins by QP's have adirect correlation I assume they're talking about all prize levels. I expect that they pay attention to such things, but don't bother publicizing the results. Giving people enough information to convince them that they don't really have an advantage is probably  bad for business.

Anything that has random results produces what's known as a normal distribution, and a normal distribution isn't an equal distribution. If each number in a lottery game was drawn the same number of times as every other number it would be almost certain proof that the results were not natural. A normal distribution results in most results falling in a narrow range and a few results falling outside that range.  For  the numbers in alottery game that might mean that  most numbers are drawn between 12 and 15 times over a set period with a few being drawn 9 to 11 times or 16 to 18 times and a smaller number drawn less than 9 or more than 18 times.

The same thing happens to winners of a random game. Most people who play often enough will have results that are fairly close to  the probable result suggested by the design of the game. A few will win a bit more often and few will lose a bit more often. A smaller few will win or lose  a good bit more often than the norm. Those few who win much more often may think they've hit on somehting that gives them an advantage, but probability demands that they exist as a result of random results.

So here's a follow up question for Jimmy or KY or anyone else that has a better grasp of statistics than I:

Do you think it's possible to have a methodology of selecting numbers that will consistently keep me in the top standard deviation for winning picks, or given enough time will my success rate go south (or north as rdgrnr would say) closer to the average odds for games I participate in?

United States
Member #116268
September 7, 2011
20244 Posts
Offline
 Posted: February 24, 2013, 10:25 am - IP Logged

So here's a follow up question for Jimmy or KY or anyone else that has a better grasp of statistics than I:

Do you think it's possible to have a methodology of selecting numbers that will consistently keep me in the top standard deviation for winning picks, or given enough time will my success rate go south (or north as rdgrnr would say) closer to the average odds for games I participate in?

New Hampshire
United States
Member #136492
December 12, 2012
322 Posts
Offline
 Posted: February 24, 2013, 10:42 am - IP Logged

You're right Ronnie316. Perhaps it's just wishful thinking on my part but I do believe it's possible to improve your hit ratio, even if only slightly, by paying attention to patterns and statistics.

By the way I followed the tip you gave me a few weeks back about staying with the same bonus ball from one draw to another on the bigger games and I actually got the bonus ball +1 last night in PB. Coincidence? Maybe, but it worked last night so thank you.

United States
Member #116268
September 7, 2011
20244 Posts
Offline
 Posted: February 24, 2013, 10:53 am - IP Logged

You're right Ronnie316. Perhaps it's just wishful thinking on my part but I do believe it's possible to improve your hit ratio, even if only slightly, by paying attention to patterns and statistics.

By the way I followed the tip you gave me a few weeks back about staying with the same bonus ball from one draw to another on the bigger games and I actually got the bonus ball +1 last night in PB. Coincidence? Maybe, but it worked last night so thank you.

I'm in your corner msharkey2001, even if our improved odds are for a single draw its ok with me because it only takes one draw to win it all and be set for life. Glad to hear you nailed a B ball, its a good feeling to pick a winner.

United States
Member #121745
January 16, 2012
4808 Posts
Offline
 Posted: February 24, 2013, 11:36 am - IP Logged

I'm in your corner msharkey2001, even if our improved odds are for a single draw its ok with me because it only takes one draw to win it all and be set for life. Glad to hear you nailed a B ball, its a good feeling to pick a winner.

Speaking of patterns, do you all consider numbers skipped, etc. to be a pattern? I was using that new tool (drawings skipped)  this morning on US lotteries to pick some new numbers to consider for Decades of Dollars.  (I had checked my ticket earlier and lost).  I took a very loose average of the number of skips in the last 5 drawings and input the 4 skip #'s (like 1, 5, etc) and then went to check the numbers to make certain the set had not won before, and the set was the DOD number that just won.   But I doubt I would have got this number had I done it yeterday since I did not have the last set of drawing numbers to use for the loose average.

J'aime La France.

New Hampshire
United States
Member #136492
December 12, 2012
322 Posts
Offline
 Posted: February 24, 2013, 2:18 pm - IP Logged

Speaking of patterns, do you all consider numbers skipped, etc. to be a pattern? I was using that new tool (drawings skipped)  this morning on US lotteries to pick some new numbers to consider for Decades of Dollars.  (I had checked my ticket earlier and lost).  I took a very loose average of the number of skips in the last 5 drawings and input the 4 skip #'s (like 1, 5, etc) and then went to check the numbers to make certain the set had not won before, and the set was the DOD number that just won.   But I doubt I would have got this number had I done it yeterday since I did not have the last set of drawing numbers to use for the loose average.

Absolutely I use drawings skipped when determining which numbers to pick along with at least one number delta of 4 or less in every line. I play Hot Lotto and Lucky For Life (analogous to your Decades of Dollars). In both of these games the winning numbers drawn have on average 3 numbers which have been drawn within the last five drawings, so I make sure to include 3 numbers per line which have come up within the last 2 1/2 weeks.

Also in both of these games a number delta of 4 or less comes up 94% of the time. I have gone back to the inception date for both of these games as my data set. Every line I play now will include at least one delta of 4 or less.

This strategy has worked pretty well for me. As Jimmy and KY would say my results have been at the very favorable end of standard deviations and random chance.

Good luck Artist77. Hopefully I'll be tied with you after Tuesdays MM drawing in the one a day MM challenge!

CHERRY HILL, NJ
United States
Member #130213
July 9, 2012
387 Posts
Offline
 Posted: February 24, 2013, 3:52 pm - IP Logged

1000 LP-members playing get each a combination sent to.

Each player is bound to play his ticket by honour of participating.

The player winning, cashes HIS ticket.

The clue is that with a big system, someone will win bigger.
To me that sounds much better than sick picks or quick picks.

After the drawing, the winners can be published, if they agree.
The price distribution should be published.

I think SergeM has a good point. To bring the odds in your favor, you will need to increase coverage (more bets). Having the \$\$\$ does not guarantee you can can actually place your wagers on time. The syndicate pools in Mass. and elsewhere had to secure their own lottery POS to do it which made it very easy to track them.

I think like-minded LP members can organize themselves, set rules and try this out. I will be among the ones who will participate.

NY
United States
Member #23835
October 16, 2005
3475 Posts
Offline
 Posted: February 24, 2013, 6:18 pm - IP Logged

Any results that depart from average should be expected to be temporary.

I don't pay too much attention to what people have to say about their results using their chosen "system", but I have  noticed a few things. I'm pretty sure that nobody has claimed to have a positive cash flow as a result of playing the lottery. People  may have a period in which they do very well, but it's preceded and followed by periods where they don't do so well. As I've already said, we should expect some people to do better than average purely as a result of random chance.

Plenty of people talk about patterns, but there appears to be near-universal acceptance that these patterns come and go. Hindsight is always 20/20, and finding a pattern that has happened in the past as a result of random chance doesn't help predict what will happen in the future. You can easily see this by flipping a coin a sufficiently large number of times. You'll find several "patterns" where 5 consecutive heads are followed by a tails, but if you check every occurrence of 5 consecutive heads you'll find that the number of them followed by a tails is about the same as the number followed by a 6th heads.

A graph of a normal distribution over a given period of time will look about the same  regardless of when that period is, but the details will change. Make a list of the numbers drawn over a few different periods spanning 100 drawings and then draw graphs showing how many numbers were drawn once, twice, three times, etc. The graphs will be similar, but the indvidual numbers won't stay be in the same place. For one period 17 might have been drawn the most number of times, but in a different period it will have been drawn about an average number of times.

 Page 5 of 6