Rhode Island United States Member #56010 October 28, 2007 434 Posts Offline

Posted: May 18, 2008, 12:21 am - IP Logged

Computers and controversies aside, choosing lottery numbers by spinning around balls inside a machine is a random event...NO matter how low or high the odds are for a game, over time it performs according to mathematical and statistical expectations...Most lottery players base their picks by looking at previous drawings results...This is done in order to extract or pick on some key statistics they think will appear on next draw based on the previous drawings occurences...Obviously i have heard of Hot/Cold or "Due" but these are more of a "feel" statistics with no precise mathematical definitions...More precise definitions are of the following type: for example In pick 3 games:

Total possibilities: 1000, Singles: 720, Doubles: 270, Triples 10...if you play only singles all the time you can expect the results to be singles around 72 % of the time...also triples..there is nothing special about them..there is only ten of them and if you total up all the occurences and the total number of drawing you'll see that they come out ~1% of the time...Matter of fact i looked at the last 100 drawings of PA nite pick 3 game and the totals are: Singles: 76, Doubles: 23, Triples: 1..pretty close to expected mathematical formulas....

When it comes to jackpot games it gets harder to predict patterns in particular statistical categories because the odds are lower...Some stats are general and don't help all that much: say for a p5 game

having 3/2 2/3 even/odd or low high balls is ok in general but it doesnt help to narrow it down all that much there are still thousand of combos left...Wheeling on the other hand has a mathematical guarantee designed in it but before you start weeling you have to take a major guess at your main wheeling numbers...i dont like wheeling because of the high cost involved and since one has to guess some numbers at the beginning anyway it is not a rigorous statistical aprroach....

There are other more important statistics that help narrow down the picks more closely...My objective of this topic is to discuss and at the end come up with the 5 most important statistical areas or picks that matter the most...give examples if you can to illustrate what you mean..

Rhode Island United States Member #56010 October 28, 2007 434 Posts Offline

Posted: May 18, 2008, 12:55 am - IP Logged

Lets start discussion with Powerball 5/55 game..some people already know about this but they simply overlook the usefulness of this statistic because they dont know how to apply it..the key is to have a systematic approach..

1) Repeats from last draw

three options:

A) NO number will repeat 60% chance of happening

B) 1 Number will repeat 30% chance of happening (about 1 in 3)

C) 2 Numbers will repeat 5.5% of happening

if you decide to go with option A the benefit is that you eliminate all 5 numbers from last draw, now the game becomes 5/50 ( improve in odds from 3,478,761 to 2,118,760 or 40% improvement)

If you decide to go with option B the benefit is that you have a 1 in 5 chance of guessing what # will repeat and if you guess correctly you have found one number and also you have eliminated 4 other incorrect choices...now the game becomes 4/50 (or 1:230,300 way down)...there are ways to do this right: most of the time the repeat is a "hot" one and it usually repeats in a different position...or you can buy 5 lines and each of them has one different number from previous draw and discard the other 4 on each line...this guarantees the 1 in 5 choice and now you go on with the 4/50 picks

Option C is the best but it happens rarely...still it can be a beauty if you happen to catch it at the right time...the right way to apply this option would be to pick one hot number and one "suspicious" number that you would normally think will not draw again ...or you can buy 10 lines with two different numbes repeatis from last draw and discrad the three remaining others in each line...after this the game becomes 3/50 or 1:19,600

Examples of occurences of the three options by date:

option A: occured on 5/14/08 19-21-29-32-46 (previous draw was: 7-18-30-41-45)

option B occured on 5/17/08 12-14-15-21-43 (previous draw was: 19-21-29-32-46)

option C occured on 5/7/08 9-25-26-28-42 (previous draw was: 2-28-36-42-46)

with other filters on top of this it becomes more easy to make succesful picks

Dump Water Florida United States Member #380 June 5, 2002 3106 Posts Offline

Posted: May 18, 2008, 1:56 am - IP Logged

When we talk about filters I assume we're talking about starting with all the combinations so we know the 5# winning combination without the PowerBall is in there.

In that case the repeats for the last 283 draws were 54.8% for none 36.5% for one 7.9% for two 0.7% for three

Games like 6/49 are generally around 45% for none, though it is still a very good number.

The best filters are those that offer roughly a 1in2 or 1in3 chance of getting it right more often then not. Powerball Pick-5 portion . . .

Adjacent or Neighboring numbers. ie: 10-11 are adjacent.

67.5% none 28.7% one

Last Digit. ie: 02-22-32-42-52 have the same last digit.

36.1% none 49.8% one 8.8% two

First Digit. ie: Decades 20-21-23-26-29 have the same first digit.

6.57% none 46.4% one 25.8% two 15.5% two and two

Sum of the Digits in the numbers (35=3+5=8) that matchs . . .

Zeta Reticuli Star System United States Member #30470 January 17, 2006 10354 Posts Offline

Posted: May 18, 2008, 2:14 am - IP Logged

I would start off with "Is the game hittable?"

Remember the threads about the Tennessee lottery and how it ass impossible for a triple to be drawn, yet the state lottery accepted tickets with triples.

I'd definitely make that one of the five factors, or a prerequisite to get to five factors.

Those who run the lotteries love it when players look for consistency in something that's designed not to have any.

There is one and only one 'proven' system, and that is to book the action. No matter the game, let the players pick their own losers.

New Mexico United States Member #58526 February 18, 2008 683 Posts Offline

Posted: May 18, 2008, 2:49 am - IP Logged

Benmas: I agree with you 100% on your first post. Lottery drawings are random and the only hope we have of making smart picks is to try to find some kind of "pattern" we can exploit. For a long time I only bought quick picks. I started getting tired of quick picks that seemed very unlikely to hit. For example for a pick 5 something like: 5-6-7-8-30. I started trying writing all kinds of spreadsheets that would analyze hot/cold, even/odd, hit/skip....etc. It would all look good on paper, but never really panned out that well when put in to use for real drawings. So, now I just use microsoft excel to generate random numbers. Well, the numbers aren't really "random" in a complete sense, because I have the spreadsheet generate them according to "patterns" that I think will help my odds. For instance, for a pick 3 game, about 70% of my picks will be singles, 30% will be doubles (with maybe one triple thrown in there, heck you have a 1 in 10 chance that way if a triple DOES hit). For pick 5 games, I use decade patterns mostly to "guide" my RNG. For instance, I analyze all pick 5 of 39 games and find the most hit decade patterns (usually like 1-2-1-1 and 1-1-2-1) and I generate my picks based on that. Also some of my pick 5 tickets would be generated by the most hit high/low patterns of 3H/2L, 2H/3L, 1H/4L, and 4H/1L. These examples are a couple that I could think of that I use to basically manipulate a RNG to produce picks within some sort of set parameter(s). This seems to produce the best results for me personally. It's all luck, but a little basic observation and smart picking can't hurt..............

When we talk about filters I assume we're talking about starting with all the combinations so we know the 5# winning combination without the PowerBall is in there.

In that case the repeats for the last 283 draws were 54.8% for none 36.5% for one 7.9% for two 0.7% for three

Games like 6/49 are generally around 45% for none, though it is still a very good number.

The best filters are those that offer roughly a 1in2 or 1in3 chance of getting it right more often then not. Powerball Pick-5 portion . . .

Adjacent or Neighboring numbers. ie: 10-11 are adjacent.

67.5% none 28.7% one

Last Digit. ie: 02-22-32-42-52 have the same last digit.

36.1% none 49.8% one 8.8% two

First Digit. ie: Decades 20-21-23-26-29 have the same first digit.

6.57% none 46.4% one 25.8% two 15.5% two and two

Sum of the Digits in the numbers (35=3+5=8) that matchs . . .

30.8% none 51.6% two 10.8% three

BobP

you highlighted some very good statistics Bobp...from your list I rate as the most important in order:

1) repeats from last draw (i said that earlier too)...

2) first digit breakdown: (DECADES)

3) the adjacent or consecutive #s (10-11)

..these three are some of the main ones.. they give the best % splits...now it would be good to show how to use these filters in the most effective or systematic way....for example if you know that three numbers are going to come from one decade the probabilities are good that it will include adjacent numbers too..(indeed last pb draw was 12-14-15-21-43) ...and it also repeated one number (21) from previous draw...

so the main stats for last PB draw hit the following %:

repeats last draw: one 36%, adjacent one 28%, first digit or decade: three 15%

just will only these three filters one could have gone a long way...not to mention the others but of course as you add more filters your chances of error increase also...I like to discuss only 5 main ones and that is it for me ...5 filters

Rhode Island United States Member #56010 October 28, 2007 434 Posts Offline

Posted: May 18, 2008, 3:02 am - IP Logged

Quote: Originally posted by Texas MadMan on May 18, 2008

Benmas: I agree with you 100% on your first post. Lottery drawings are random and the only hope we have of making smart picks is to try to find some kind of "pattern" we can exploit. For a long time I only bought quick picks. I started getting tired of quick picks that seemed very unlikely to hit. For example for a pick 5 something like: 5-6-7-8-30. I started trying writing all kinds of spreadsheets that would analyze hot/cold, even/odd, hit/skip....etc. It would all look good on paper, but never really panned out that well when put in to use for real drawings. So, now I just use microsoft excel to generate random numbers. Well, the numbers aren't really "random" in a complete sense, because I have the spreadsheet generate them according to "patterns" that I think will help my odds. For instance, for a pick 3 game, about 70% of my picks will be singles, 30% will be doubles (with maybe one triple thrown in there, heck you have a 1 in 10 chance that way if a triple DOES hit). For pick 5 games, I use decade patterns mostly to "guide" my RNG. For instance, I analyze all pick 5 of 39 games and find the most hit decade patterns (usually like 1-2-1-1 and 1-1-2-1) and I generate my picks based on that. Also some of my pick 5 tickets would be generated by the most hit high/low patterns of 3H/2L, 2H/3L, 1H/4L, and 4H/1L. These examples are a couple that I could think of that I use to basically manipulate a RNG to produce picks within some sort of set parameter(s). This seems to produce the best results for me personally. It's all luck, but a little basic observation and smart picking can't hurt..............

Just a quick note: This post is not to say that Quick picks are not good or to discurage them..to the contrary: if you can buy Qps in addition to your own picks thats even better...the primary reason of this post is to find the most important statistical categories or filters that have the most potential in prediction future draws...and some comments how to use these filters

Coin Toss: i said at the very beginning "Computers and controversies aside"...i'm assuming that we are playing live ball drawings only and that they are random...You do make a good point about is it hittable, rigged etc..but i dont want to go off topic and discuss other things..there are other topics for them...your comment is valid nonetheless..the random and fair assumption comes first consideration

Dump Water Florida United States Member #380 June 5, 2002 3106 Posts Offline

Posted: May 19, 2008, 3:01 am - IP Logged

The filters I listed, Repeats, Adjacent, First Digit, Last Digit, Summed Digits, are the only ones with strong 1of2 or 1of3 choices. All the others are like Odd/Even, Low/High, Sums with 1of5 choices or more to make.

Even with 10 different filters a single choice from each (for which there is no rational decision to be made) you end up playing in a lottery that's generally as hard to beat as Pick-4 straight.

With too many filters you end up playing the filter lottery instead.

Still it would be nice to have a dozen or so filters that when correctly chosen could give us the winning numbers. Lotwin has about 27 filters, half of which are based on play slip positioning, with an analysis of any draw that shows what the exact filter settings would have been to have the winning combination. Interesting if not predictive. Now we need a way to predict what filter settings to use. Gail Howard projection suggests what filter setting based on the last draw, but not what to do when that already happened once as it happening three or more times in a row is less likely.

The filters I listed, Repeats, Adjacent, First Digit, Last Digit, Summed Digits, are the only ones with strong 1of2 or 1of3 choices. All the others are like Odd/Even, Low/High, Sums with 1of5 choices or more to make.

Even with 10 different filters a single choice from each (for which there is no rational decision to be made) you end up playing in a lottery that's generally as hard to beat as Pick-4 straight.

With too many filters you end up playing the filter lottery instead.

Still it would be nice to have a dozen or so filters that when correctly chosen could give us the winning numbers. Lotwin has about 27 filters, half of which are based on play slip positioning, with an analysis of any draw that shows what the exact filter settings would have been to have the winning combination. Interesting if not predictive. Now we need a way to predict what filter settings to use. Gail Howard projection suggests what filter setting based on the last draw, but not what to do when that already happened once as it happening three or more times in a row is less likely.

BobP

The discussion here is not limited to filters by using them in lottery software....A person can use filters by using his mind only when going to a store and choosing numbers...For example: One goes to store to play and notices the previous winning numbers posted on a board...He/She then decides that from last draw one number will repeat (lets say last PB draw he saw was 12-14-15-21-43) He/She chooses 14 as a repeat (and eliminates the other 4)...after that decides that since there were adjacent numbers last time, this time there will be none in his picks...There were no 30s last time and he/she decides that 30s will stay out one more time...(this will fall in the decades filter)...He/she notices that the last digits were all small numbers so decides that this time there will be two high last digits and repeated last digits...

SO his/her $2 own picks may look something like: 2-14-29-46-52 or 14-24-29-40-46......

So by talking about filters we are really talking about is what statistics are the most beneficial when making picks...thats what i want to discuss in this topic & at the end come to a general conclusion what are the 5 most important ones..whether you use lotto software or not is not that important here...Also the reason i said 5 filters is to keep the strategy relatively simple....if you go up to 27 different statistics then chances are pretty good you will confuse yourself and not make the right picks....

United States Member #5599 July 13, 2004 1185 Posts Offline

Posted: May 20, 2008, 2:54 pm - IP Logged

Hi,

I would suggest two catagories for the 5 most important filters. Individual number elimation and elimination of combinations. In either case, it is getting the highest amount eliminated with the least amount of error.

You are a slave to the choices you have made. jk

Even a blind squirrel will occasioanlly find an acorn.

I would suggest two catagories for the 5 most important filters. Individual number elimation and elimination of combinations. In either case, it is getting the highest amount eliminated with the least amount of error.

What exactly?...Just eliminate some Cold numbers?..Thats not a statistic, its just a thing that occurs because some numbers come up more meaning some others will sit idle for some time...

We all do that anyway..we eliminate certain numbers when making our picks...elimination helps but the filters should be geared more towards getting hits because thats where the money is

United States Member #5599 July 13, 2004 1185 Posts Offline

Posted: May 21, 2008, 10:59 am - IP Logged

Hi,

By eliminating numbers you increase your odds that you will hit. Mainly because you've reduced the amount of combinations. Examples in a Fantasy 5 game might consist of 1) numbers that have occured two times in a row or 2) numbers that have occured 3 times in the last 10 games or 3) numbers that have occured more than 8 times in the last 39 games. As always, there are examples where these things have occured, but it is a very low percentage of the time. Because cumulitive errors are multiplicative, I only works with error percentages of 5% or less. A few of those 20, 30 and 40 percent errors in your filtering almost guarentee your going to pick losing numbers.

Just an opinion. *S*

You are a slave to the choices you have made. jk

Even a blind squirrel will occasioanlly find an acorn.

By eliminating numbers you increase your odds that you will hit. Mainly because you've reduced the amount of combinations. Examples in a Fantasy 5 game might consist of 1) numbers that have occured two times in a row or 2) numbers that have occured 3 times in the last 10 games or 3) numbers that have occured more than 8 times in the last 39 games. As always, there are examples where these things have occured, but it is a very low percentage of the time. Because cumulitive errors are multiplicative, I only works with error percentages of 5% or less. A few of those 20, 30 and 40 percent errors in your filtering almost guarentee your going to pick losing numbers.

Just an opinion. *S*

OK, what you said is true most of the time but there are only few numbers that exhibit the 1) 2) and 3) you mentioned..even if you are right you can only reduce 3-5 numbers...ok i agree with you that eliminating certain "unlikely"numbers eliminates more combos and leaves better chances for success...however there are still many balls left..a better coverage strategy is needed... thats why i want to get to 5 most important statistical considerations also...you can still do your number elimination before or after the filters....the statistics mentioned above can tell you a lot more than just eliminating some numbers

United States Member #5599 July 13, 2004 1185 Posts Offline

Posted: May 22, 2008, 1:34 am - IP Logged

Hi,

Yes, there are methods that eliminate more numbers somewhat reliably. Simple observation and testing through pick histories appears to be a valid approach. All I would try to impart, is that even as strong as statistics is, it won't be enough. How many statistical experts have frequented the LP and other lottery sites? The result always appears to be the same....no winning system based on statistics alone. Therefore, because so many have taken the pure statistical path before with no success, there is no need to retrace a losing path. The answer might be a combination of methods with statistics playing a part.

As far as giving additional number reduction methods.....well...a work in progress. *S*

You are a slave to the choices you have made. jk

Even a blind squirrel will occasioanlly find an acorn.

Yes, there are methods that eliminate more numbers somewhat reliably. Simple observation and testing through pick histories appears to be a valid approach. All I would try to impart, is that even as strong as statistics is, it won't be enough. How many statistical experts have frequented the LP and other lottery sites? The result always appears to be the same....no winning system based on statistics alone. Therefore, because so many have taken the pure statistical path before with no success, there is no need to retrace a losing path. The answer might be a combination of methods with statistics playing a part.

As far as giving additional number reduction methods.....well...a work in progress. *S*

what you say is right but even if you reduced to half the numbers succesfully it is still too many...

How many statistical experts have frequented the LP and other lottery sites?

the statistical geeks use too many statistics thats why the end up confusing themselves most of the time...if you use only three to five key stats combined with your observations and you can do the picks in your mind with those (no software neccesary) than it works better..keep it simple...also the lotto programs they have way too many stats (even odd, high low, hot/cold)...these are not the most crucial filters, they are just a way to group numbers...