Pick-3 Positional Frequency

Reply #22

Lkydeb*594 — Thu, 02 Aug 2007 22:36:07 GMT

thank u BobP, yes, time sure did flew by we don't think about that. I will look into it see what comes up. If by any luck, I get anything out of them, I'll inform u.

Reply #21

BobP — Thu, 02 Aug 2007 19:03:03 GMT

Time does fly, back then it was 5/26 drawn three times a week, later 5 and later 7 just keeps expanding ways for people to put ever more into the game. In the same sense Pick-6 used to be once a week, now twice. Send an email to the Florida Lottery, they used to pretend they didn't know what I was talking about, would reply the ballset was changed for each draw or give me the date of the last draw. BobP

Reply #20

Lkydeb*594 — Thu, 02 Aug 2007 14:05:10 GMT

hmmm, well, p5 is drawn every day, the same as p3, so that makes me wonder why every 2 yrs. for p3. How do we know when they change the ball sets

Reply #19

BobP — Thu, 02 Aug 2007 01:29:47 GMT

Years ago when I last asked the Florida Lottery said they replaced the balls . . .

Every 4 years for Pick-6

Every 3 years for Pick-5

Every 2 years for Pick-3/4

Which makes a certain amount of sense considering how often the games are drawn. Maybe the balls get dirty or the numbers show wear, or they simply want to keep us from getting enough draws to make a better choice as to what numbers to play. If the process was as random as they claim they wouldn't need to keep rotating ballsets... [ More ]

Reply #18

Lkydeb*594 — Thu, 02 Aug 2007 00:43:56 GMT

BobP, I thought that Florida toss the ball sets every 3 yrs.

Reply #17

BobP — Wed, 01 Aug 2007 22:23:27 GMT

Somewhere on LP is a post by me explaining how to use the GH one ball like Powerball feature to setup three positional columns for Pick-3 and then use the D chart to pick the top five from each to wheel positionally. Problem is Florida tosses the Pick-3/4 ballsets every two years and starts fresh so the previous draw history just throws you off. BobP

Reply #16

dx123 — Wed, 01 Aug 2007 03:06:08 GMT

Hello BobP>>>

I am using 1645 draws with daily input [approx. 4 years worth of data input] in the Ontario P3 game for each of the 3 Positions using GH. The '*' numbers are worth looking at and usually make for an exact hit in some cases. Is this what you were looking for? I feel that around 2000 draws would make for a better review of the numbers. But, sometimes I hit the F4 in Board 6 and remove all the games until it shows the last 1000 draws only. The new data would then indicate the hotte... [ More ]

Reply #15

BobP — Tue, 31 Jul 2007 02:42:14 GMT

Working with Pick-3 positionally and wheeling positionally is only for straight play. example: I believe digit A will appear in the first position, digit B will appear in the second position and I'm unsure whether digit C or D will appear in the third position so I will play both ABC and ABD straight. No box to worry about. With three columns of top 5 digits by position = 125 straight combinations.

BobP

Reply #14

BobP — Tue, 31 Jul 2007 02:35:26 GMT

Draws in July. Drop a draw add a draw to the 22 for each draw's positional workout. I need to test this deeper. BobP

Reply #13

Stack47 — Mon, 30 Jul 2007 21:51:08 GMT

A smaller range of draws probably is better because the top 5 will change faster. Even with a 30 draw range, the top 5 might remain the same for two or three straight weeks.

How many draws did you use the past 22 draws

Reply #12

Stack47 — Mon, 30 Jul 2007 21:33:25 GMT

Many years ago when I was using the Scientific Number Selector, every Sunday I tracked positional frequency for the past 5 weeks (30 draws). And unless a digit hit two or more times the previous week, there was hardly any movement. If a digit in the bottom 5 hit twice the previous week it was still possible that it would remain in the bottom 5 the next week if its last hit was more than 5 weeks ago.

I never tracked how many times the top 5 digits created wins but the results were probably abo... [ More ]

Reply #11

sysp34 — Mon, 30 Jul 2007 13:47:56 GMT

BobP

using only top 5 (sort order) would be useless because there are others 125 top 5 combinations

i mean if we use combin(10,5) we got 252 unique combination and there are 126 combination for top 5

so far the 3 number groups 4-3-3 with combination 2389 - 014 - 567

to play straight we have 6 positional 123, 132, 213, 231, 312, 321 unfortunately there are expectation limits in 1000 draws because of the groupped numbers

220 boxed number for 3 number groups with combination

Reply #10

Big Hitter — Mon, 30 Jul 2007 13:09:44 GMT

I use positional frequency, but I weight each position. Position 1 gets a weight of 5 points, position 2 gets a weight of 3 points and position 1 gets a weight of 1 point. I do this to give credit for digits that pop first or second since the odds are a little higher against them. I usually look back 14 days and I mix midday and evening draws together. Good luck

Reply #9

RJOh — Mon, 30 Jul 2007 10:10:07 GMT

True but the odds of all 5 winning numbers being among those 12 numbers in a 5/39 games are 1:727.

Reply #8

BobP — Mon, 30 Jul 2007 08:47:41 GMT

I found a sweet spot for the month of July in Florida using 22 past draws with a fair number of straight wins with 125 numbers using the top five digits in each position, but it has now been 8 days since the last hit on paper. Have to do more research to see it this would have been profitable enough to carry over dry spells over a longer time period. Funny how it could go from winning nothing with 50, 60, 100 past draws and start winning at 22, guess I was trying too hard. btw: You don't have... [ More ]

Reply #7

BobP — Mon, 30 Jul 2007 08:39:55 GMT

I didn't do that right for Pick-5. To determine the number of combinations necessary for a full wheel matrix you would treat the wheel as a mini-lotto game, determine the odds for that game to learn the number of combinations necessary.

Say you want to know the number of combinations needed to make a Pick-5 full wheel matrix for 12 numbers.

12x11x10x09x08 = 95,040

The game is Pick-5 so 5x4x3x2x1 = 120

Dividing 120 into 95,040 = 792 combinations or tickets needed to make a full whe... [ More ]

Reply #6

BobP — Mon, 30 Jul 2007 01:59:02 GMT

Yes there is always at least one winning ball missing out of the columns of five. The old theory about the numbers rising to the top like cream must have gone out when the state homogenized the drawings. BobP

Reply #5

BobP — Mon, 30 Jul 2007 01:51:22 GMT

RJoh: Say you are selecting Pick-3 digits by position and you like three digits in the first position, two in the second and two in the third. Your playing matrix would be 3x2x2 = 12 numbers.

A-B-C = ABC ABF AEC AEF

D-E-F = DEF DEC DBC DBF

G = GBC GBF GEC GEF

If you like five digits in the first position, five in the second position and five in the third position the playing matrix would be 5x5x5 = 125 numbers.

In a Pick-5 game the full matrix would be AAxBBxCCxEExFF.

B... [ More ]

Reply #4

Stack47 — Sun, 29 Jul 2007 23:30:53 GMT

The digits that hit the most often in each position (or tied for most) is the first play(s).

It probably shows that the top 5 digits in each position had about 60% of hits and shows the smaller percentage even/odd and low/high bias too. It might even show had you played the most frequent digits in all positions, you may have broke even but the majority of the hits were combinations of high and low frequency digits.

There might be a spot-play by playing the low/high - even/odd bias.

Reply #3

JAP69 — Sun, 29 Jul 2007 13:10:24 GMT

I use positional frequency and positional due for exact order wheeling.

I try to cut my plays to a minimum so I loose out often.

I look in a lot of areas to find digits for reapeat or due by position.

One example is one that I just posted in my blog on finding digits in the front pair. It was the 2 ball that I felt was ready to drop either in position 1 or 2 in S.C.. It fell in both positions yesterday. So putting the digit 2 in position 1 and 2 for the wheel was a good to go for me.

I... [ More ]

Reply #2

RJOh — Sun, 29 Jul 2007 11:15:32 GMT

Sounds like you tried that positional frequency on a pick5 game using a 5x5x5 matrix. In my lottery program, when ever I check combinations within a file or with another file, all the winning numbers are high lighted and added up. At the end of each search the total hits per file and position are shown. Since the numbers in the file are sorted the information is useless because number one is always in the first position and the highest number is in the last position. It might work in a game... [ More ]

Reply #1

LosingJeff — Sun, 29 Jul 2007 08:20:25 GMT

Hi Bob,

Try your test on Indiana. It should be easier to figure out seeing how they paid 3 straight winners on Pic 3 yesterday (0 midday/3 evening) With the Lotto being as high as it is you would think that there should have been 500 straight winners.

Pick-3 Positional Frequency

BobP — Sun, 29 Jul 2007 07:14:17 GMT

I conducted a short test of Positional Frequency for Florida and got nothing. I was wondering if anyone is using Positional Frequency for Pick-3 with any success anywhere and if so, where and how many past draws are you working with.

For those who don't know. Positional Frequency counts how often each digit has hit in each position over a range of draws. The digits that hit the most often in each position (or tied for most) is the first play(s). For example, working with the top 5 digits in... [ More ]