Can it pass the "So what" test?

Reply #26

RJOh — Sun, 06 Sep 2009 22:34:12 GMT

I would think players having 3 numbers right in every combination they played in a 5/39 game would upset the apple cart even if they only played one line and the same for a 6/49 game if they had 4 numbers right. Either one is not likely to happen very often.

Reply #25

BobP — Sun, 06 Sep 2009 22:25:40 GMT

I was thinking what worries lottery directors doesn't have to be what any one player does, and could include an idea distributed across enough players to make a difference.

For example a random number selction system for Pick-5 that includes a wheel with 3 key numbers against all the other numbers. For 5/39 this would take only 18 lines and guarantee two 4 number prizes when the 3 keys are correct.

For Pick-6 a wheel with 4 key numbers against all the other numbers. For 6/49 this wou... [ More ]

Reply #24

time*treat — Fri, 04 Sep 2009 18:44:43 GMT

The lotteries do their rigging in the odds/payout space; any type of deliberate encoding of the numbers runs the risk of being DEcoded.

Reply #23

Coin Toss — Fri, 04 Sep 2009 04:58:15 GMT

RJOh,

Yeah, I'm pretty sure the lotteries have people doing just that (what a job!).

bashlet572

Thanks. It is nice when we get discussions going.

Reply #22

bashley572 — Thu, 03 Sep 2009 09:54:20 GMT

GREAT discussion everyone!! Coin Toss, I like the idea you are proposing as I try to look at the lottery every way possible to look for trends or 'ways' it might be moving.

This is the BEST of LP as there is an open discussion of ideas

Reply #21

RJOh — Thu, 03 Sep 2009 00:41:42 GMT

For the lotto and a so what test, I'm trying to do a reverse approach here. Picture ourselves as on a lottery commission and we're given an assignment to see if anyone out there has come up with something we should be concerned about, not cheating, but something that produced consistent jackpot and second tier winners.

I would think that's a permanent assignment for every lottery commission. They probably constantly monitor sales and payouts as was mentioned in the PowerBall story concerning... [ More ]

Reply #20

Coin Toss — Wed, 02 Sep 2009 21:54:18 GMT

BobP

He meant that of the ganes a team lost, they beat themselves , not that the other team outplayed them. John Madden basically said the same thing about pro football, statistically more games are decided by turnovers.

For the lotto and a so what test, I'm trying to do a reverse approach here. Picture ourselves as on a lottery commission and we're given an assignment to see if anyone out there has come up with something we should be concerned about, not cheating, but something that produ... [ More ]

Reply #19

BobP — Tue, 01 Sep 2009 23:18:06 GMT

Most ball games are lost, not won. Casey Stengel

BobP,

Since Casey Stengel meant games are lost by errors, bonehead plays, bad base running, etc., how can we apply that to lotto?

Meant it as bit of humor because lotto is a ball game, now I'm not so sure what he meant as every base ball game has one winner and one loser, so how can more games be lost then won?

I appreciate your attempt to look for something beyond the usual suspects.

BobP

http://www.geocities.com/wal... [ More ]

Reply #18

Coin Toss — Tue, 01 Sep 2009 23:00:05 GMT

BobP,

Since Casey Stengel meant games are lost by errors, bonehead plays, bad base running, etc., how can we apply that to lotto?

Falling in live with a system even though it just doesn't produce winners?

RJOH,

Goos examples.

time*treat,

The Union Plaza (at the time, now it's the Plaza I think) in Vegas wanted to increase the take in Keno. This was prety simple and did the trick- for the same numbers chosen they offered a second drawing, so called the first one Red game and th... [ More ]

Reply #17

BobP — Tue, 01 Sep 2009 22:58:48 GMT

Aw crap, never get old. That was the 40th line I skipped when posting the build, then it got mixed in when I sorted the lines and another was left out.

Reply #16

time*treat — Tue, 01 Sep 2009 18:46:16 GMT

We had a game called red-white-blue; ten numbers in each color, two numbers pulled from each group (2 red, two white, 2 blue) and you had to match number and color so 1,2,3,4,5,6 was not the same as 1,2,3,4,5,6. 'Took one look at that game and said ... ok, well, nevermind what I said. That game died rather quickly, though.

'Could probably do a whole thread on Games that needed to be taken behind the shed and killed with a blunt object .

Reply #15

RJOh — Tue, 01 Sep 2009 18:22:14 GMT

I have never heard anything other than lack of public interest as a reason for dropped games.

When Ohio dropped SuperLottoPlus (6/49+1) they used a similar excuse but I suspect they could no longer afford to increase the jackpot by $2M after every roll because it wasn't attracting the number of players it once did. When the game first started it was the best game in the area and many of its players were from the surrounding states. At surrounding states added better local games along with Po... [ More ]

Reply #14

time*treat — Tue, 01 Sep 2009 14:52:12 GMT

There are two basic ways to go on short-term trends:

1) design a system that adjusts itself to short-term trends or

2) design a system that is based on a longer haul pattern and ignores short term variances.

The hit-the-target exercises in Statistical Process Control come to mind... and also roulette (will RED come up for the fourth time in a row or will BLACK finally show?)

When a game gets dropped, I don't expect that even if someone had cracked its code that that fact would be pub... [ More ]

Reply #13

RJOh — Tue, 01 Sep 2009 13:47:59 GMT

When you sorted your list of 39 combinations, you ended up with 40. Where did the 1 7 13 24 31 come from

Reply #12

BobP — Tue, 01 Sep 2009 07:30:38 GMT

Most ball games are lost, not won. Casey Stengel

Reply #11

Coin Toss — Tue, 01 Sep 2009 05:45:38 GMT

time*treat

No need to stick to Pick 5, but as far as jackpot games go if you're going to hit a jackpot in your lifetime it's certainly more likely to be a Pick 5 than a Pick 6 or 5 + 1.

But for Pick 5 a 5/39 matrix is good.

As far as systems, for example, let's ay that we know for a particular game (specific game, Pick 5 and state) with a 5/38 matris that there are 19 perimeter numbers - along the outside of the playslip, top and bottom lines and outer columns, and 20 inside numbers.... [ More ]

Reply #10

time*treat — Mon, 31 Aug 2009 23:09:59 GMT

@CT

Ok, if we stick to pick 5, then I think we have to ask how big X% is. With P5 (in most states) you don't have to be right very often to do quite nicely.

Also, we need to know the matrix size to use certain filters; your odd-count/even-count distribution will be different if the maxrix size is odd (35,37,39), than if the matrix size is even (36,38,40

Reply #9

Coin Toss — Mon, 31 Aug 2009 21:30:49 GMT

BobP

Using the Illinois 5/39 payoffs, 2 of 5 pays $1, 3 of 5 pays $10, 4 of 5 pays $100.

However, if the list was played live (real money), you have 13 states listed above, the wild-Card Multi, and Tri-State Megaball.

That comes top 15 X's 39 = $585 (dollar plays).

Using the Illinois pay table above, you have

2 X 39 = $78

12 X 10 = $120

1 X 100= 100

________________

= $298

$585

-$298

_______

= - $287 loser

= SO WHAT

If I'm runing a lottery I'm going... [ More ]

Reply #8

BobP — Mon, 31 Aug 2009 19:11:40 GMT

After I posted the list of 39 combinations further up in this thread I went to the Lottery Post Results page and looked at every Pick-5 draw showing on Sunday, so if the last Pick-5 draw was Friday, I used that as the draw for that state.

All in all there were multiple 2 number hits for every game, in Florida you win a free Quick Pick for 2 numbers.

There were 12 hits for 3 numbers, more if you count two were multi-state.

Hit IL+IA, VA, MA, MN, FL, PA, NC, WA, TN, SC, KY, WildC... [ More ]

Reply #7

JKING — Mon, 31 Aug 2009 13:03:46 GMT

Hi,

Why beat a stst to death that doesn't yield what you want? For me, I'll take that stat in a different direction. For example, how often do 2O/3E repeat, how often does 30/2e follow a 2o/3E pick, ...etc.

What Odds/Evens are hot or cold?....Compare Odds/Evens to other filtering criteria.

Just a thought.

Reply #6

time*treat — Mon, 31 Aug 2009 11:18:34 GMT

Kansas lottery drawing glitch picks same numbers for 3 days (2005)

https://www.lotterypost.com/news/125196

A computerized drawing system glitch caused the Kansas Lottery's Pick 3 game to draw identical numbers in the same order three days straight, from Dec. 16 through Sunday.

Apparently some players caught on. By day three of picking 5-0-9, ticket sales were $12,223,and the payout was $23,480, which is 192 percent of sales. The usual payout is 50 percent to 60 percent, Kansas Lottery... [ More ]

Reply #5

BobP — Mon, 31 Aug 2009 06:28:50 GMT

Pick-5 let's say we're talking about a brand new 5/39 game.

I would open up CoverMaster and select . . .

Pool - 39

Pick - 5

Hits - 3

Match - 5

[Generate][Timed Random Search][1sec][Up To Max Tickets][8]

[Start][Finished]

[Tool Bar][Optimize][Options][Preserve Order][Start][Finish]

[Tool Bar][Lock All]

[Generate][Timed Random Search][1sec][Up To Max Tickets][16]

[Start][Finished]

[Tool Bar][Optimize][Options][Preserve Order][Start][Finish... [ More ]

Reply #4

Coin Toss — Mon, 31 Aug 2009 02:55:35 GMT

time*treat

The lottery loves it when anyone loads up . Even if there were 2 odds and 1 even for a week nothing says that would continue, it may very well change the night you load up, and without knowing the specific numbers, it's still a so what.

Unless of course, some body is booking that kind of action, or you are the one booking that kind of action, then you have something.

Like RJOh said, this applies to most systems.

The only thing I can think of that's vague yet may be good to... [ More ]

Reply #3

RJOh — Mon, 31 Aug 2009 01:21:50 GMT

Even numbers in the last 437 MegaMillions drawings since the last matrix change had a small edge, both WB and MB. So what?

WB MB

0 = 4 2

1 = 55 28

2 = 146 79

3 = 156 81

4 = 67 33

5 = 9 4

1128/2185 227/437

51.6% 51.9%

With 56 WB and 46 MB, there are an equal amount of odd and even numbers in both groups. The combination that won the $333M jackpot had 4 odd and 1 even numbered WB with an odd number MB.

Reply #2

time*treat — Sun, 30 Aug 2009 23:56:53 GMT

More odds than evens, X% of the time isn't exactly So what? depending on the value of X.

That skews certain combinations forward. Imagine if a state's RNG pick-3 started to pull 2 odds and 1 even. Every day. For a week.

Few people would say So what? Some would kvetch, but a smart few would start loading up on 2 odds, 1 even bets by the 3rd day.

Reply #1

RJOh — Sun, 30 Aug 2009 23:54:07 GMT

What you have described pretty much covers every lottery system I've ever seen both personal and commercial. They all reveal some interesting facts about combinations drawn in the past but so what? . Most players want to know how to pick the next winning combination not something that happens 5% of the time and if they pick their combinations a certain way, they'll have a good chance of winning when it happens again. I'm not sure any software or system can do more than that.

Can it pass the "So what" test?

Coin Toss — Sun, 30 Aug 2009 23:04:01 GMT

Let's say we're creating a system, any game, starting completely from scratch.

And let's say we're going to try to to build on it by factors that can pass the so what test.

By so what I mean things like this, say we're working on a Pick 5 game and someone says that X% of the time there are more odd than even numbers.

Until some lottery out there starts book action on More odd or evens on the Pick 5 , this stat really doesn't tell you that much.

Nice to know, but thus far, so what? If... [ More ]