Lottery Lotto Gambling Software Systems

[mitachoo]

I had read the discussion in the "Original Pick 3 Strategies" section of http://saliu.tripod.com/Newsgroups.htm#Strategy concerning playing the Pr >=100 filter only when the Running Skip was at a certain level. This seemed like a powerful way of "fine-tuning" when to play or not to play, but because it was based on looking at historical data, I had to ask myself 1) how effective of a playing strategy is this going forward, and 2) do filters really "prefer" to hit at certain skips?

I did extensive back-testing using Pr from my state's Pick-3 lottery (data current through 10/13/09, PM drawing). My data and analysis is shown later on in this post, but to summarize my conclusions:
1) Yes, Pr does appear to "prefer" to hit a certain running skips, and the value of the "preferred" skip changes from time to time.
2) No, this is NOT an effective playing strategy because there's no way to tell which skip is "preferred" until after the fact!

Here's my thought process. Please feel free to comment if you see any glaring holes or inconsistencies in my analysis.

Using the same procedure as outlined in the link posted above, the current Running Skip for (Pr1>=100) is 3. Using a Parpaluck of 200 (the default Parpaluck from Bright 3), there were 4 hits that had a skip of 3 versus 20 hits that had a skip of >=3, so I would say the “Projected Hit Rate” (PHR) for (Pr1>=100) for the next drawing is 4/20 = 20%. I repeat this procedure for a shitload (I know, a highly technical term) of drawings. Go to http://sites.google.com/site/pr1phr/ to see what the top part of my file looks like.

I did this for 5,000 drawings and looked at a Histogram of PHR (x-axis) versus "Hit Rate on Next Drawing" (y-axis) and found there is no correlation at all between the PHR and the likelihood of hitting on the next drawing! Go to http://sites.google.com/site/phrhistogram/ for the data and Histogram. If there was any correlation I would expect the data to track along the red line but as you can see if runs straight through it and tracks around the average hit rate of (Pr1>=100).

The data gets "less spikey" if you use more drawings (I went up to 50,000 at one point). If you use much less than 5,000 drawings, the Histogram gets VERY spikey, which led me to the false conclusion that some skips are EXTREMELY preffered over others, but of course the real reason is that you have less opportunities (i.e. data) for the "true" hit rate to manifest itself.

And one last point . . you can clearly see from the data in http://sites.google.com/site/pr1phr/ that skips 2, 4, and 7 seem to be "highly preferred" but again that is in retrospect and it's not clear to me how one would make good use of that information.

Thoughts anyone?

[marcher]

How did you get to that site? There is also a well placed caveat on that web page:

“Please be advised that this is a replica of an original web page serving the purpose of backup — in case the real page becomes unavailable. This page is NOT updated. For the latest, always go to:

Free Lotto, Lottery Software, Excel Spreadsheets, Strategies.”

Most filters in that version of the software were changed — some of them dramatically. My software calculates now the median of every strategy. In that case, the median was 5. You could have made it 6, even 7. Play the strategy only for current skips equal to or less than the median. It saves lots of money. Also, apply other filters to reduce the cost.

If the current skip of a strategy goes beyond the median (or 6 or 7, in this example), just put it aside. Wait for the next hit. Meanwhile, play another strategy where its current skip is under the respective median.

With more layers, there are more filters, therefore more strategies are possible.

Best of luck!


Ion Saliu
PS
You can really be tedious, Mitule! You told me it was so because of your engineering training!
IS

[marcher]

The lottery has far, far worse odds than the casino games. Yet, the lottery offers a big luxury. You can study lottery strategies in the privacy of your home, undisturbed. You don’t have such a luxury in a casino. You can see easily the streaks and the skips. The software also calculates the median now (5, in this case):

[mitachoo]

You're not kidding! Once I get off on a sidetrack like this I can lose many hours to endless tinkering. It has occured to me on several occasions that I spend much more time on tinkering with things like this rather than looking for strategies!

[marcher]

Okay, while we are at it. This particular pick-3 strategy based on pairs is one of the most talked about. It is one of the first lottery strategies I discovered. I saw it as early as 1986, when I used an Atari XL! The strategy also attracted the most attention from the users of my software.

PR = 100 is actually the median times 3. You can go higher than that, like median * 5. The reverse is also true for the MAX of the filter; e.g. median / 3 or median / 5.

Look at this portion of the skips:

0 4 1 0 5 4 1 6 2 5 5 5

There are 12 consecutive hits under 7. If we add the number of draws as in that string, the calculations are not accurate. As you know, the skip deducts 1 from the gap between hits. If it is zero, actually it means playing 1 drawing. So, the total number of draws to play actually is:

1 5 2 1 6 5 2 7 3 6 6 6 = 50 drawings to play.

You will notice that it is fairly easy to reduce the total of combinations generated by the strategy to 50. The cost would be 50 x 50 = 2500.

Total wins: 12 x 500 = 6000

You can reduce total amount of combos to 20, for example. You can afford to miss a few winning situations and make a better profit. You can even raise the bet: Up to $5 per ticket in pick-3.

No matter what, you only play for under the median, plus 1 or 2. It goes above med + 2 (7, in this example), you stop playing. It is natural also that there are not many consecutive hits of above median. Like we had in this case: 2 10 12 5; 0 8 9 2; 0 14 9 15 4; 6 31 21 3. Looks like 3 consecutives is the worst…

I’m gonna play something tomorrow, after I mail some court documents to Michigan…

[Advertisement]