Patterns

[george]

Hi lotto experts,
I play Lottario for about one year now, and I made up some basic strategies, as all of you probably did, too (based on what I read so far): exclusions, hot/cold/due numbers, then wheel and filters, stuff like that.
I play 21 numbers.
I'm not that successful: max winning was 4 numbers few times. I return only 40%-50% of what I paid. That's just to introduce myself and what I do

I have a feeling that these strategies could be improved, actually they could be reconsidered essentially and even replaced by what I call "patterns".
If we assume that the draws are not really random, and the next draw results can be somehow predicted based on previous draws results (otherwise all "systems" just make no sense), then there must be some patterns. I assume that some of the balls tend to hit more often than the others, for some period of time, say until the set of balls is replaced by another set (I believe OLG does it from time to time). I believe the following pattern could be found: there are 21 numbers (or N numbers) out of which at least 3 numbers (or maybe even 4) hit during the last D draws (D could be 40 or 50, or 25). What do you guys think? Has somebody tried that?
Thanks!

 

[Icewynd]

Hi George & Welcome to the Forum!

I do believe that the lottery draws are random, but I also believe that randomness has trends and characteristics that we can learn and exploit. There is much of interest in the archives of this board, so use the search feature to see what you can find.

Recently there were some posts by member Frank on the Markov Chain theory and he introduced his method of finding which lottery numbers may follow the current set of numbers.

You might also be interested in the Number-Pickers Challenge posts to see how other players are doing with their picks.

Good Luck!

 

[george]

Thank you Icewynd, I will review those posts (I already reviewed some of them). Here are my thoughts though: if the lottery results are computer generated, then you are absolutely right, I 100% agree with you that lottery draws are random. But we deal with the real balls, that have some physical characteristics (zillions of them, I know), and if the pattern (in terms of pattern that I described) can be found - we could exploit it
Of course, it's only theory, I will try to find such pattern(s), interesting if somebody tried it already... I have few more questions for those guys

 

[Icewynd]

Yes, you are right. The ball machines, which we use here in Ontario, do introduce some physical forces on the balls which factor into which balls are eventually chosen. The thing is, how can we measure and use this information?

I would rather rely on two characteristics of randomness: trendiness & reversion to the mean.

By "trendiness" I mean the way that what is hot stays hot for a while and what is cold can stay cold for quite a while.

Reversion to the mean is the principle that dictates that what is too far out of its normal distribution will eventually return to normal.

Good luck!

 

[george]

Yes, you are right. The ball machines, which we use here in Ontario, do introduce some physical forces on the balls which factor into which balls are eventually chosen. The thing is, how can we measure and use this information?

That's the thing! I don't care about ball weights, or how ideal their shape is, or how slippery they are, whatsoever. If pattern can be found - that's all we need!! Some of their physical characteristics contribute to the fact that some balls tend to hit more often, and I don't care which exactly characteristics do that

 

[Frank]

Predicting the lottery is a bit like predicting the weather. Todays meterologists can give a flavour of the weather to come in terms of temperatures, rainfall wind speeds and the approximate parts of the country which will be affected. Due to chaotic elements in weather behaviour, its not an exact science in terms of exactly when and where. I think lotteries have 'seasons' where for a short time certain families of balls are in favour, such as mainly hot or mainly cold balls, seasons when balls from a particular decade are missing, that kind of thing. If you can spot it and correctly assume that it will continue in the short term, then you are on to something.

On the subject of ball physical characteristics leading to drawing patterns, I can only speak for the UK lottery but with 8 sets of balls to randomly choose from and currently two of 6 drawing machines being used, its hard to believe that physical characteristics are not being ironed out. Having said that, theres something about balls 13 and 38 that make them polar opposites. if you include the bonus ball number 38 has been the clear leader for about 20 years, whilst 13 has been the laggard over the same period! You might think some size and weight variance could account for this, but not so with so many machine/ball combinations. Its bizarre.

 

[jack]

Hello, in lottery can only be mathematically up to 80% or 75% accuracy, as in the law of pareto80 / 20, the remaining 20% 25% will always be random, then around that try to predict Mathematically the 3-4 numbers of 49 / 6 for use as fixed, as are the remaining random, see for its frequency cycle and random 80/20 or 75/25. There must be rotating patterns, ai the number or not this is in default

 

[jack]

Hello Frank, you have Excel features, could be studied to predict 75% of a draw, the lottery example 49/6 predict 3rd 4 numbers to secure, use as fixed, the rest would be random (uncontrolled) as the repetiçaoes last draw etc, then we would have two parts, 75% research and 25% random, creating a system on top of it, another example of a lottery 39/5 predict trios, to secure the number two missing, random

 

[cdrake]

Years ago I remember the lottery commission televising the draws generally around the evening news. And what bugged me was the fact that they didn't seem to mix up the balls with any continuity from draw to draw. They talked about how they ensure the balls are of the same weight and the same size blah, blah, blah. But I never heard them say anything about ensuring the balls are mixed for a set minimum time before the extraction begins. When you see some of the "Kooky" combinations where you get 4 or more numbers coming from one decade especially the first two you gotta wonder if this is there way of manipulating the draw in order to thwart any of the real "Playa's" out there. The commissions whether here in N.A. or in any other part of the world can see patterns in the draw developing as the draw date approaches and maybe this is a way that they can throw a kink into mix to prevent any one person or syndicate from garnering an advantage. I could also be totally wrong but that's the "Conspiracy Theorist" coming out in me.

 

[blitzed]

yeah, they must filter out draws with too many sequential numbers, or goofy sets...test-draw procedures for many lottos sound convenient eh?

6/54 Lotto Texas is the most recent one I've read about:
http://www.txlottery.org/export/sites/lottery/Games/Lotto_Texas/Pre-test_Results.html

such test-draw procedures for any lotto, lead to conspiracy theories...what if? in this case, they tape 7draws...then cherry-pick the one with least payouts, so they can keep that jackpot rolling over, to build more lotto frenzy?

cya,
blitzed

 

[george]

Thank you All,
Very interesting, especially blitzed's post. It actually means that there are few lotto machines (two for the lottery?) that they switch once a month in the beginning of month, and few ball sets (four?) that they select randomly for each draw. I believe OLG does pretty much the same.
Hmm.. That's a bad news.

 

[george]

I think here is the proof. Look at the test results in Oct - Dec 2013:
http://www.txlottery.org/export/sit...o_Texas/Pre-test_Results.html_1782164351.html, especially at 11/16/2013 results, in red. They failed. Ball #53 hits 6 times out of 7 consecutive draws!

11/16/2013
Test# 1 J* 16* I 12 47 - 21 - 40 - 53 - 28 - 44
Test# 2 J* 16* I 12 53 - 44 - 51 - 34 - 22 - 31
Test# 3 J* 16* I 12 18 - 2 - 13 - 15 - 12 - 8
Test# 4 J* 16* I 12 48 - 53 - 14 - 44 - 54 - 3
Test# 5 J* 16* I 12 24 - 54 - 49 - 48 - 43 - 53
Test# 6 J* 16* I 12 23 - 53 - 36 - 44 - 41 - 33
Test# 7 J* 16* I 12 40 - 33 - 22 - 20 - 53 - 21

What is the probability of such event? Next to zero. For 6/54 game, the probability that a particular ball hits in the draw is ~0.11 (correct me if I'm wrong). The probability that it hits in 6 consecutive draws is 0.11 in power of 6, which is 0.000001772, i.e. 1 in 564,473! OK, we have 7 draws here, so I (roughly) assume it is 1 in 200,000-300,000, something like that. Still almost impossible, but it happened.
If you look at the other draws on this page, you will see that #53 hits very often when ball set #12 is in play.
Look at #44 - it hits 4 times in 7 draws on the same date, and it hits twice again in the next six consecutive tests.
Lottery is random?

Few problems here in finding how to exploit this idea:
- Algorithm of finding such patterns... There should be different patterns for each ball set, but as opposed to Texas' lotto company, OLG doesn't disclose which ball set was used in each draw (or maybe somebody can point me to such disclosure?), so we are blind here. Still, I believe it's possible, but it complicates the whole thing a lot.
- We don't know which exactly ball set they will use for the upcoming draw (I don't think lotto machines matter too much, I think it's ball sets that matter, at least I assume that). This can be predicted somehow, or we can try and cover all cases (expensive, yes...)
- Another thing is that (don't laugh), among zillions of factors that contribute to lotto results, are such things as temperature, humidity, pressure, etc, in the lotto room, so the pattern may vary from one period of time to another (maybe from month to month, or season to season?).

If anyone has any productive idea(s) of formalizing these things - just let me know please... I don't use Excel, it's all in my DB (i'm a developer), the results from Sep 1997, Lottario, so I can write the code that would do whatever we want, but I just kinda stuck. I probably need a mathematician
So, if you:
1) share this idea,
2) have some interesting idea, and, most importantly:
3) are mathematician:
we can meet offline, I live in Toronto. Please send me PM. Short description of what you think and few words about yourself would be greatly appreciated.

 

[george]

I just noticed that there is no such thing as PM, is that correct? Anyways, we can figure something

 

[george]

Roughly, I did something myself. Not perfect for sure, for instance I cannot figure when they switch ball sets. But I can figure when they play the same set, it happens in 50% of the cases, not sure yet though, because statistics is too low for now. Still looking for fellow mathematician