Following, then playing hot combinations

paul762's avatar - lion

How good can this method be? if you chart say 100 draws and then play the best combinations of numbers for your lottery is there an increased chance of those combinations coming out in future drawings, I say this because as GH says follow the trend/the trend is your friend.

paurths's avatar - underground

Hi Paul,

100 draws?

One draw every day, or 1 draw a week?

That shouldn't matter, right?

Assume one has 2000 draws or something, for backtesting.
The way to go, theoretically ofcourse, assuming again that there is such a thing as "the way to go", would be to split up the 2000 draws in blocks of 100.
Then, per "block", see if combinations of numbers that seem hot within that block hit again somewhere in the next 100, or do they stay away for e.g. 300 draws and then surface again.
Not an easy task, b/c how to define the 100-draw-block...? With what draw should one start the creation of them blocks? Draw 1? Draw 25? Or draw 143?
To test this manually is pretty much the same as growing wings and fly off to the moon.
Even to write software that does this is not that simple. Not that it can't be done, but i think it would take quite some work.

Some numbers seem to "like to fall with each other".
Find these, see how many times they come back within an acceptable playing-period, find out how long it has been since they last fell together and decide when to start playing them, along with some other numbers. Wheeling, most likely.
Might be one could wait years for the combination to return, especially with let's say 2 draws a week, no good. One could go insane lol.
Might be the combination returns several times within an acceptable timeframe, hopefully accompanied by the other numbers choosen.

Very nice, all, but first a definition of "best combinations of numbers" must be found.
How would one identify this "best combinations"?
Just b/c some numbers have fallen very much together in the past, near or distant past, doesn't mean they will fall plenty of times together again.
In the end, the goal is symmetry, balance. A number can grow hot and fall like crazy within a hundred draws, then it fades away, for hundred or more draws. (depending on the amount of numbers in the game) And another, or a few other numbers become the hot ones.

In Belgian lotto, 6/42, i found that since 1978, it never happened that the same 3 numbers fell the best/most during a timeframe of a year. Every year other numbers climbed the ranking.
How many times did it happen that the winning 6 numbers came from the top 10-numbers? (numbers that have come out the most)
Answer: 1 time. (in Belgian lotto, that is)

These kind of findings wont help directly in finding the winning combination, but it will help in eliminating useless, statistically!, combinations.

RJOh's avatar - chipmunk

If you have enough data you don't have to depend on GH's words about trends, you can check it out for yourself. 

It been my experience that while numbers do follow a trend, the trend cover a group of numbers too large to guarantee a win with a few lines.  For example the other day I played Ohio Rolling Cash5 using a group of 20 numbers that were mostly likely to hit based on a trend and all 5 numbers were in the group, but the best I could do in 25 lines was to match 3 twice and match 2 three times.  There were 15,504 possible combinations using that group and my 25 lines didn't contain even one of the 75 combinations that would have matched 4 but two of the 4550 combinations that matched 3. 

I've yet to find a trend that covered less than 15 numbers which would still include 3003 combinations.  Even then most times those trends will only include 3-4 of the winning numbers.

paul762's avatar - lion

the reason I ask is because If I play the irish lotto and if I play just 3 numbers and those numbers are drawn in the 6 numbers from 42 it pays 500/1, so I was thinking about using hot combinations triads to use for the next draws, however after checking through GH it shows that this method of selecting the 3 numbers doesnt work, any other suggestions?

time*treat's avatar - radar


The website I checked showed small or no prizes for match 3/6 in Lotto, Plus 1, & Plus 2. Where did you get the figure 500/1?

paul762's avatar - lion

its not the lottos that pay the 500/1 - its the uk bookmakers

time*treat's avatar - radar

At those rates, I'm on the wrong side of the pond Eek, if you only need 3/6, unless you only get 1 chance at it. Thing is, some of us say "numbers" when we mean "combinations". But if you have as many chances as you want, I say go for it. As to your orig. ques., IMO 100 draws will give you too many "best"s. If I were going to use that strategy, I'd cut the lookback period by at least 1/2. 50 draws, or less. By the time you're up to 100 draws, the 'old' popular groups of 3 have had their run and new ones are coming to the fore.

cahaba's avatar - moon

                                                                                    I Agree!

paul762's avatar - lion

there are 7 numbers drawn which include the bonus but if you pick 3 numbers and get 3 from the main 6 that are drawn they pay you 500/1.

I had not thought about just going to find the best 3 number combination/triad from the last 50 draws, I was using all history and leaving the last 100 draws to see if the best combo came out in those 100 draws, ideally as there are 3 draws on the same night I would like to find a set of 3 numbers that can be played for the next 100 draws for each of the 3 draws every wed and sat night. it would be good to find the ideal number of draws to look back through to find the best 3 number triad that produces the most wins in the following 100 draws, but I dont expect to find the solution.

time*treat's avatar - radar

I will leave out the bonus number for the moment 

42-Choose-3 gives you a possible 11480 'triads' in the game 'universe'

6 choose 3 gives you 20 possible 'triads' in the drawing, of which, you need 1. 

For your method, I would go back and see if there were/are any 'pairs' that tend to repeat. If so, then use that as a base for picking your triads. If not, then it is probably better to look for 'new' (not drawn before, or in a very long time) triads & pairs.    

JesterNow, if you'll pardon me, I have to go and brush up on my 'other' English. SmashCheerio. Going on holiday, and all that.

TopEnd of thread (1 page)

Welcome Guest

Your last visit: Thu, Oct 29, 2020, 12:02 am

Log In

Log InCancel

Forgot your username?

Forgot your password?