How to Train Your Excel File to Play the Lottery


It is not that hard to do but it takes a Macro.  Just about any formula can be created and then checked back in the Static Data.  You can also Track the Options by themselves in the Static Data but the Columns grow and So does the File Size, eventually becoming to large to function.  That is where the Macro comes into use. 

The Macro functions by bringing a Game from the history and playing a Value for a Filter Option to see if it hits.  With a Positional System in a 5/39 Draw Game 5 Guesses are required to at least get the winner into a sizable Group.  Getting smaller Groups requires more filters. 

Training is really nothing more than Bing Watching a “few” draws to see what Values are hitting in each position and trying to figure out what Values like to pair up. Kind of sounds like a Lottery within a Lottery Except that there is a direct connection between the outputs from Both.  Training involves making a list of combinations of values that you can afford to play and then Binge watching the Draws again to see if that worked.  Each time you do this you look at the game summaries hopping to develop patterns that get the most correct Values.  Trying to develop a set that wins and turning to the Next Game with those Sets in hand.

To Give your Excel File a “say” in the output you can program in a few if/then or Select by Rank formulas that select the individual values and their sequence.   While this might seam Dynamic, it is not.  No matter how creative you get in your Ifs, as long as the answer for the Draw is always the same it is in fact Static. 

If Tucker Black was still active, he would tell you that there is not connection between the static and the future.  As near as I can figure out, there is no Dynamic answer.  Dynamic is the Answer and I am it. 

The Question becomes “Why Train you Excel File to Play the Lottery History?”    All I have is this.


Then Get Out a Little Bit More and log a Few more Steps on your Fit Bit or whatever they call those Dick Tracy Watches.  A Few more Hours of Training and I might be able to take my own advice.  I just don’t think I’ll get that Wrist Watch (don’t like being watched!)

garyo1954's avatar - garyo
In response to AllenB

Great idea! I need to do it. I really miss Excel. I should just bite the bullet and get another computer for the CNC.

But I'm not getting a Dick Tracy watch either. Got a phone that stays in my pack most days, sometimes until it discharges. Coolest game on it is that one called Cal-Cu-La-Tor, but I like the pc version better. 

Guess I need to set down and do lotto work. Haven't done much but update files in the last week or so.



There are other Breeds just like Dogs.

The Point is to have a Standard Regime for Training.

I have Settled into a Training Regime that puts the "Little Guy" through a 60 Feature Obstacle Course.  Many of those Commands produce the same Number.  I keep Track of that.  I do this for each Position (in the Macro).  At the Same time I run a Filter Mixer using things that I observed when I run the Commands (Formulas) on Binge Watch.  The Mixer can be programmed to combine filters based on several factors.  One Factor is Hit Rank.  Probably the Most Potential is in the Inventory Filter, where I count how many Formulas produce the same Number and if that Number is a Winner. 

Baby Steps 1  In 1000 Draws, the Filters Combining for the most number of same output (the Mode in Excel) is a miss 93% on P1 and 96-97% on P2,P3,P4 and P5.

Still searching for the Next Step.  A Few More Runs (Few is dimensionless) and I will go On Auto.  No Watch, just a goal to go from A to B by moving something other than my fingers.



I did not start this Topic as a means of saying Look at Me and My New Golden Retriever.  I did it to ask a Question of Members as to how they go from all of the Historical Data to actually picking combinations to play.   It really does not matter what program you are running the same Question Applies.  At some point you have to act on some statistic or combinations of same to make a selection. 

How many of you Do Back Tests with Sets of Filter Values, be they fixed or derived from formulas?   Training your Program requires many of these to find successful combinations that have been winners in the past.  It is still no guarantee that they will be successful again but the Point of all that data in the first place was to find the best or what happens the most. 

So, sitting on top of all that Historical Data, how do you do your selections?  Does your program do it for you or do you Plug the Values in based on what you See in the Data

garyo1954's avatar - garyo
In response to AllenB


I do the filter backtesting some, with the caveat that it includes a time frame or or draw period when the pattern occurred.  Sometimes you discover this great pattern, only to learn it occurred 600 draws back and hasn't occurred since. Yet I'm not sure it would be considered filters. More like "key values used to determine if a draw is in the norm or an aberration."

So yeah, I tend to be more concerned with recent patterns. And unless something really stands out, I don't pay attention to what my eyes see. Lately, the number 2 has been prolific. Including pretest it has hit 15 times in the last 34 draws. That's one number occurring at a 44% return rate. We're on the downside of that trend at this point. At its peak it hit 13/26 (50%).

But a single number is not an easy thing to take advantage of unless you are aware of other trending nuances and patterns. Coupled with other trends and you might/possibly/could have something.

I haven't given up on finding a pair and building play sets around that, but 17-23 made me rethink that to some extent. Up until March, it seemed to be a pair you could count on. After (very) little success with it, it hasn't been seen in 224 draws. (Maybe I jinxed it? Maybe I'd have better luck posting sets to come out and playing sets I didn't post? Maybe it reached the end of the trend?)

There are too many variable with pairs. Consecutive pairs are a conundrum as well. Albeit they are consistent, not staying out more than 9 draws (only twice I recall), and most often hitting several times in a week.

I regret triples don't come out as often. Something like 49 total, staying out as long as 119 draws. Two pair is actually a slightly better play (67 hits/114 draws longest out).

I look at a lot of different areas in anticipation of what is where and where what is going to be. I don't get the urge to run out a play a set that I feel is "definitely without doubt hitting tonight or find new numbers." When I start a set, I plan to pay them at least a week to ten days before pulling the plug. And I'm not as likely as I once was, to go play numbers just to be playing.

Mostly I find it better to use restrictive values to eliminate possible outcomes rather than to allow them. In any case, its much easier to eliminate what we know won't happen than to pinpoint what will.


In response to garyo1954

I have been Instructing my Golden Retriever for about a Week.  I have now got her up to 228 Commands.  There are actually only about 12 Basic Formulas that I do +/- 0 to 10 making 20 iterations of a single formula.  I have a 3000 Game History File that tells me when Each Formula Hit and the Formulas that the Hit paired With in the Other Position.  The History is not impressive.  Only 1 Formula was near break-even.  One Formula hit a 5 Spot in Ohio on an Out of State training exercise. 

These Formulas When Tested did the Same thing to Each Position.  To Find better Results the Formulas must be mixed up for Each Draw.  To get to some way of Mixing, I have Run a 3000 Game history of Every Formula that hit in each Position in every Game.  I have extracted from those results to find Formulas that hit along with the formula that hit in Position 1. 

With 228 Formulas and only 39 Numbers, it is obvious that many of the formulas select the same Number.  The Spread is all over the Map.  It is not uncommon to see 50 of the formulas select the same Number.  The Baby Step holds True.  96% of the time, the number picked most by the formulas does Not Hit in any Position.  The most common Same Count is 6 or 7.  I have developed an inventory Selector to try to combine the correct formulas based on those counts.

I am prepping the Old Girl (I have a 12-year-old Golden Retriever named Sally) to do another Run.  Not Sally, she’s too Old but My New File.  Come to think of it, If I get this to be successful, I think I’ll name the Program “Sally”. 

I’ll put her through an exercise later today.  It is getting to the point where I need to do these tests at night so that I can walk away and not watch anymore because the Training Run is approaching 2 to 3 Hours to complete.

OK I’ll say it for you.  Just because it hit before does not mean that it will hit again.  To be honest, I don’t want to think of that right now.  I am locked in a battle with the History, trying to find a way to beat it.  A voice keeps telling me that if I can’t beat the History with the Old Girl, She will never pick a winner. 

And I also know that any Hit is Pure Luck.  I am just hoping that after all of the Training that the Old Girls Luck will improve enough to Hit the Big One.

It is huge and confusing but I have to see it to completion before I consider any thing else.

garyo1954's avatar - garyo
In response to AllenB

At some point I'm going to have to quit playing around and do some serious work. Been pushing too much to the back burner thinking I'd get caught up and do it, but you know when you start something, it always seems to expand to fill the time you have to do it.

I use between 23 and 40 keys depending on what's happening. But there are a lot more in the program that pares the possibilities down. Doubt its near 228. And sometimes those are too much. And sometimes I wonder why I even bothered with some of them.

For the most part, we find something that works, add it to the program, and a week later it no longer works.  Maybe we're teaching the lottery more than learning from it?

But I wouldn't stop. I've found probably 10,000 ways to be wrong already. All we need is one right way and it will all be worth it!


garyo1954's avatar - garyo

Keeps happening.......18 2s in the last 38 draws.....


11 8s in the last 32 draws!

Texas is playing this game with missing numbers!


garyo1954's avatar - garyo
In response to AllenB

Oh my! I'm going to need a bigger Labrador!
33k into writing I now see WHY you have 228 formulas! 

Baby Steps is good. At the moment I'm teetering to stand upright without face plopping. It appears what you are using is an echelon above the prediction program I wrote a couple of month's ago!

I suspect to face plop a few times before I get a working model. 

"One small step for lottomaniacs; one giant task to program." Or whatever Neil Armstrong said when won MegaMillions. 


In response to garyo1954

OMG That's Funny GARYO.  Hoaxers will tell you that he did not really win, but I believe your version.

Interesting thing about the Inventory Filter, Turns out that there are Often less than 15 Formulas to worry about because many of then produce the same Number. 

I just ran Sally through a 1000 Game test trial.  Long Way to go.  The Only thing I can say is the File Hummed.  Unfortunately it made Sally chase her Tail and I saw for the first time what it is like for a dog to switch sides for the chase, but continue to spin in the same direction.  I still can't shake the Image and Me and Old Sally are both Dizzy as Hell.


garyo1954's avatar - garyo

BOOM! Way to go!

Like I said, you're an echelon above where I am. I went ahead with the idea we IM'ed about and face-plopped my way through several variations of data files until I came up with one I almost liked late last night.



Data always looks bland. Definitely far-fetched, a bit complicated to explain, (some might even think I'm crazy if I posted an explanation). I wouldn't blame them, some days I wonder myself.

Personally I don't like all those zeros, but it is what it is; young game with not much data to work with. Sometime today I'll run a game with 6000 or so draws and see where it goes.

Got things to clean up, things to add, and formulas that aren't going to write themselves. (They aren't doing me much good on paper)

Going to get another cup of coffee, open my mind, and see what I can make of this of all this garbage before the day is over. 

Day light's burning!




In response to garyo1954

Pick a Row. Pick a Column.

Do that 5 TimeCorrect  you get the GP.

I can't quite see what to make of the chart, but no matter.  Just throw together some formulas for the Selection and Let her Rip. 

I may have to rethink naming my System after such a good dog as Sally. 

Considering Blind Squirrel



Since this has turned into a public conversation between You and Me only, Perhaps we should continue by PM or Email. 

No One Else here is interested enough to join the conversation.

garyo1954's avatar - garyo
In response to AllenB

Yes. It was a good talk too! Got some pointers and some ideas. Thanks!

Since 899 people have looked in, I'm guessing there may be only two people talking (on the forum) but plenty listening, taking notes, and watching to see where this goes.

We're stimulating brains as we speak!


In response to AllenB

Cricket Master, I believe the reason there are no other contributors here may be due to the fact that most here do not use Excel, or even know what a macro is. When I say "use" I mean using Excel's formulas, formats, and / or macros.

As for me, I am following you here, just no time to compose a post with my thoughts and findings. I have yet to dig into the last PDF you sent me. I hope to have some free time in a couple of weeks, at which time I will look at what you have presented.

Keep doing what you do!

Welcome Guest

Your last visit: Thu, Nov 26, 2020, 3:48 am

Log In

Log InCancel

Forgot your username?

Forgot your password?