Predicting Lottery Numbers From Past Draws

So you’re ready to pick your numbers – but you want to have a go at predicting what numbers might come next based on the previous draw results.

Let’s get into how you can approach this, what does make sense and what’s a complete waste of time!

First Rule Of Prediction

… is never talk about predicting lottery numbers…

(OK, I’m joking! Literally everyone talks about how to predict numbers. But the vast majority of it is total horse dung.)

The first rule is: The more data you have the better.

So don’t just look at the last couple of weeks and think that’s going to be useful for lottery prediction. Get the full history of draw results for your game.

This can be easy if your lottery company is one of the few that provides downloadable results files on their website. Or it can be a pain in the neck if you have to page through years of results manually retyping them!

Here’s some quick tips on getting results together:-

1. Ask the lottery company for a file first (look for their contact page and email or use the contact form)
2. Try to find a results page that gives multiple results on a page rather than just one result per page
3. Try to find a results page where you can copy and paste blocks of results (or at least one complete result) to avoid typing every single number
4. Do this on a desktop computer with a big screen, then you can put the lottery website and a spreadsheet side by side (on Windows just hold the Windows key + left or right cursor) to minimize switching between them
5. Get each number into it’s own spreadsheet cell (may take some manipulation – sometimes pasting into a text file is easier and then defining the separator character – i.e. comma, space etc – when importing into a spreadsheet)
6. Make sure it’s clear which are your main balls and which are supplementary/powerball numbers
7. Ideally also include the date or draw number as it can avoid confusion later

Even if you do have to manually compile all the past draws data, it’s only painful the first time. After that you can just add on the new results as they come in.

Second Rule – What Makes Sense

When analysing your past results you need to always be thinking about what makes sense, not what appears to be happening.

Just because something happened in the past results does NOT automatically mean it is more likely to happen in future.

This is an important point to grasp. Otherwise you are doomed to end up in certain forums arguing about stuff that makes no sense in the real world…

The truly difficult thing with prediction is identifying what is and isn’t significant – because even when something is completely and utterly random weird results can and do happen.

If you flip a coin the chances of getting 10 heads in a row are small – but not impossible. If you did get 10 heads in a row it does not automatically mean that getting an 11th head is more likely. It might be, but that simply isn’t enough data to prove it.

This is the problem.

You will never have enough data to prove any theory you come up with. You can either trust me on this one, or spend some time reading up about statistical significance…

My advice here is to simply trust common sense.

By all means go looking for patterns but then ask your common sense if it makes any sense, e.g. why would ball no. 25 skip every 4 draws, and then only appear on the next Wednesday draw but only if the month has a ‘y’ in it and the moon is a waxing crescent…

I exaggerate of course, but people come up with crazy pediction rules like this all the time to explain something they saw in the past results, which is really nothing more than random results! People looking for ‘lottery systems that work‘ get lost in this kind of stuff all the time.

Third Rule – Stuff You Can Use

Spreadsheets are fun – no, really, they are! 😉

There are lots of fun functions and formulas you can use with a spreadsheet to analyze your results and look for interesting stuff. You can use Microsoft Excel if you like, but LibreOffice Calc is free and will do pretty much everything Excel can.

Some quick functions/formulas you can use:-

• COUNT() – lets you count how many times a number appears in a range, e.g. =COUNT(A2:A200,10) will tell you how many times the number 10 appears between A2 and A200.
• COUNTIF() – can do fun things like tell you how many times numbers appeared greater or less than a specific value, e.g. =COUNTIF(A2:A200,”>40″) tells you how many times a number greater han 40 appeared in that column. You can even use ‘wildcards’ so “?0” would count numbers ending in a 0 (so 10, 20, 30 etc).
• FREQUENCY() – is a powerful function that can do all of your counts in one hit. You feed it a column of results and a column of ranges to break things into and it spits out a table of counts for each range.
• MAX()/MIN() – as they suggest, give you the minimum or maximum value in a range. There’s also MAXIFS and MINIFS where you can add a criteria to this.
• Statistical functions – there are a whole ton of these. There’s everything from binomial distribution to normal cumulative distribution. But be careful as just because it sounds mathematical doesn’t mean that it gives you anything useful in terms of predicting what comes next. Always apply the ‘what makes sense’ rule!

Now tell me spreadsheets aren’t fun? 🙂

Just Remember

So, is predicting lottery numbers from past draws really possible?

Not to any level of ‘certainty’, no. But lottery draw machines are not perfectly random, because making something that is perfectly random is pretty much impossible. The lottery companies are happy that they are random enough, so you are unlikely to ever see anything dramatic in the results that means you can reliably predict anything that comes next.

And as I said above, you can never have enough data to actually prove it anyway!

But that’s not really what lottery prediction is about – not in the real world – what it’s really about is the potential to gain an ‘edge’, a small advantage over other players who can’t be bothered (and a small part of your overall lottery strategy), and that is something that’s technically possible. Just not statisitically provable.