Can you use statistics to win the Lottery?


Now there are many ways to approach this feat. Some chose numbers at random, some chose so we can make pretty patterns on the ballot sheet, some chose to use the birthdays of family and friends, some use their lucky numbers while some might choose to use overdue or numbers which occur most often.

Like my approach to nothing else in life, I decided to approach this endeavour analytically.

I decided to try to use statistics.

Now, most would argue that this cannot be done. There’s the age-old saying that the lottery is simply a tax on persons who are bad at math.

A quick estimation of the probability of winning reflects that the odds are extremely poor:

Firstly, we operate on the assumption that all outcomes are completely random and independent. Random because they are not based on changeable values such as in horse racing or a sport where the outcome is reliant to some degree on strength etc.

The numbers also do not repeat themselves and discounting the Powerball for the time being since it only goes up to 10, we have the following outcome:

– The first number has a 1 in 36 chance of winning ie. 1 x 36

– The second number has a 1 in 35 chance of winning ie. 36 x 35

– The third number has a 1 in 34 chance of winning ie. 36 x 35 x 34

Unfortunately, the odds are pretty low.

Yet, maybe the numbers aren’t totally independent after all. Perhaps we can look for patterns, sequences or frequencies instead of taking each number in isolation.

If we use a coin; it has two sides and therefore can only have 2 possible outcomes when flipped. However, if we were to test this theory in practice, the outcomes are not as uniform as you would expect.

A quick search on the Lottery website will confirm this theory as you will be able to recognize that each individual number occurs with varying frequencies.

In that case, we can try using the Law of Large Numbers which states that the expected frequency of a particular number should closely match the observed frequency given a sufficiently large sample of draws.

We would, therefore, be able to use the following formula:

  • Expected Frequency = Probability x Number of draws
  • Observed Frequency = Number of occurrences in an actual lottery draw

Assuming we use the broadest probability of 36, this only narrows the possibilities very slightly.

Maybe the key is to look for patterns and assume that each pick is somehow related to the next one.

Since we know that the outcomes are completely independent, maybe we can try the Multiplication Rule:

P (A, B) = P (A) P (B)

But then again, as everybody knows, correlation does not automatically equal causation.

There are also websites which use metrics to predict outcomes such as Lottometrix.com.

This website uses what they refer to as ‘the geometry of chance’ formulated by Renato Gianella (2014) and it extracts patterns to come up with a range of possible outcomes. I myself have not tried it yet, so am unable to comment on its efficiency.

The big question remains, therefore, can someone win the lottery using statistics? Or even use its predictions to narrow the possibilities within an acceptable degree of certainty?

I was able to locate two instances online. The first is that of Joan D. Gunther, who purportedly won the lottery 4 times using statistics and the Statistics Professor Nicholas Kapoor, who was able to win $100,000.

On the other hand, to the average person like myself, the chances of winning are depressingly slim. As the saying goes:

the quickest way to double your money playing the lottery is to fold it in half and place it back in your pocket”.


So why do we continue to play?

An analysis of the psychology of this type of gambling blames the heuristics of marketing, a general ignorance of probability theory and the most common – addiction.

We all seem to fall under the illusion of control or our tendency to believe that our ticket will be the winning ticket.

Hindsight bias also known as the ‘knew-it-all-along’ effect allows us to believe that we are better than we truly are as predictors of outcomes.

Representativeness bias or gambler’s fallacy, which ties into the previous theory in line with the Law of Large Numbers where the gambler wrongly assumes that because the last outcome when he flipped a coin was heads, then the next one will be tails.

Nevertheless, in spite of all the reasons why I shouldn’t, I still feel determined to try.
There’s an old story as follows:

A man finds himself in dire trouble. His business has gone bust and he’s in serious financial trouble. He is so desperate that he prays to God for help.

“Oh God, please help me, I’ve lost my business and if I don’t get some money I’m going to lose my house as well. Please let me win the lottery.”

Lotto night comes and somebody else wins. The man prays again.

“Oh God, please let me win the Lotto. I’ve lost my house, my business and I’m going to lose my car as well.”

Lotto night comes and the man still has no luck.

He prays again.

“My God, why have you forsaken me? I’ve lost my business, my house, my car and my wife and children are starving. I don’t often ask for help and I have always been a good servant to you. Why won’t you just let me win the lotto this one time so I can get my life in order?”

Suddenly, there’s a blinding flash of light as the skies open and the man is confronted by the voice of God:

If you don’t see or hear from me again, I guess you would know why, and if you can help me figure it out in any way, I promise to share it with you. 😊
Please let me know what you think.



“Here’s to the crazy ones.!

The rebels.

The troublemakers.

The round pegs in small holes.

The ones who see things differently…

Because the ones who are crazy enough to think they can change the world are the ones who do!”

  • Rob Siltaren

download (1)

Like this article?

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.