Standing in front of a roulette table you have a choice to make, red or black.

It’s possible to select a specific number, odds, evens, streets, corners and even green, but the most popular choice is betting on either red or black.

Each seemingly gives the player an equal chance.

On a standard UK roulette wheel there are 16 reds and 16 blacks, and 1 green. We can therefore instantly rule out green, as this offers just a 1 in 37 chance of landing.

Statistically, the odds of landing either red or green are exactly the same. There is a 48.6% chance of landing red, a 48.6% chance of landing black and a 0.8% chance of landing green. Mathematically, 100 spins of the which should produce 48 reds, 48 blacks, 1 green and a variance margin of 1.

What should land and what does land are two different things.

Many years ago, I tried playing the national lottery statistically. Lottery wins should be completely random of course, but being the statistician I am, I wanted to see if I could leverage an advantage.

I analysed all previous lottery draws, and found which balls came out more times than any others.

I would have expected each ball to have been drawn a relatively similar number of times. What I found was very different.

Some lottery numbers were drawn less than half the number of times others were.

How could there be so much variance when every ball has an equal chance of being drawn. This comes down to the random factor.

I then played the highest drawn six number balls. The result, 1 number drawn.

Odds and statistics can’t be concluded using just one example, so I played these numbers continuously keeping review of the highest drawn balls in case there was a major shift.

A few weeks later reading a Sunday newspaper I saw the previous nights results. The familiarity of numbers I had analysed, and researched, were staring back at me.

Hastily running for my ticket, I held it in front of me against the newspaper results. I had **5 of the 6 numbers drawn!**

My win was just under £1,000, and the unclaimed jackpot for all six numbers stood at **£3.2million**. I was one number away.

Was it **just coincidence that 5 of the 6 numbers** I had chosen, that **were drawn** that day, were all the highest drawn numbers to date?

Could this advantage leverage be achieved on a roulette table, with just a red or black to choose from?

To do this we would need data. I set out the task of watching 100 real live roulette spins, broken down across four individual sessions of 25 roulette spins each.

Even a 1% advantage would make an incredible difference.

**100 Real Roulette Spins – Analysis**

The analysis for the 100 real roulette spins was broken down in to four different segments, with four different dealers and two roulette wheels.

Watching 100 real roulette spins takes time!

The first 25 roulette spins resulted in 10 which landed as red, 14 of which landed as black and 1 landed as green.

- Based on this initial 25 spins the percentage split calculated to be
**56% black**,**40% red**and**4% green**.

The next 25 roulette spins resulted in 11 which landed as red, 14 of which landed as black, and none landed as green.

- This next set of 25 roulette spins has a percentage split of
**56% black**and**44% red**.

There seems to be a slight pattern emerging with the data so far, and a clear lead for black, but we still have 50 roulette spins to go and now swapping to a different roulette table.

The penultimate 75 spins, on a different roulette wheel and different dealer, came in identical to the last 25. There were 11 which landed as red, 14 which landed as black, and none landed as green.

- Therefore, this set of 25 roulette spins has a percentage split of
**56% black**and**44% red**.

Just as we start to see a pattern emerging, the last 25 roulette spins produced a slightly different result.

In the last 25 spins, 12 landed on red, 12 landed on black and 1 landed on green. This is consistent with the laws of odds and probabilities.

- In this very last set of 25 roulette spins they had a percentage split of
**48.6% red**,**48.6% black**, and**0.8% green**.

What was interesting in the data analysis in that red did not land more times than black in any of the four sets of 25 roulette spins we made.

**Overall, in 100 roulette spins, we saw that 54% landed on black, 44% landed on red, and 2% landed on green** **– black a clear winner!**

**Roulette Wheel Landing on Black**

As we saw in the case study, 54% of the time the ball landed on black.

In fact, if we had bet black on the first 3 sets of 25 roulette spins, we would have provided a better return than our stake – although the 4^{th} set would have seen a £1 loss.

During the 100 roulette spins the black landed a maximum of five times in a row. This happened just once, whereas black landing four times in a row happened an astonishing four times.

If we had **bet £1 on black for each of the 100 spins, we would now have £108** and would have an **£8 profit margin**.

**Roulette Wheel Landing on Red**

Although statistically we should have seen 48 lands on red, we saw just 44.

Throughout our case study we would have lost on any of the 4 sets of 25 roulette spins if we had bet on red.

Red only managed to appear consecutively a maximum of 3 times during the 100 spins, which happened 3 times, in comparison to black which appeared up to 5 times consecutively.

A £1 bet on red for each of the 100 spins would now see our roulette account balance with £88, **an overall £12 loss**.

**Roulette Wheel Landing on Green**

Staking on green is often overlooked.

Yes, the odds are against you, but as the chances of green appearing is 1 in 37, and the odds are 35-1, is it possible that green appears more times in real world randomness, that gives it further study?

In this example, no.

The first green appeared on spin 10. If we stopped here, we would have a profit margin of £25!

This wouldn’t be a fair analysis though.

The second, and only other, green did not appear until spin 97.

A £1 bet on green for each of the 100 spins and we would now have a £70 return and **a loss of £30**.

**Best Colour to Bet on Roulette – In Conclusion**

Our case study showed across a large sample of 100 roulette spins that **black was the best colour to bet on roulette**. Black appeared 54 times, red appeared 44 times and green appeared 2 times. A £1 bet across 100 roulette spins would have accounted for an £8 profit margin staking on black.

The power of randomness is though what statistics are all about!