What is Roulette?

If you’re reading this post the chances are you’re familiar with roulette. It’s one of the simplest and most instantly recognisable casino games.

If you’re not familiar with roulette then check out this wikipedia article for more detail.

What is the Roulette House Edge

On Roulette it actually makes no difference what you bet on, if you played an infinite number of games you would end up with the same result whatever you bet on, this is due to a house edge which is built into all casino games. For European roulette this is 2.7% meaning for every £100 you spend on roulette you will statistically lose £2.70

For more information on why this happens check out the video below, which although based on US roulette provides an excellent overview of why it doesn’t matter what you bet on.

Just remember that European roulette has a 2.7% (1/37) house edge rather than the 5.26% (2/38) house edge on the US tables.

Why do I play roulette?

Playing roulette is rarely classed as advantage gambling but there are a few exceptions.

I only play roulette as a way of taking advantage of Bingo offers and typically talk about this in the form of a ‘treble up’. By playing my initial deposit on the roulette table I can massively reduce the ‘cost of bingo bonuses’.

Why did I analyse my roulette play?

The diary only covers play with a friend of mine where I support and guide him in his Advantage Gambling. Separate to the diary I also do my own ‘personal play’.


Following a terrible run of roulette results (sometimes variance can be a killer) I decided to analyse my roulette play to see just how fair below the expected value I was and the answers surprised me….

What was my roulette analyse based on?

I’ll start by highlighting the sample size, although this covers a decent number of Roulette spins (£26,355 spent over 1,127 hands to be precise) it’s by no means a large enough sample to offer any mathematical proof.

We;ve already discussed how on roulette EV is unaffected by bet size or the type of bet, for this reason ,y analysis is based on all my combined play, across all bet sizes and bet types (double up, treble ups and other bets all the way to betting on individual numbers).

What did the analysis show?

The graph below shows a comparison of expected value versus cumulative profit over every one of the 1100 hands.


Probably the most surprising thing is actually how close I’ve ended up to EV. After spending over £26,000 on roulette I’m down £711 which is remarkably close to the £685 loss that EV would suggest.

What do I conclude from this?

In short it’s reinforced my belief in expected value. I never serious doubted it but when you’ve had a run of results like I did emotion can overtake cold hard maths. Performing this analysis helped remind me that they maths works for a reason.

It’s interesting to see that even after the terrible roulette run where I lost £1,150 of £3,650 spent on 125 spins roulette (a 31% loss) all it’s done is put my cumulative result bang on EV. I had no idea  prior to doing this how much luck I’d been having, just prior to this bad run I was over £1,000 up on the expected value.

As I stated above though this shouldn’t be taken as scientific proof, and we should be careful how much we read into it, if the analysis had been done after 1000 results it would have shown a very different story (See below)

1000 Roulette spins (£22,500) shows a very different story

I thought I’d share this as it’s a great example of both variance and how expected value works.

I should highlight that my personal play has changed as I’ve got more experienced, initially I played more double ups and treble ups before moving onto higher variance more aggressive play (which is also more profitable). On the graph you can see this clearly shown, after the first 300 offers I moved to a much more aggressive, higher variance approach.

Please comment below and let me know your thoughts on this, I’d especially appreciate feedback on whether this style of post is useful.


8 thoughts on “£26,000 on roulette, what happened and what I learnt

  1. I presume this is Roulette at online casino s not “Live Roulette” which would give different figures as it is not a RNG .

    1. Yeah it is at online casinos, but live casinos would give the same result as it’s the same RTP (assuming European Roulette)

  2. Hi Matt! Another great post. I was wondering how to use this kinds of statistics!? is it make sense to change a way of you play accordingly to results on roulette in relation to expected value. I mean when you are much up on ev just to change style of your play (change numbers etc) or if you much dawn then increase stakes is that make sense?

    1. Hi Lukasz!

      It makes absolutely no different to ev so from a purely rational perspective I shouldn’t let historical play affect future decisions. Having said that, I use it more on an emotional level, a bad run of luck can be disheartening and I find calculating the EV can remind me that it’s just a bad run and that it will recover.

      Finally, I’m just a bit of a geek and find it interesting to analyse data!

  3. This website is so great, but is there any chance you could do a step-by-step example for a 15network site? I did one and lost money and was wondering if it might just be my bad variance luck or if I misunderstood the instructions. The method used develops through the challenge posts in a way I don’t think is covered well by your initial post!

    Other than that, this is all so great, thank you so much!

  4. Great post, thanks. I’m at one of those low points (after just starting) and it is getting difficult not to take the losses emotionally. I know the maths is good but it was getting rather difficult to stay the course. It is good to be reminded with a real life scenario.

    1. Glad you found it useful. Remember if you are going through a bad patch is better to reduce variance or take a break than it is to start chasing losses!

Comments are closed.