Side Project – Betistics

Free sports stats to even the odds.

Two fed up guys who wanted to stack the chips in their favour.

Betistics was a free service that stored and analysed historical sports data mixed with live betting odds to identify where the bookies might have got it wrong.

Betistics was built by myself and a good friend Nick Cliff (a prominent Sydney Futurist in the payments space) just because we had nothing better to do, thought it would be fun, was a good way to not let our sub par coding skills deteriorate the point we couldn’t bash out a website and whilst we aren’t full on gamblers we do love a punt especially when we have more information at our disposal.

Gambling and Startups

Gambling can be a polarising subject (for good reason), but if you are in the startup space chances are you love a punt. You’re interested in starting your own business which is something society deems to be a risky endeavour and you’re going to do what you can to get an edge, make sure you are making the right decisions by whatever means possible and this is exactly what Betistics aimed to do for sports betting.

“Problem gambling is only a problem when you’re losing”

This was the tagline which we used and went on to win a fishburners pitching competition so feel that these sentiments match those of many others in this space who are just looking for an edge in everything they do.

We don’t mean to make light of what is a legitimate issue here in Australia, but a memorable pitch is a good pitch so do admit we were walking quite a fine line here.

How Betistics Worked & The Data

Betistics never told users what they should be betting on but looked at some key main bets and portrayed the data in an easy to digest way over a variety of time frames. From there, you could deduce your own insights cross referenced for what the book-keepers were paying and make your own risk assessment. In order to do this though, we needed a lot of data and here’s how we got it.

1. Sports Data

We were very reliant on sports data for Rugby, NRL and AFL but we didn’t start this project to spend our time on data entry. It turns out accurate sports data commands quite a premium and is consumed by news channels, bookies, news websites etc and whilst easily programatically available at the time it commanded a premium which is something we didn’t want to pay for for an MVP / fun technical project.

In the end, we hired Justin from Ugie in the Eastern Cape of rural South Africa via upwork. Justin loved rugby and earned his living doing odd jobs via upwork. In his spare time, he watched all of the super rugby games and was happy to take a look at NFL and AFL Data for a relatively nominal fee per weekv whilst entering them into our custom build backend. Without Justin, Betistics would not have got up and running!

Ugie, Eastern Cape, South Africa where our sports data hero lived

2. Betting Data

If there’s a set of companies who have ulterior motives with providing resources if it might mean the slightest chance of gaining more business, it’s betting companies.

We utilised William Hill’s open betting API’s that allowed us to populate all upcoming games for all sports and the associated odds for a number of different bets that we were analysing. This was the final data piece we required.

3. Sports Data Analysis

Our statistical analysis was simple looking a number of basic metrics which ourselves, and anyone else could bet on.

  • Straight up team win rate and form
    • Last 3 games
    • Season to date
    • Last Season
  • Head to head win rate against their current opponent
    • Last 3 games
    • Season to date
    • Last Season
  • How often they covered the line
    • Last 3 games
    • Season to date
    • Last Season
  • Home vs away game win rates
    • Last 3 games
    • Season to date
    • Last Season

Business Model

If people liked our data and thought they saw a gap, we were loaded with William Hill affiliate links for easy access where we got paid a commission on any money deposited by our users.

Over our first season of operation, we made a whopping $72!

The real business model…

Each week Nick and I looked at the data and if we thought there were big gaps would put faith in our system and lay some bets (obviously with William Hill). Surprisingly we managed to grow our balances from around $50 to $500 over the season sticking to our data. Huge gains we know…

What happened to Betistics?

We had a lot of fun building it, and it looked like there was some potential here, both from a business POV as well as refined a proper data driven gambling model!

At the end of the day, Nick and I built this for fun as well as some technical exercise. It was essentially a well put together joke and did soak up our spare time so after a season, made the call to focus on other things now that we had had our fun. At the end of the day, funnelling people into gambling sites to make a $ or operating full time in that industry was something that we didn’t feel matched with our own personal goals at the time.

Hope you had fun reading about this side project!

Leave a Reply