4 rules for using data in football content


Football data is everywhere. There are free sites, such as WhoScored and FBRef, and premium offerings from Wyscout, StatsBomb and Opta. Stats are used to antagonise and inform. They’re used as argument makers and argument breakers. However, all of this data doesn’t come with a manual and there are plenty of potential pitfalls when using it. 

Stats can’t be wrong, but the use of them can be. Especially if you aren’t familiar with it all. 

Through our experience as a Content Services provider in the football industry, we feel well placed to identify four key rules for using data in football content in 2021. 

1. Context 

Context is always key. The urge can often be to list a number of positive stats for a player to show off how good they are. Readers who aren’t well versed with what is or isn’t ‘good’ aren’t to know whether 0.45 goals per 90 is impressive or not. If you’re dabbling in stats, be sure to provide that extra bit of information when possible. For example, mention a stat and then highlight where they rank for it. 

A return of 0.45 goals per 90 sees him rank sixth in the Premier League

It sounds simple, but arming readers with as much knowledge as possible without overwhelming them is far from easy. Hit the sweet spot, though, and readers can come away feeling clued up on a player they’ve never seen before. 

2. Sample size matters 

On Twitter recently, a graphic had Arsenal’s Emile Smith Rowe near the top for key passes and expected assists for players aged 23 or under, despite the fact he’d only featured in 220 minutes in the Premier League. He was outperforming a number of talents who had appeared in 1,500 or more minutes. At first glance, you don’t see that and a narrative takes root. It’s an unfair comparison weighted towards those with fewer minutes to their name, purely because of things such as purple patches and hot streaks. 

Sample size helps separate those who are overperforming from those who are at a consistent level. The latter is the desirable trait. 

3. Relevant stats 

This one is perhaps trickier, but it is just as important. When analysing players, people will often gravitate towards those which paint them in the best light. But those stats aren’t necessarily the metrics to define the player by. This seems to be more prevalent in the post-match tweets. You sometimes see accounts using shots to bulk out how well a defender has performed due to the fact their defensive numbers weren’t as high as expected. 

4. Don’t neglect totals 

Availability is an underrated trait when profiling players. This sort of overlaps with sample size. The larger the sample size, the easier it is to read into the stats they’re posting. As aforementioned, there needs to be a minimum number of minutes played to make it a fair comparison, but there’s also a requirement to flag how many games a player has started. On a per 90 basis, players may have a similar goals return, but then when totals are looked at, there’s a significant difference because one has twice as many minutes within the same time period.

If you’d like to learn about how we can help you make the most of data in your content – either through our products or services – drop us an email and we’ll be in touch.