What can football teach us about data analysis?
So your team had their best ever year last year, pulling in some very impressive results and winning an award in the process. But this year it has been much tougher. Ok, so last year was exceptional, but at least one European project still seems to be on track.
Does this sound familiar? Well, if you have followed the remarkable last 18 months of Leicester City football club it certainly should.
It is not all that surprising that Leicester has struggled to achieve the same fantastic results as last year, football teams who win the Premier League often struggle the next season. However, with Leicester City only a few points away from the relegation zone their decline is much more marked than anyone would expect, so why are they struggling?
Some years ago I was part of a team facing a similar issue. The management team was complaining that the organisation was being left behind by the opposition because our IT changes were not getting through to the system that ran the company fast enough. This was crucial because we were operating in a fast changing industry. The question was how we could improve?
As the development team lead I was asked along with the other team leads to help improve the process and from my perspective, the development team was performing fairly well and getting through tasks quickly. But that is only part of the process of delivering change - we had to look at the whole chain.
Analyse the problem
The comparison between Leicester City and my own team is that you have a scenario where teams of people are not delivering as you would expect. What can be done to help resolve this? Can organisations learn about how top level sports teams address their issues? What can we learn from them? First of all, we need to analyse the problem, then work out different resolutions, and finally, pick a solution.
So what do we know about the issue Leicester City are facing on the pitch? Well, the armchair analysts (me included) say opposition teams have worked out how to play against Leicester and this is exacerbated by the fact they have lost a key player in N'Golo Kanté. That might well be right, but can we prove it? And how can Leicester fix it?
These days sports teams are awash with stats. Opta, a sports statistics company, records around 1,500 "events" from every fixture and all of the top teams use analysis to help them improve. With that in mind, this excellent article from Statsbomb (link here) based on the statistics produced by Opta has done the number crunching for us.
I don’t want to repeat the Statsbomb article here, but in short, the problem is that teams tactically countering Leicester have been successful, and the impact of losing Kanté means they don’t get the ball forward fast enough. This, has limited the number of goal scoring opportunities for Mahrez and Vardy.
What is the solution?
Much like my situation where the business users were not getting the service they needed to hit their targets, the Leicester forwards were also not getting the opportunities they needed to score. Statsbomb identifies while the midfield players Drinkwater and Simpson are getting the ball forward well the other midfielders need to be more positive in their passing and giving more game time to Demarai Gray who has some good passing stats could also help.
As an aside Leicester have just signed a new midfielder, Wilfred Ndidi who hopefully can help replace Kanté and start to free up their midfield passing. The good news is there are options for Claudio Ranieri to try as a solution, even though he only has a very small amount of time left of the transfer window to bring in anyone new.
So how did I get on with my issue? Well, I suspected that we had some issues with the approval of change being delayed and a lack of testers, which led to a log jam of changes and then difficulties with rework. However, in those days I did not have the option of having a statistics tool like Opta at my fingertips, so I had to go away and do some digging around at a technical level.
I was able to find some statistics that proved that the development team delivered, but that there was a real backlog in testing with some items being with the test team for months. Ironically the business was not freeing up people to test. There was also a problem with people approving in a timely manner which was losing us vital days in the process.
Like the Leicester midfielders, our approvers needed to be more positive and like Leicester we needed some new recruits to help with the testing. I could not easily produce the stats on rework to prove what I suspected, but by clearing out the testing backlog, our changes were able to flow faster to help our business achieve its goals.
The world of analytical tools
Of course, my issue was back when football analysis was in its infancy. In May 1999, Prozone (one of the first football analysis companies) had two clients and no revenue. By August 2000, six Premier League clubs were paying customers. Today every Premier League team and many professional clubs in the UK are using analytical tools to give them greater insight into their problems, offer solutions and the ability to track improvement.
The world of IT and development has also changed a lot since then with greater adoption of Agile methodologies and DevOps meaning being able to analyse the throughput of change through IT systems is hugely important to continuously improving the process. This led us at Basis Technologies to develop DevAnalytics - a tool specific to SAP, which provides the means for insight into performance and process of change moving through your environments.
The tool provides thirty-five pre-defined metrics including:
- Release Velocity: How often are you releasing software into production?
- Average Cycle Times: How long is your development process taking?
- Quantity of Rework: How much waste is being generated in your process?
- WIP and Approval Times: How long is it really taking you to work on this and approve that?
If only I had this tool all those years ago then I could have had much more information quickly in order to find our issues, provide a view on rework and even be able to benchmark our performance against industry values. I also would have been able to continually monitor and improve team performance and deliver better results for our business, enabling them to react quickly to the changing marketplace.
Wired also have an excellent article here if you are interested in finding out more of the impact of data analytics on football.
Visit our DevAnalytics products page to find out how you can benefit from a better analytical tool.