fcbusiness speaks to Rob Esteva, Managing Director of The Stats Zone to discuss the growing influence of data and analysis in football.
FC: Sport, and in particular, football has become increasingly reliant on data. What has been the main driver of this?
RE: The large sums of money involved in the sport have led to investors and key decision makers demanding more thorough analysis in many different facets of the game.
It is not good enough anymore to rely on a single person’s (ie. the manager’s) opinion and judgement when it comes to transfers of seven, eight, or soon to be even nine figure sums for example.
Whether it be a transfer, or an English Championship play-off final to reach the cash bonanza in the Premier League, if data can be used to minimise the risks and give the team a better chance of succeeding, that has to be taken with the money at stake.
Is there a danger we’ve become too reliant on data for decision making or is that being dismissive of its full impact?
I think we’re only touching the surface of what can be done. In March this year we attended the MIT Sloan Sports Analytics conference in Boston, and when you see all the developments within various different sports in North America, it is clear that there is still plenty of scope for developments this side of the pond in football.
More and more clubs and organisations recognise the value of data – The Football Association has recently been on a recruitment drive for a Data Scientist.
One of the main risks however, is that expectations become too great just because data is being integrated in the decision-making process. Returns and benefits are often minimal or take years to pay dividends, but patience is hardly a virtue in football in particular.
That is not compatible when managers are losing their jobs at a quicker rate than ever before.
One of the biggest challenges remains the buy-in from Managers and Head Coaches, and that is incredibly difficult when they know their job may only last 11 months or so, on average.
Clubs and sports bodies are constantly searching for the 1% in additional performance. How far can we go with data collection and analysis to achieving this?
One of the biggest developments in football in recent times has been analytical research by goalkeepers when it comes to facing penalties.
Conversion rates have dropped from 75-80% historically to approximately 60% in the last few seasons in the Champions League, and a big part of that is down to goalkeepers studying data on their opponent’s actions.
Gains have been achieved in this particular facet of the game and there will be others as the game continues to evolve.
Now that is one example where the data is largely available, or at the least, the video is, so that is relatively easy for a team to embrace. However, there is plenty of scope for data collection and analysis outside of what is in the mainstream domain.
We have been working with one club to analyse their high press and how effective it is. We watch every instance of it in every game and collect bespoke data to report back to the club on how they can improve their pressing and also counter it.
To give an example of one specific data point, we look at the angles attacking players close down players in possession to give them the best chance of creating an opportunity to turn the ball over.
That may sound obvious, but intelligent pressing is not always applied and can be masked by the high workrate of a player. A slight adjustment on the angle of approach can quickly reap higher turnover rates and that demonstrates the lengths that teams can go to for that extra percent.
Given the speed of change do you think there’s a skills gap when it comes to data analysis at clubs and do you think there needs to be more investment staff as well as technology?
This is arguably the biggest threat to development. There are some great technological developments on the market – the folks at Prozone / Stats are doing some great work in their offerings to clubs – but there is a major shortage of suitably talented individuals with a numerical background.
There are literally thousands of Performance Analysts coming out of universities with degrees seeking opportunities at clubs, but that is not the case with talented data analysts. Working directly in sport is generally not on the radar of those with numerical qualifications because they know there is much greater remuneration in other sectors.
I was heartened to see Everton recently recruit an Analyst with an economic background, but until clubs and organisations start paying the market rate for educated candidates, they will struggle to attract the most talented people.
Going forward, what trends do you envisage in the world of sports data?
As knowledge and best practice is shared and data becomes more available and widespread, I ultimately see the cost of data reducing – an outcome which can only serve to benefit those outside the elite sports and teams.
The knock-on effect of that may see more top clubs bringing their collection ‘in-house’.
That may not necessarily mean they do it themselves, but I envisage many building teams or using companies that exclusively collect data on their behalf similar to what Arsenal have done with StatDNA. Gone are the days where ‘distance run’ is seen as a relevant stat. It tells us next to nothing.
The example of the project we are working on with regards to the high press data collection is one where a team has invested in trying to prove or dispel theories, and fine tuning their tactical approach in matches. That is where I see the greater gains as more clubs and organisations have access to the base level of data.
A similar trend could occur with sports federations and tournament organisers. Preserving the integrity of their sport is key and proprietary data is a big part of that. There is a fundamental conflict in buying data from companies who also provide the betting industry with the very same data, which is ultimately obtained from the one person collecting the data in the stadium or arena. The penny is dropping with some rights’ holders, but this is one area I expect to develop further in coming years.
And finally, I expect major developments to be made to enhance the safety aspect within all sports. To give a specific example, there is some great work being done with data on impacts in heavy contact sports like NFL and Rugby.
The only way some of the more physical sports will survive long term is to improve in this area to prevent major lawsuits akin to the concussion issue in NFL, so I fully expect safety improvements to be made in similar sports to improve equipment and safety using such data.
About Rob Esteva
Rob is the Founder and Managing Director of TSZ. He launched TSZ in 2015 after identifying a gap in the market between those who can analyse sport, and those who can crunch the numbers in sport. TSZ provide a wide range of data-related services to leading clubs, associations and federations and help them be a little more informed and smart about how they use the data they have and maximise the value of it.
