The excellent Chris Anderson has posted a blog asking what analytics actually is. This struck me immediately as an important question, particularly given the way that the word has become so overused that it is perhaps difficult for people to know what it refers to. I don’t disagree with what Chris says but I would like to address his piece with my own views.
Later this year, it will be 25 years since I graduated from the then Sheffield City Polytechnic (now Sheffield Hallam University) with a degree in Applied Statistics. I don’t think I appreciated how unusual my skills were in the world of the late-1980s and this led to plenty of opportunities in the job market which I am very thankful of.
So, I know a little about analytics because I have been doing it in some form or other for my whole career. It just wasn’t called that at the time. Analytics is just a fancy word for data analysis regardless of the bizarre addition of communication to justify this ‘new’ word. It isn’t as if my career in data analysis didn’t involve the communication of my findings. However, there is a problem as analytics seems to generally be associated with numeric data. I don’t know where this has come from in the period since my graduation and it strikes me as being one of the major obstacles to pushing football analytics forwards.
I spent my early career in market research and later in academia. In neither industry did we treat quantitative and qualitative data as mutually exclusive. Quite the opposite in fact, one supported the other. As an example, the academic research I was involved in dealt with health variations in England and Wales. Having modelled quantitative data to find where potential health extremes appeared to exist, those areas would then be visited and qualitative data collected in order to provide more information on what was actually going on and why those extremes were being observed. Obviously I am simplifying the process massively but you get the idea.
The other thing that the word analytics suggests is that people are doing something very complex. I would argue that that is not the case in general and it also doesn’t need to be, at least at this stage. The data needs to be understood first and, crucially, the numeric data needs to inform the qualitative analysis via, for example, video which is the current analysis method of choice at both clubs and in the media. There may be a time for complexity but it isn’t now.
The sport of football is only complex if you want to analyse all 22 players on the field throughout an entire match but who is even attempting this? Will it ever be relevant to do so? The fact is that much of what happens on a football pitch is irrelevant to its outcome. That is why it is important to begin from first principles when trying to understand it. What happens on the pitch that IS important? How big a role does luck – good or bad – play? Many similar questions are there to be answered. These are not complex but I rarely see them being addressed.
I would therefore like to ask everyone to embrace qualitative AND quantitative analysis. It is very rare that quantitative or qualitative approaches can answer questions on their own which is why collaboration is so important. Using only one, whichever it is, cannot answer questions fully. Patterns can be found with either approach of course and some of these are already insightful. However, combining both approaches provides better and more answers which leads in turn to greater understanding.
To finish, here is an example, albeit heavily simplified. I discovered from numeric data that FC Twente created a goal attempt from 40% of their corners last season in comparison to a league average of around 25%. Discussing this with a colleague, he suggested that this may have been so high because Twente often brought a man out of the penalty area as if to take a short corner. What generally happens if a team does this? The defending team brings two men out of the box leaving more space for the attacking team if the corner is then hit long rather than short. With quantitative data it was possible to identify this pattern in the first place but only communication with others leading to qualitative analysis using video produced a full picture. Using either approach separately from the other would probably not have led to seeing the whole picture.
It is important to work together and not fall into the trap presented by Hollywood in both Moneyball or The Trouble with the Curve. Both of these films suggest that there is some sort of battle going on between quantitative and qualitative approaches. Buying into that will only hold football analytics back.