I Don’t Create Stats Just Because I Can!

Today 05/03/2014 I engaged in a morning time twitter conversation with @SimonGleave and @CPAUWIC that was based around the question of small or big data and how do we choose the right KPI’s.

The question will have many different answers and be significantly impacted on the environment and context in which data is being or to be collected. While many grass roots sports can function and run smoothly with no formal performance measurement tools the further you progress and in many cases wish to progress the process of performance analysis can be indispensable.

One of the key challenges for us all is deciding what to measure. For me the priority here is to focus on quantifiable factors that are clearly linked to the drivers of success. Bear in mind that I don’t see quantifiable the same as statistics. While statistics of performance are among the most widely used, non-statistical measures can be just as important. For example in football positioning of units such as your midfield during transitions from attack to defence and vice versa may be the success factor in assuring adequate defensive cover or support in attack, yet this is never displayed as ‘sexy’ stat. The quantifiable measure here is against the desired game plan. So before we jump in and eagerly create our analysis process based on collecting lots of statistical data consider the discipline of performance analysis combines data mining and research with strategy, understanding of athlete behaviour, and partnering with the coaching/fitness/medical team to improve performance.

A recent blog by Keith Lyons quoted from an article by Geraint Lewis and Chris Powers:

“it is probably unsurprising that many fields are awash with poor, inefficient codes, and data-sets too extensive to be properly explored.”

We can so easily drown our focus in a maze of data and actually lose sight of the fact it’s sport with many variables. While physics is by far no strength of mine I often consider the quantum physics observer affect that refers to changes that the act of observation will make on that being observed. Does the fact an athlete/coach/official knowing they are being observed change the behaviour? Do stats infleunce athletes when decision making?

Will they make more wrong decisions in an attempt to improve his/her stats?

So for me I don’t look to create numbers just because I can I look at many of the tactical, technical, physiological and psychological factors then seek to identify the interconnections between them relevant to our gameplan in search of the ‘game-changers’. Once these are established I ensure the data is shared with the people that have the knowledge to change behaviour i.e. the coach, S&C, Phsyio etc.