Every question you need to ask about running stats

Skip the next four paragraphs if you want to go straight to the questions.

There's a line repeated so often that I'm half-expecting it to turn out he never said it, but there's a Johan Cruyff quote that goes: “What do you do during 87 minutes when you do not have the ball? That is what determines whether you’re a good player or not”

A football coach of a certain persuasion might look at that and say "what do you do during the 87 minutes when you don't have the ball? Well, you're running."

Public use of running stats is exactly the mix of 'very appealing' and 'rarely scrutinised by the analytics community' that makes said community antsy. The data is being scrutinised, though, in research papers, but that's even further removed from the TV pundits usually using the figures.

It's from these papers that I hoped to put together an 'everything you need to know' explainer post, but I still don't quite feel equipped enough to write one. However, informed by that reading, what I do feel equipped enough to do is to write down all the questions you should want answers to. It's a surprisingly useful starting point.

If you have further questions or answers for the current ones, please get in touch. You can find me on Twitter @get_goalside or email getgoalside.newsletter@gmail.com

What type of running?

'Running' is actually pretty vague: does 'jogging' count as 'running'? Do you care about total distance or just the distance when players are really working for it?

There's some variation in usage of categories and how to define them, but generally the speeds that players move get broken down into: walking, jogging, running, high-intensity running, sprinting. Now, how much does each of these contribute to overall distance...?

What is 'normal'?

How much is 'a lot'? How much is 'noticeably little'? How much is 'normal game-to-game variation'?

Is 'more' better?

It seems reasonable to assume but is also worthwhile questioning. The following questions probably help you on your way to working that out.

Possession and game state: difference-makers?

Do teams 'run' more when they don't have the ball (and vice versa)? Do they 'run' more when they're trailing a game (and vice versa)? How would this affect your interpretation of running data after a match?

If you want to ask another, even tougher question: can you separate the influence of the two of them? A trend we often see is that teams that win (therefore lead in matches) are also teams which have a higher share of possession. So if teams who lead in matches/have more possession do run less (if they run less), is that a result of the possession share or the game state?

Possession: difference-maker? Part Two

The previous question was more about possession as a percentage, the share of the ball between the two teams. But how might a team or player run differently when their team has the ball and when they don't?

Crucially, if you're trying to make a point (say, on TV) about a team not working hard enough defensively, would it be worthwhile trying to get your hands on specific out-of-possession stats?

Is style of play a factor?

This might also be linked to the possession share question, given that higher amounts of possession can come from a team's simple higher quality or a marked desire to retain the ball.

Crude volume of running aside, style of play will certainly affect what the running 'looks' like. You'd imagine that teams who look to frequently counter-attack at pace would have different physical requirements to those that don't.

Player position and role

I haven't framed this heading as a question because it simply isn't one. Different positions - different roles within a team - have different physical expectations. The question here would be what the differences are, and what is the magnitude of them.

You don't tend to get people comparing the running stats of players from different positions, but even comparisons for the same one could be a little misleading (note: famous walker, Lionel Messi).

Do substitutes run more?

'Sub effects' is one of the terms that early stats Twitter coined, meaning stats of a player who often came on as a substitute and had perhaps misleadingly high figures. Maybe they came on against tired defences and racked up dribbles, shots, or goals.

Similarly, do substitutes have a different running stat profile to players who start matches? Obviously if you counted pure distance, the players on pitch longer would have higher values, but is 30 minutes of a sub appearance the same as 30 minutes as a starter?

The previous questions are ones that I could probably have a stab at actually answering based on my reading, the ones that follow are more speculative

Can you control for team quality?

'Team quality' is a vague concept, but if two teams were coached to the same level and had the same technical ability, would physical output be the deciding factor? Even if it turns out that running more doesn't strongly correlate with being more successful, might it be a factor in teams outperforming their 'ability' levels?

Does the sequence matter?

If a player jogs for 5km and sprints for 500m, how much difference does it make what order they do this in? Is there significance - in effort levels or fitness/injury risk perhaps - in them doing the sprint in regular amounts at regular intervals compared to in one or two big blocks?

How many questions is too many questions?

At the end of the day you probably want to use the running stats for something. How granular and individualised can you slice the running data without it becoming too individualised? How broad can you keep things, in the quest of a meaningful sample size to compare to, without losing important specificity?

Get Goalside is just about the only place you can get this kind of work. If you like it, you can become an official supporter for just £2 a month.

Suggested reading:

Some research papers that I particularly enjoyed reading recently

'Are current physical match performance metrics in elite soccer fit for purpose or is the adoption of an integrated approach needed?' (2018), Paul S. Bradley and Jack D. Ade

'Match-related physical performance in professional soccer: position or player specific?' (2021), Stefan Altmann, Leon Forcher, Ludwig Ruf, Adam Beavan, Timo Gross, Philipp Lussi, Alexander Woll, Sascha Härtel

'Is it worth the effort? Understanding and contextualising physical metrics in soccer' (2022), Sergio Llana, Borja Burriel, Pau Madrero, Javier Fernández

Less directly about physical metrics but provided the research for 'famous walker, Lionel Messi':
'Wide open spaces: A statistical technique for measuring space creation in soccer' (2018), Javier Fernández and Luke Bornn