Fans Partnerships

Q&A: What does GPS data tell us?

GPS data captured through devices worn by AFL players was shared with broadcasters and licensees (Champion Data) on an ongoing basis this season for the first time through changes in the new CBA.

This includes identified individual (top five only) and team metrics such as distance covered and maximum and average speed. GPS data has now joined statistics, highlights and opinion in the analysis of players, teams and AFL games. But given broadcasters and media outlets have only commenced using GPS data under this new agreement since week one of the AFL finals, it remains to be seen how it will contribute to the conversation.

Geelong was criticised for its work rate in the first week of the finals after its 22 players combined for the least distance covered of the eight teams, but how strong is the correlation between distance and desire? editor Simon Legg sat down with a high performance manager from an AFL club to learn more about the application of GPS data ‘inside the four walls’.

How do clubs analyse GPS data from a team perspective?

We don’t look at global volume on a team level. We do it more on an individual level and we don’t look at it as a performance-based thing or a performance-outcome thing. It’s just how you work that day because of the position you played on that particular day. So if we have run a lot in a game, then I’ll be looking at what we need to do that week from a load perspective. We might have certain players that are doing a tagging role, so for instance, our negating midfielder or best runner goes to a player like Tom Scully, and we know that Scully won’t come off in the second half, so we’re going to put our best runner on him because we know he can run with him. We may then manage his load at training during the week depending on who he is playing on or the specific role he needs to play. So that analysis is not actually done on a team level, it’s all individual.

And from an individual perspective, how do you view that? How does it change from a negating midfielder, to a defender and so on…?

It depends on the makeup of your team, because ideally you want to manage energy as best you can. We might need our best rebounder to play a whole quarter because he’s just intercepting everything and rebounding. So therefore on an individual level, we see that he has played all of the second half, and he is a high-level running player with high speed. It’s all about trying to get them back to the most normal level you can based on the output they have provided. That half-back might have 35 touches and run 12 km, which has happened this year, but he might also run 15.5 km and not go near it as much, so everyone thinks he hasn’t played as well but he has run his backside off so no one cares. But when he runs less the week before and has a strong game, his running patterns and distance covered is overlooked. It can be misconstrued in that way, because touches don’t equal effort.

That’s a good point, because if a team loses and someone runs one or 2 km less than the week before, that’s when the stories start to come out…

You’re probably far enough across it in terms of your understanding of what GPS provides us, but if you’re getting beaten by 30 points, and you are sort of within a sniff but you’re not really going to claw it back, most teams that are losing are running more and most teams that are winning are running less. It’s very common across all team sports that if you’re losing, you’re running more. I’d love for every player in our team to move the ball so well that everyone runs 10 km because by the end of the season they have played two games less than everyone else.

Do the KPIs differ from a half-back, to a midfielder or ruck?

It’s not so much a target that we’re trying to hit, everyone has a profile that they run. If there are players that are significantly under or significantly over what they normally produce, we’ll talk to them about the reasons why. A gun midfielder may not have slept well because his son was sick, or perhaps someone’s diet has changed throughout the week. We’ll consider what we can tweak to help them be more consistent.

Also, if the player is significantly under, how did they play? One of our mids ran about about 12 km and had 40 touches in the wet and that was his best game of the year, but it was just about his lowest running game of the year. But he’ll have games where he runs 14.5 km but doesn’t get a sniff.

Obviously, we have individual profiles for their positions, but it’s not a strict KPI where they have to run a certain distance by three quarter time. We don’t fill their head with that sort of stuff during games because it is irrelevant. A player sticking their head over the ball and getting the pill when it needs to be won is more important than jogging fast enough when we didn’t have the ball and you’re in the back 50.

There’s so many different ways that GPS data can be read because when you look at the tactical role they’re playing, there’s a reason why a half-back isn’t running as much in a game because he is the plus-one in a quarter.

We need to be really careful with how we comment on the numbers and shy away from just assuming that less miles travelled means a player is being lazy. I guarantee he won’t be called lazy if he gets five intercept marks and six rebound 50s. When our midfielders play inside, their acceleration, high speed running and metres gained drops off completely as opposed to when they are on the wing or across half-back.

When it comes to the high-end speed and those running numbers, which position is more likely to record a higher end speed?

You’ve got special midfielders like Tom Scully and Marc Murphy that have got this ability to run at a high speed while playing inside, but position wise, you find that back flankers and wingers run at the highest speed because they’re not in as much traffic. There’s a difference between if the game is coming at you, or you’re running at the game.

In your role, if you’re seeing that the team hasn’t covered that much distance defensively, is that sometimes a good thing because it might mean that your field positioning is strong and you’re repelling a lot?

Absolutely. If we have a really poor set up, what will happen is we do a lot of high speed running and we’re gassed really quickly because all we’re doing is chasing. If the opposition isn’t scoring off us doing not a lot of work, that’s great because it means we’re well positioned and we’re conserving energy.

Talk to me about how you calculate training loads based on an individual’s game, like with the earlier example of a game where they have hardly had a break…

We have a lot of different methods in place like physical screening and longitudinal data across the season and the pre-season — we know what guys can tolerate. It doesn’t then mean that we will definitely pull someone back at training the next week, because if he doesn’t need it then he doesn’t need it. But if he is running around with Andrew Gaff the week before and has run 15.5 km then a physical toll starts to accumulate and we’ll act accordingly. We have databases that are strong enough to let us know that if someone’s base line measure is 16 km and he runs four games in a row at that level then I’m not worried, whereas if a younger guy who isn’t at that level does it, you would have alarm bells ringing. Everyone is different and that will determine what we do at training.

The crux of it is that GPS for us is more of a load monitoring tool than a performance tool. I’m not going to make any decision on a game based on the GPS numbers on game day, at all. My job is to make sure that if a player is required to play three quarters straight, he can do that.

Thanks for your time.

No worries!