Interview with Adrian Caruso (Australian Football)
Adrian Caruso (@FoxFootyAnalyst) is an analytics expert in the Australian Football League (AFL). Adrian is an analyst for Fox Footy, a subsidiary of Fox Sports that is dedicated to the AFL. In this Q&A, Adrian explains the difference between Australian football, American football, and rugby. Adrian also talks about some of the dynamics that are a part of working with analytics in Australian football. As is customary in all our Q&As, Adrian finishes the interview with advice for students and young professionals.
Can you give a brief description of your job (your responsibilities) and the path you took to get there?
My role at Fox Footy Channel is to feed analytical information to our producers and on-air talent. The Channel is taking the analysis of AFL matches to another level in order to give the viewers a more in-depth view of why things are happening on the field. We have invested in virtual technology to try and illustrate best what is happening on the field in a graphical sense in terms of patterns of play, possession locations and trends of scoring, which has never been done in the sport before.
I am involved in coming up with ideas for our weekly shows and providing content for themes in each of the matches we cover. During games I am watching multiple camera angles and I have access to detailed statistics and communicate with the producer what I am seeing happening in the game. Relevant information is put up on screen in a graphic sense, passed on to the commentators or used to create key storylines.
I previously worked for the AFL’s official stats provider (Champion Data), capturing and analyzing data as well as co-writing the AFL Prospectus. I then spent three years at an AFL club as a full-time opposition analyst (advance scout) as well as doing some list analysis, recruiting and video/stats related game analysis.
What are the basic differences between the AFL, NFL, and rugby?
Australian football is difficult to describe. It is a kicking-based game (throwing isn’t allowed) that is continuous, so players need to be skilled. There is no protective gear worn by the players, yet the sport is very physical. It is also endurance based as players are required to cover plenty of ground with no restriction on where they can go on the field. The principles of the game are basic – win the ball, get it going your way, and try to kick the ball through the big sticks when you are within range.
What difficulties arise when trying to use analytics in the AFL?
Similar to a sport like hockey, the game is predominantly continuous so therefore there are limited set plays to analyze. A lot of what happens on the field comes down to simple things like players’ instincts or the bounce of the ball, so we need to be aware of that when trying to work out what is happening.
Teams have offensive and defensive game styles but passages of play generally roll into one another, apart from stoppages of which there are roughly 90 per game.
Also, similarly to a lot of sports, analyzing the defensive aspects of the game is still very difficult.
What are some traditional AFL stats that are poor representations of performance? How have analytics improved these metrics to provide more reliable stats?
When AFL statistics were first recorded back in the 1970s they were very basic – kicks, handballs, marks, tackles and goals. These stats tell only part of the story.
For example, we can now break down each kick into a number of categories including its location, distance, effectiveness and degree of pressure the player was under. If it’s a kick inside the forward 50m there are even more layers. These breakdowns tell a far better story than just a raw kick total.
Analytics have also helped us now break the game down into three phases – turnovers, stoppages and behind kick-in. Some teams are strong in one area and weak in others. This helps clubs identify trends but also helps us at Fox Footy by giving the viewer a breakdown of the sources of scoring for each team in a game, for example.
The MLB, NBA, and NFL have all experienced some kind of debate between “old-school” and “new-school” ideas. Traditionalists favor regular statistics and scouting, whereas others argue in favor of analytics and advanced metrics. Is this kind of rift present in the AFL as well?
To some extent yes. The older generation tends to use what they are seeing on the field as well as basic indicators such as disposals (kicks and handballs), tackles and contested possessions to analyze a game whether as a commentator, coach or scout.
With the advancement in analytics in the game though, more and more attention is being paid to the detailed statistics that are available and those which correlate to success. This is partly why Fox Footy hired me as an analyst to try and identify the important advanced metrics in the game that we can implement in our coverage. It is tricky though to not over-complicate things as the heart of the game is still very simple.
I think the most effective form of analysis in the sport remains identifying something while watching a game and then using analytics to support that.
What’s your favorite part about working in sports?
It would have to be learning from former greats of the game. I was never a good player but always had a good understanding and interest in analyzing a game. Being able to work with past champions and listen to their theories and knowledge is invaluable to someone like me.
At Fox Footy we have some of the greatest players ever to play. Being able to work with them in trying to support their theories through analytics is very enjoyable.
What advice would you give to someone interested in working with sports, specifically analytics?
Have a point of difference. There are more people who are interested in sports statistics than amount of jobs available so it’s very hard to get in. Write a blog or do a research project – do something above and beyond just being interested, something that will get you noticed.
Special thanks to Adrian for his time and insights. Follow him on Twitter @FoxFootyAnalyst.