Australian Football

A brief & personal view of the evolution of Australian Football statistics

As a high school kid playing for the Moe Seniors I can recall dad and mum keen to tell me after the game how many ‘touches’ I had. It was dad who took count and on reflection I have since discovered this is how stats sheets and tallies were collected back then (sometimes kept in cardboard boxes and tins)—ad hoc and by a variety of privately committed resources such as parents, club volunteers, perhaps newspapers and occasionally forward-thinking radio stations. During this era it is surprising how little interest the official bods— club officials, through to head bodies such as the VFL had in the maintenance of game day statistical records.

When I was recruited to play with Carlton I found coach Ron Barassi was indeed interested in statistics (again provided to him by volunteers); an interest that he attributed to his coaching mentors, Norm and Len Smith. While at Carlton I was pleased to hear Ron say I was a good reader of the play. I not only counted my ‘touches’ but also concentrated on where my disposals ended up.

I’d had one other brush with statistics. In my first year at Monash University completing an Economics Degree (1969), I sought a major in Economics Statistics 1. I barely scraped a pass and for years after took little interest in stats of any kind. My professor said I was hopeless at the caper and I sometimes wonder if I got back into stats just to prove him wrong?

Either way, in that study year I did get a fair picture of the principles governing the collection and interpretation of proper verifiable data, but it was not until 1993-94 that my stats interest was rekindled and I became aware of the exceptional data collection that was then available historically, mainly due to the sterling contribution initially by media mega star of the day, Harry Beitzel and his sidekick statistician Ray Young; and then Ray and his family under the aegis of the remarkable APB Sports.

Back then, the main spur for more analysus came when I was co-presenter with former tough Richmond and West Perth legend, Mal Brown, on a tipping segment on Ranald MacDonald’s ABC 774 morning show. The segment was billed ‘Beauty and the Beast’ and I referred to the computer generated tips and stats provided to me by APB Sports and Swinburne's Professor Stephen Clarke’s ‘Tin-Head’ computer tipping program that had previously featured in the Herald Sun and The Age newspapers. Meanwhile, Mal thrived on presenting his ‘strong opinions’ and having a crack at the stats and scientific approach.

I also became fascinated watching games while listening to "expert" commentators. There was a gulf between the respective viewpoints of what I witnessed — and what the data suggested — and what the "experts" said. I was resolved not to get involved in the subjective opinion game being peddled by past players from the commentary box. Instead, I wanted my commentary to be formed from an evidentiary perspective. At the time the rudimentary existing data was inadequate for the purpose.

For example, North Melbourne, coached by Denis Pagan, was winning lots of games during the mid-90s, yet the Kangaroos' statistical profile said they had the least or second least number of any team for the available stats such as kicks, marks, handballs, frees for, ruck hitouts and tackles. I deduced this is because the existing stats lacked a comprehensive and qualitative dimension. There was no accounting for how players won the ball other than via marks or frees, nor the quality of a disposal, and where it occurred on the field.

Because there was nothing of this sort available on a competition basis for all games, all teams, and all players, I ramped up the data collection—based on my observations of what was important, in terms of winning and losing—and provided the underlying analysis and commentaries. With my life and business partner, Angelika Oehme, we invested and incorporated Champion Data in 1995 as the vehicle for doing so, and in 1996 launched the REV Rankings in the Herald Sun on a weekly basis.

Initially we used Prowess Software, which was a world-class system for collecting data and linking to vision replays. Their systems were widely used by all AFL clubs, mainly in the video room and to collect the occasional coach’s ‘secret’ stats that were kept undercover and pretty much copied by all club insiders. In contrast, the philosophy of Champion Data was to spread the stats word in the public domain. Within a few years of incorporation my terminology and measures such as hard and loose ball gets, handball receives, contested and uncontested marks, various effective and non-effective kicking categories, inside 50m entries, rebound 50s, zones and CLANGERS were adopted throughout the media and coaching circles.

From an anecdotal and mostly cottage industry level, suddenly the stats were everywhere, and of a high quantitative and qualitative nature produced professionally at all stages from capture, distribution, presentation and interpretation. The stats and analysis were equal to the best stats being collected anywhere else in the world. Importantly, there was a distinct home-grown flavour, not just copied from other football codes and sports. In sum, it all looked like a “Stats Revolution” had besieged Australian Football. The speed and influence of stats surprised many and there were, and still are, sources of resistance; now there is a genuine concern that overuse is killing the golden goose.

Within a frantic period between 1996 to about 2006 the original Champion Data stats eventually morphed into the current ‘official AFL stats,’ definitions and schema. During this time there were four notable events that boosted the rapid growth and acceptance.

Midway through the 1996 Pagan became the first and most enthusiastic adopter. While North, as mentioned, had low counts of the old familiar basic stats, I was able to demonstrate to Pagan that his team was number one at winning contested footy and employing effective long kicking. He appreciated this because North’s Wayne Carey-orientated game plan proved successful. Because North won the premiership that year, other coaches who were previously doubtful or disparaging started to take notice.

Eddie McGuire, as early as 1998 had also embraced the new Champion Data stats and started presenting them on the Channel Nine Footy Show. Providing the stats and commentaries for the show was funded by Swinburne University. The ribbing and bantering between Eddie and Sam Newman was classic entertainment viewing and “hard ball gets’ mentions soon popped up in pubs, taxis, and in tea room conversations. By the start of 1998 the first internet boom was underway. The AFL, News Limited and Channel 7 joined forces and were quick to establish market leadership for our code in this booming new media space. Stats were a desired item by a new generation of computer literate consumers. After a tough tendering process competing against a consortium lead by player manager, Ricky Nixon, it was Champion Data’s record of service and industry support that won the contract.

The fourth major boost in stats coverage was the 2002 change in television broadcasting rights; previously held by Channel Seven, and thence transferred to a consortium of broadcasters including Channel Nine, Channel Ten and Fox Footy. The competitive pressures between the new broadcasters and a deep-seated desire to look better than the previous Channel Seven effort meant a significant boost in budgets and resources all round and as this happened, stats and graphics became key focus—and separator—between each of the competing networks.

By the time Angelika and I sold the business to new private owners in 2007 and I continued working under contract for the company, I noted the appetite for innovation waned and the stats product itself was becoming more routine and embedded in the official lexicon. My observation, during this period, was the game itself had changed dramatically from the mid-2000 period and onwards and was hurtling into unchartered waters compared to the fundamentals of yesteryear. Yes, interchange numbers started rocketing, but so also the handball to kick ratio plummeted, significant periods of congested play began emerging, there became an ever-increasing use of strategic ball use such as lateral, sideways, and backwards kicks, and a high percentage of ‘handball hot potatoes’ and sprays.

I count myself as not just a student but a cartographer of the game. Australian Football has such a high degree of chance ball and can look so chaotic, but in my view, that’s its main appeal. It invites the astute observer, and there are a great many of them, to draw their own pictures of what they like and what makes sense to them and why the stats perform a vital role. For a reference point, the observer can always ask; “how many touches did he have?” My job as a principle involved in the stats revolution is examining the ‘influence’ of those ‘touches.’

My current TedSport project which is integral to the new AFL Predictor App co-produced and published by Slattery Media (free on the AppStore) is founded on the principles espoused by coaching legend, Allan Jeans who famously said: “There’s no point winning the ball unless it is put to good effect.”

It’s so simple and so true. The main factor that wins a game of footy is more and better quality use of the ball compared to the opposition. Always has and always will. It’s just that now what constitutes quality use of the ball is not reflected in the type of stats I came up with more than 20 years ago. Footy has moved on in recent years and so must the metrics capable of measuring it. Once again, I am leading my company to pioneer the next generation of stats and stats thinking.

I liken this challenge to Masterchef. To anticipate and predict the best footy nowadays, you must have the best and right ingredients and also the knowledge and will to serve up the best cake there is.

The AFL Footy Predictor app is available through iTunes by clicking here .

Ted Hopkins' book, The Stats Revolution is available through the Slattery Media Group .