This post is the first of a series where we reveal a bunch of new methods to evaluate the performance of players (and teams), and to better contextualise what you see on the field each weekend. These measures will use a basic kit of statistics to dive down a little deeper, hopefully in an accessible way.
What are the Player Rate Statistics?
The first range of methods we are introducing are the Player Rate Statistics, which largely try to look at a player’s style of play on the field and their potency. Unlike Player Approximate Value (PAV), these are not the “one big all-encompassing number” style of comparative valuation; instead they’re small slices to see who a player is on the field, and how well they do what they ostensibly do.
Let’s introduce them one by one:
- SPD – Scores (Goals + Behinds) Per Disposal.
- SPD is a way of measuring the scoreboard potency of a player when they get the ball. For the opposition, the key to stopping players high on this measure is to deny them the ball in the first place. The players you’d expect have a high rate of scoring per touch do well on this measure. Players such as Josh J Kennedy, Lance Franklin, Eddie Betts and Jeremy Cameron have done well over the last five years in this measure, meaning it is somewhat predictive from year to year.
- PPD – Points Per Disposal.
- PPD is the slightly more direct version of SPD, working out how many actual points a player puts on the scoreboard every time they get the ball, which accounts for accuracy and shot position. Like SPD above, Kennedy, Franklin, Lynch, Hawkins and co. rate strongly here.
- IDPD – Impact Disposals (Goals, Behinds, Inside 50s, Rebound 50s, Clearances and Goal Assists) Per Disposal.
- Impact Disposals are any type of disposal that gets a special designation from the statisticians due to carrying extra meaning. Some disposals accrue more than one of the classifications above; eg some goals are also inside 50s. This measure takes a rough look at how “worthwhile” a player’s individual disposals are; attempting to ignore the empty carb disposals that take place in general play. A range of players can poll highly on this measure, from ruckmen to forwards to ball-winning mids. Like the two measures above, there appears to be a year to year correlation for those appearing at the top. The best players with a decent sample size here in 2017 were Paddy Ryder, Lance Franklin and Gary Rohan while the lowest were Jake Kelly and Neville Jetta
- NDIPD – Non Disposal Impact (1%ers, Bounces, Tackles, Frees For and Against) Per Disposal.
- Non Disposal Impact is the other half of the stat sheet – the involvement measured by the statisticians that does not end up as a possession (although some may lead directly to one). Expressed as a rate per disposal, it highlights players whose impact is disproportionately “off the ball”. Unlike the measures above this tends to identify genuine defenders more strongly.
- MPD – Marks Per Disposal.
- MPD is an easy measure of how much a team looks for a given player to target – most commonly lead up forwards like Nick Riewoldt and Levi Casboult. At the other end of the scale are the guys who win their own ball in order to get a touch.
- K:H – Kick to Handball Ratio.
- This measures a players’ confidence to dispose by foot or by hand; a measure to how they approach the game from week to week within their given role. For the opposition, they may look at this and potentially force players to overplay their weaker method to force a potential mistake.
- CMPM – Contested Marks Per Mark.
- CMPM looks at how often the marks a player drags in are under some pressure – usually the domain of big bodies such as Naitanui and Sandilands. Players rated highly here don’t get targeted a lot in clear space.
- MIPM – Marks Inside 50 Per Mark.
- MIPM takes a look at where a player snares their marks – a great way of looking at who is most likely to be a “stay at home” forward. This statistic tends to be split across both KPFs and goalsneaks.
- GAPIF – Goal Assists Per Inside 50.
- Not all Goal Assists are Inside 50s, but a fair proportion of them are, so this measure looks at how damaging and effective a players’ usage is when they’re the player going Inside 50 is.
- CPR – Contested Possession Rate
- This compared contested and uncontested possessions and simply looks at whether a player is more likely to get the ball in the contests or in space. At a basic level this can tell you whether a player is “inside” or “outside” with regards to stoppages and other contested ball situations.
Where are they?
Well, we linked to the 2017 table above, but here are the links to 2014, 2015, 2016 and 2018. We will be updating throughout this season, and we will try to add previous years when we get free time.
How did you come up with the Player Rate Statistics?
Over time we have been creating these measures in order for us to understand the game better, to assist what we are seeing on the field. These aren’t designed to replace watching football, which is a cool and good thing to do. They’re not even designed to replace the “eye test” and subjective opinions about footy, but merely to add some extra sources of possible insight.
The real story behind the story is that I (Cody) was in Melbourne and read this Kane Cornes brainfart, just before catching up with Figuring Footy to talk about the state of footy stats.
The problem most people have with the use of statistics in sport (such as Kane) is seemingly that they:
- don’t know what they are meant to represent;
- don’t agree with the findings; and/or
- think they ruin “how the game was played”.
We can’t solve the last one (which is bullshit by the way in our opinions, because footy is fantastic to watch right now), but the first two are easily solvable by transparent communication and modest realism about what certain information represents, and by linking the information to real-world on-field results. Our discussions with Rob, and other like minds in the field (such as Ryan Buckland, Matt Cowgill, James Coventry, Tony Corke and Daniel Hoevenaars) have also lead us to a similar conclusion, and a realisation that there are some internal measures that we use that we haven’t shared publicly. This is an attempt to publicise some of these ideas and add to the field of footy analysis.
Spoiler: HPN has been working on a number of big projects over the last 18 months, and entering this year was on the edge of burnout. We did not have most of these measures ready until early March, however we have used some versions of some measures in the past. Most of these were created in a hotel room in East Melbourne while watching late-season ice hockey.
Like with PAV, we decided to use data published by AFL Tables, perhaps the best resource to football on the internet, as the basis. Using the information above, you should be able to recreate, supplement, or improve on these measures.
Questions?
Hit us up in the comments below, via twitter at @hurlingpeople, @capitalcitycody or @arwon, or via email at hurlingpeoplenow [at] gmail [dot] com.