Key Performance Indicators

[Quicky]

I read an interesting article in Inside Football today which highlighted the importance of missed tackles as a statistic. We are all pretty familiar with the total tackles statistic but missed tackles is something I have never seen data on before and seems like a pretty important statistic. We all know as a club we have developed our own Key Performance Indicators. To quote Trojan Pies:

Beat me to the punch - yep Desire Indicators. Developed by Senior Football Analyst, Marcus Wagner (also somewhat known for coming up with the idea for Trav to try the iPod with crowd noise).

The theory behind the measurement was to give the forwards a level playing field when assessing defensive influences on a game. Obviously, it is far easier for backs to rack up a score based upon the more traditional measurements like spoils and tackles so giving the forwards weighted credit for their efforts was required (apparently they b!tched a lot about it when doing game reviews).

This measurement is also used as a combined team score when reviewing each game – i.e. they total up all the DIs across the playing group and then reference it against their benchmark to determine whether they performed well as a team. I learned this all this around two years ago and given Hine’s reference to it in the interview, it seems it is still one of the most coveted player and team performance measurements down at the Westpac Centre.

It makes you wonder what other important statistics we are overlooking and what impact it has.

 

[Quicky]

For your interest.

Most effective tacklers:
1. Patrick McGinnity WCE 89.6%
2. Darren Jolly COL 86%
3. Nick Lower FRE 83.5%
4. Nic Naitanui WCE 83.3%
5. Jack Redden BRI 83.1%
6. Luke Breust HAW 83%
7. Josh Hunt GEE 82.4%
8. Sean Dempster STK 81.7%
9. Brad Ottens GEE 81.7%
10. Sam Jacobs ADE 81.1%
11. Josh Kennedy WCE 80.5%
12. Clancee Pearce FRE 80.1%
13. David Rodan PA 79.8%
14. Michael Firrito NM 79.5%
15. Ryan Crowley FRE 79.5%
16. Jordan Gysberts MEL 79.2%
17. Brad Sewell HAW 79%
18. Hamish Hartlett PA 78.9%
19. James Kelly GEE 78.7%
20. Matthew Boyd WB 78.5%

Broken Tackle Rate:
1. Lindsay Gilbee WB 57.3%
2. Nathan Djerrkura WB 51.1%
3. Luke Russell GC 50.0%
4. Justin Sherman WB 48.1%
5. Ben Jacobs PA 47.1%
6. David Swallow GC 46.9%
7. Dustin Martin RICH 46.1%
8. Stephen Hill FRE 44.7%
9. Jarrad Grant WB 43.6%
10. Dion Prestia GC 43.3%
11. Adam Goodes SYD 42.7%
12. Harley Bennell GC 42.3%
13. Shane Savage HAW 42.2%
14. Daniel Cross WB 41.6%
15. Matt White RICH 40%
16. Levi Greenwood NM 39.8%
17. Bachar Houli RICH 39.4%
18. Rory Sloane ADE 39.3%
19 Hayden Ballantyne FRE 38.8%
20. Matthew Stokes GEE 38.7%

Not surprising to see plenty of rucks feature in the best 20 tacklers.

The top three effective tackling sides are Geelong, North, Collingwood.

The bottom two are Gold Coast and Footscray.

They didn't provide the full ladder for that unfortunately.

Some pretty surprising results in there.

 

[jefffff34]

I'm working to become a scout, racking up quite a portfolio at the moment.

Can confirm tackling percentage is one of the things I look at, as well as competitiveness in marking contests, contested ball, clean hands, 2nd efforts, and most importantly, I have developed a way of rating how players kick a ball. Almost sabremetric like the last one is.

I have met with one club already (not CFC), and the above are the important things.

 

[Ed D]

It makes you wonder what other important statistics we are overlooking and what impact it has.

There would be a lot of them. Champion Data have so many different types of stats but only the basics are shown to the general public. You know whenever a coach or asistant mentions something about metres gained, pressure acts, attacking 50 tackles and all that they're going of stats that are given to them during the games.

 

[doppleganger]

The days of using disposals as a performance indicator are long gone.

The days of using disposal efficiency are over.

Tackles mean feck all.

There is an entirely new world of statistics waiting to be used, competitive advantages can be found....but unfortunately we dont get access to them, or people are unfamiliar with them and hence forth gravitate back to the tried and true.

Football clubs should be spending much more money on analysis of the games, but at the moment it is all about coaches and fitness staff.

 

[Quicky]

The days of using disposals as a performance indicator are long gone.

The days of using disposal efficiency are over.

Tackles mean feck all.

There is an entirely new world of statistics waiting to be used, competitive advantages can be found....but unfortunately we dont get access to them, or people are unfamiliar with them and hence forth gravitate back to the tried and true.

Football clubs should be spending much more money on analysis of the games, but at the moment it is all about coaches and fitness staff.

What sort of statistics?

Where can they be found?

 

[doppleganger]

What sort of statistics?

Where can they be found?

It is about interpretative statistics, people correctly developing methods in how to evaluate each isolated incident and its impact to the scoreboard.

Does a clearance with poor disposal that results in a turn over, hurt a team more than losing the clearance but ensuring a 50/50 contest further away from the team goals?

Basic counting of statistics is of limited value, and the definition behind effective is questionable at best.

All clubs have their own internal KPIs, and some statistical groups, ie TedSport, are already looking at ways of actually re-aligning how football is measured.

Yet unsurprisingly, AFL clubs seem to be shy in data mining and backing in football analysis, preferring to spend their resources on a bevy of coaches and fitness staff...but the tide will turn eventually.

 

[Quicky]

It is about interpretative statistics, people correctly developing methods in how to evaluate each isolated incident and its impact to the scoreboard.

Does a clearance with poor disposal that results in a turn over, hurt a team more than losing the clearance but ensuring a 50/50 contest further away from the team goals?

Basic counting of statistics is of limited value, and the definition behind effective is questionable at best.

All clubs have their own internal KPIs, and some statistical groups, ie TedSport, are already looking at ways of actually re-aligning how football is measured.

Yet unsurprisingly, AFL clubs seem to be shy in data mining and backing in football analysis, preferring to spend their resources on a bevy of coaches and fitness staff...but the tide will turn eventually.

This is the sort of thing I was hoping to discuss when I started this thread. It wasn't about the importance of missed tackle stats. It was about the importance of underappreciated or unknown stats.

Does a clearance with poor disposal that results in a turn over, hurt a team more than losing the clearance but ensuring a 50/50 contest further away from the team goals?

Does it? and why?

Basic counting of statistics is of limited value, and the definition behind effective is questionable at best.

This has been gone over a lot on the board and without being a statistical expert I tend to agree. But why? And how can statistics be reinvented to be more accurate in this interpretation of what is a very important facet of the game?

Yet unsurprisingly, AFL clubs seem to be shy in data mining and backing in football analysis, preferring to spend their resources on a bevy of coaches and fitness staff...but the tide will turn eventually

Is this the next area of football evolution?

 

[Knightmare]

Good thread.

I was thinking along these lines some time ago.

Champion Data have all the stats - but I understanding is that those stats are available to AFL clubs, not the public.

I'd really enjoy stats on things such as (To be recorded on afl.com for public access):
Production of direct opposition. Kicking efficiency for specifically long kicks then kicking efficiency for specifically short kicks. Meters gained through running the footy. Total tackles inside 50. Marks off opposition kicks. Spoils. Total hitouts to advantage + % of those to advantage.

I'm sure others can think of any number of addition stats.

 

[Ed D]

I could be wrong but I'm pretty sure Champion Data is available to the public as long as they pay for it (and it's certainly not worth it for the general public). I'm certain most commentary boxes also have access to Champion Data, although apart from Bruce McEvaney I don't know how many get much use out of it.

I've heard Hine and also coaches from other clubs talk about how clubs have their own rating systems, and it's all be based on using tailored stats that they think are important. From my understanding what they focus on changes as well, and I'm sure different coaches have different criteria and rating systems for their own teams. I've also heard the comment that there is a stat for everything that happens in a game, which doesn't really suprise me when SuperCoach claims to use over 50 different stats for it's own scoring system.

 

[doppleganger]

This is the sort of thing I was hoping to discuss when I started this thread. It wasn't about the importance of missed tackle stats. It was about the importance of underappreciated or unknown stats.

Does a clearance with poor disposal that results in a turn over, hurt a team more than losing the clearance but ensuring a 50/50 contest further away from the team goals?

Does it? and why?

It is an example of something that needs to be investigated....is it better for a club to potentially lose a clearance, if that will enable them to have a greater impact on the scoreboard.

Basic counting of statistics is of limited value, and the definition behind effective is questionable at best.

This has been gone over a lot on the board and without being a statistical expert I tend to agree. But why? And how can statistics be reinvented to be more accurate in this interpretation of what is a very important facet of the game?

The definition doesn't quantify any situation...was the guy under intense pressure, were they in the open, what area of the ground etc

And the actual definition was defined what 10 years ago....is it still suitable?

Yet unsurprisingly, AFL clubs seem to be shy in data mining and backing in football analysis, preferring to spend their resources on a bevy of coaches and fitness staff...but the tide will turn eventually

Is this the next area of football evolution?

Who would have guessed that a fitness bloke who have just as much input as the head coach??

Very much an underdeveloped area, competitive advantages are to be had if clubs are brave enough to adopt new football analysis.

 

[doppleganger]

Good thread.

I was thinking along these lines some time ago.

Champion Data have all the stats - but I understanding is that those stats are available to AFL clubs, not the public.

I could be wrong but I'm pretty sure Champion Data is available to the public as long as they pay for it (and it's certainly not worth it for the general public). I'm certain most commentary boxes also have access to Champion Data, although apart from Bruce McEvaney I don't know how many get much use out of it.

Tis available to all if you are willing to pay.

The different broadcasters ask, and pay for, an individual suite of stats.

I've heard Hine and also coaches from other clubs talk about how clubs have their own rating systems, and it's all be based on using tailored stats that they think are important. From my understanding what they focus on changes as well, and I'm sure different coaches have different criteria and rating systems for their own teams. I've also heard the comment that there is a stat for everything that happens in a game, which doesn't really suprise me when SuperCoach claims to use over 50 different stats for it's own scoring system.

The clubs get plenty of stats, again a tailored package from champion, but they also have specific areas of focus that they often do themselves.

 

[Max Levern]

There are definitely flaws in the supercoach scoring system. I agree there should be weighting but how can they work out how it should be weighting, its a completely subjective opinion. If a defender gets 20 uncontested possessions but gains a 100% effeciency does he help his team win more than a midfielder who gets 20 contested possessions at 50% efficiency? In my opinion no, not even close.
Its wrong to say the midfielder wasn't as effective based solely on that number, what if the midfielder had get a quick clearance or attempted a more difficult kick than in the defender who only kicked in 20m sideways under no pressure. If the midfielder did get a quick clearance and it fortunately landed to his teams advantage he is awarded a contested poss, clearance and effiecient kick yet if the same kick went forward but the opposition player outmarked his opponent why should the midfielder not recieve the same points, the resulting play has no effect on his specific play so why should he be penalised?
I also have problems with non-attacking defenders not scoring. How can a defender that stops a goal not register the same points as a defender that helps get a goal, the end result is the same yet defenders such as glass, gibson do not recieve the same points as enright because they dont get high poss numbers. Both defenders help there team win more than the attacking half-backs yet recieve twice the numbers?
A key area in the 2012 season is spread from the contest, well not just 2012 i can remember joel corey saying the same technique in 2007 "get numbers to the contest, win the hard ball then spread as hard as possible". I believe the same theory is applied to this season with great importance placed on winning contested ball which every team aims to do, even port adeliade and melbourne place great importance on this area but fall down when they gain that hard ball. That place is spread and effiecient disposal. I have said before correct effiecient disposal isn't measured correctly based on lack of a degree of difficulty but spread while hard to measure isn't measured at all! It can to a certain extent be measured in uncontested possessions but you also have to separate those poss from from the junk possessions of going sideways and backwards that have nothing to do with spread from a contest.
Also think ruckmen are way overrated in a teams winning ability. If a ruckmen win a hitout that isn't to advantage and results in another stoppage or even an opposition clearance why is he awarded points for the hitout? its serves his team no extra benefit of winning. Hitout should only be rewarded points if it is to advantage.

 

[HFF]

Forget tackles and chases all the like as KPIs.

Our main KPI focus should be on "outside ball movement and running" and decision making.

These are two areas we are lacking in. We focus too much on contested footy but opposition kill us on the outside, why? At every contest or stoppage all of our players get sucked in to the contest. Nobody is on the outside waiting to spread. It is why opposition waltz the footy away from us from inside 50 or stoppages routinely.

Seeing guys like Maxwell come from the back half to create outside run is indicative of where we are at- there's no way known we should be feeding him the ball to create. He is good at what he does but he is not a creator of play.

Our mids are great at winning their own ball but they don't seem to win enough easy ball such as teams like Hawthorn etc.

Decision making has been rubbish all season. Bombs everywhere, play-on in stupid areas under unnecessary pressure, kick-ins a joke etc.

 

[DermiesLuvChild]

It is about interpretative statistics, people correctly developing methods in how to evaluate each isolated incident and its impact to the scoreboard.

Does a clearance with poor disposal that results in a turn over, hurt a team more than losing the clearance but ensuring a 50/50 contest further away from the team goals?

Basic counting of statistics is of limited value, and the definition behind effective is questionable at best.

All clubs have their own internal KPIs, and some statistical groups, ie TedSport, are already looking at ways of actually re-aligning how football is measured.

Yet unsurprisingly, AFL clubs seem to be shy in data mining and backing in football analysis, preferring to spend their resources on a bevy of coaches and fitness staff...but the tide will turn eventually.

Maybe we can get Russell Crow (A Beautiful Mind) to develop a formula to guarantee victory
Or maybe we just get the footy, use it correctly & finish off on the scoreboard.

 

[Paul.]

Watch the movie Moneyball. It's a good movie.

With the players we have, what is the benefits we can get. Eg. Simon Buckley is an 80-90m player with the ball. That's good for a long kick. Ignore the opinions people have on him, he can be effective.

Good tacklers in the forward line give us a higher chance of a turnover, etc. This is a brilliant thread.

 

[Stifler's mom]

Most important KPI for us is to lose to Hawthorn before the finals and then go undefeated after that allowing for one draw.
Worked a treat in 1990 and 2010.

Heres hoping for the same in 2012!

 

[lozza63]

I think that the stats available to us are indeed only the tip of the iceberg. There are many key indicators which would be very interesting to know.

What dalethomas13/Max touched on in his post above re: spreading from the contest is indeed very important - to me it's all part of one overall factor which is individual effective player work rate - ie distance (GPS tracked) traveled by players where by their run they are involved directly and indirectly on contests and or providing an option or forwards covering opposition attempts to spread - to me we seem to find players losing concentration ball watching. We could break down individual performances linking how effectively players can run to offer options and or assist giving greater numbers around the ball.

Decision making - measuring how hesitation or not taking first options can affect teams.

!%ers could easily be expanded on for example - spoils, shepards, leads, smothers and chases.

The number of totally ineffective inside 50s - ie bombing the ball straight to opposition resulting in turnovers and scoring opportunities up the other end. at times players run too far or overuse with handball and they effectively cut forwards out of the play and against well drilled disciplined sides results in goals due to turnover.

Acts of indiscipline.

 

[Tbone McGraw]

Great thread.

however, to me, the biggest question in afl is drafting. How can you judge a kid at18 and how they will develop.

This is why i have no issue with pies trading their draft picks for known products. I will have no problem if we do again this year.

I haven't seen moneyball but should watch it.

 

[blaze036]

I'm working to become a scout, racking up quite a portfolio at the moment.

Can confirm tackling percentage is one of the things I look at, as well as competitiveness in marking contests, contested ball, clean hands, 2nd efforts, and most importantly, I have developed a way of rating how players kick a ball. Almost sabremetric like the last one is.

I have met with one club already (not CFC), and the above are the important things.

once u get the scouting job at another club, talk to the pies to become an under cover agent.. feed them dud info and give us the inside goss