Freo Draft Policy / Methods

[Arcassius]

Curious read in the Herald Sun about Isaac Smith and in particular Geelong's draft strategy - mentions Moneyball

http://www.afl.com.au/news/newsarticle/tabid/208/newsid/104974/default.aspx

"Such a recruiting practice subscribes to the Moneyball principle of drafting mature players rather than teenagers because there is more hard data on which to make judgements. (Moneyball is a book about Major League Baseball club Oakland Athletics’ revolutionary recruiting practices.)"

Got me thinking as to what methods recruiters use in assessing available talent in a draft given the competition - and more specifically speculating on the methodology that Freo currently use in the recruiting process.

I think we definitely have moved to a different model of thought than other clubs - curious as to how the landscape will change in the future as people start following our lead (or was it luck). Read somewhere on here that Mick took the Collingwood recruiters to task and asked them to think outside the box after Barlow wasn't on their list of top 100 prospects.

 

[greatgatsby]

No doubt we're leading the way in looking at some different things in prospective recruits. I think, in particular, we've led a trend back towards 'footballers' first rather than the obsession with athletes. But I think the fundamentals of building a list of young players are pretty constant.

I thought the 'Moneyball' thesis was essentially about value for money. In fooball terms the 'A's had (some) success with a team of role players. Of course, the Yankees have offered plenty of evidence that money works too!

 

[The Body]

I thought the 'Moneyball' thesis was essentially about value for money. In fooball terms the 'A's had (some) success with a team of role players. Of course, the Yankees have offered plenty of evidence that money works too!

Sort of. Basically baseball teams were overvaluing certain traits in players (i.e. the ability to look good at doing something), and undervaluing the actual ability to do it. As such pitchers who achieved good results with an unconventional style were ignored, or the former catcher who could no longer catch was put on the scrap pile despite being good enough to be in the lineup as a pure batsman. The A's (and a few other teams to a lessor degree) quantified each players performance and evaluated the effect they would have on team performance rather than just using the predominantly qualitative analysis of scouts. Ultimately they ended up taking older players because there was more data available on them, and they were also willing to take a lower pay packet to get their chance when no one else would (drafted players still need to accept a negotiated contract, rather than a standardised one).

Unfortunately such an application is likely impossible in AFL as the stats that are taken at lower leagues aren't necessarily adequate to evaluate a players performance. It's also much harder to gather the necessary data given that AFL is a continuous game whereas Baseball can be measured in discrete events (although this is not impossible, such approaches are taken in evaluating NBA teams though it would be an extremely expensive exercise to obtain the data for a sufficient sample of lower league matches). What can be taken from Moneyball is the idea of evaluating a player on his effect on the game rather than how he looks doing it, and the concept that information (performance measures) on mature players is more likely to be correlated to future output (effect on a game on AFL level) rather than just be a fluke performance.

 

[greatgatsby]

Sort of. Basically baseball teams were overvaluing certain traits in players (i.e. the ability to look good at doing something), and undervaluing the actual ability to do it. As such pitchers who achieved good results with an unconventional style were ignored, or the former catcher who could no longer catch was put on the scrap pile despite being good enough to be in the lineup as a pure batsman. The A's (and a few other teams to a lessor degree) quantified each players performance and evaluated the effect they would have on team performance rather than just using the predominantly qualitative analysis of scouts. Ultimately they ended up taking older players because there was more data available on them, and they were also willing to take a lower pay packet to get their chance when no one else would (drafted players still need to accept a negotiated contract, rather than a standardised one).

Unfortunately such an application is likely impossible in AFL as the stats that are taken at lower leagues aren't necessarily adequate to evaluate a players performance. It's also much harder to gather the necessary data given that AFL is a continuous game whereas Baseball can be measured in discrete events (although this is not impossible, such approaches are taken in evaluating NBA teams though it would be an extremely expensive exercise to obtain the data for a sufficient sample of lower league matches). What can be taken from Moneyball is the idea of evaluating a player on his effect on the game rather than how he looks doing it, and the concept that information (performance measures) on mature players is more likely to be correlated to future output (effect on a game on AFL level) rather than just be a fluke performance.

That's it!

 

[Docker Dolph]

Something else that came from Moneyball that could definitely be used in AFL recruiting is the notion of which stats are over valued and which are under valued.

In Baseball the A's theorised that some key stats like Runs Batted In (RBI's) for position players & Wins for pitchers were being over valued by other teams. They concentrated on other stats when evaluating players and it paid off.

They went away from the high profile stats and looked for "ball players with old fashioned skills".

Sounds a lot like the "footballer versus athlete" debate that is often raised. Perhaps Barlow typifies that more than anyone else in recent times.

Along those lines I reckon if the Super Draft group came along now, Sam Mitchell would go top 10 rather than 35 or whatever it was back then.

 

[DockerLand1]

Something else that came from Moneyball that could definitely be used in AFL recruiting is the notion of which stats are over valued and which are under valued.

In Baseball the A's theorised that some key stats like Runs Batted In (RBI's) for position players & Wins for pitchers were being over valued by other teams. They concentrated on other stats when evaluating players and it paid off.

They went away from the high profile stats and looked for "ball players with old fashioned skills".

Sounds a lot like the "footballer versus athlete" debate that is often raised. Perhaps Barlow typifies that more than anyone else in recent times.

Along those lines I reckon if the Super Draft group came along now, Sam Mitchell would go top 10 rather than 35 or whatever it was back then.

The draft pick that we included with the Number 1 pick in exchange for Mr Croad. Yay! Worst trade EVER!

 

[Docker Dolph]

The draft pick that we included with the Number 1 pick in exchange for Mr Croad. Yay! Worst trade EVER!

You're forgetting that we also got Luke as part of that package so, no, not worst trade ever. Not even our worst trade ever.

 

[Alfonz]

You're forgetting that we also got Luke as part of that package so, no, not worst trade ever. Not even our worst trade ever.

We could have left Croad and #1 and #4 out of it and got McPharlin separately for a lower pick. Probably still would have stuffed it and taken Polak and Sampi.

 

[prattsta]

Sounds a lot like the "footballer versus athlete" debate that is often raised. Perhaps Barlow typifies that more than anyone else in recent times.

Only because for some strange reason Barlow isn't seen as much of an athlete.

 

[rotto]

Just reading off the internet, that the Moneyball method is to determine what contribution a player to winning.

Boston Celtics famous GM and coach Red Auerbach always pay salary according to contribution to winning. Not stats or potential but contribution to winning.

So why not recruit not according to strength or weakness but contribution to winning.

Not sure how recruiting mature age players is actually a moneyball principle.

 

[bfff]

Just reading off the internet, that the Moneyball method is to determine what contribution a player to winning.

Boston Celtics famous GM and coach Red Auerbach always pay salary according to contribution to winning. Not stats or potential but contribution to winning.

So why not recruit not according to strength or weakness but contribution to winning.

Not sure how recruiting mature age players is actually a moneyball principle.

Recruit mature age players who have proven they are game-winners.

 

[theGav56]

The draft pick that we included with the Number 1 pick in exchange for Mr Croad. Yay! Worst trade EVER!

In retrospect that may be true, but as another poster said we possibly would have gone Polak and Sampi.

How do you rate Croad anyway? He was a top 5 pick from memory, outstanding athletic ability, had already played international rules, became AA CHB, and has a Premiership medallion. If you said you were going to get that from your #1 draft pick (Waterhouse, Fraser etc), most clubs would take it.

By the way I would rate the Bell and Carr trades as worse, and possibly Headland too, even though the only one that was perhaps satisfactory for us was Bell. Still gave up a truckload for him.

 

[Selggad]

Just reading off the internet, that the Moneyball method is to determine what contribution a player to winning.

Boston Celtics famous GM and coach Red Auerbach always pay salary according to contribution to winning. Not stats or potential but contribution to winning.

So why not recruit not according to strength or weakness but contribution to winning.

Not sure how recruiting mature age players is actually a moneyball principle.

Can you tell me what that means? How would it be quantified? SOunds too hard to apply. It would be possible to break a squad into various roles and types and then develop KPIs against them (which i imagine the clubs already do in some form) but contribution to winning seems a little too esoteric.

 

[Clay Davis]

Can you tell me what that means? How would it be quantified? SOunds too hard to apply. It would be possible to break a squad into various roles and types and then develop KPIs against them (which i imagine the clubs already do in some form) but contribution to winning seems a little too esoteric.

Things like score involvements (Barlow had huge numbers in this regard) is one. Effective spoils

One day, when the resources are there, serious attention will be thrown at highly detailed stats. We currently have crude positioning stats like I50s, but stats like effective passes into the corridor, running into open position, etc. All depends on the granularity - we've gone a huge way from simply assessing how many kicks a player has racked up, to things like contested possessions - but there's even more room to make more assessments.

Champion Data might already assess these figures - I'm not sure.

 

[Lach72]

Things like score involvements (Barlow had huge numbers in this regard) is one. Effective spoils

One day, when the resources are there, serious attention will be thrown at highly detailed stats. We currently have crude positioning stats like I50s, but stats like effective passes into the corridor, running into open position, etc. All depends on the granularity - we've gone a huge way from simply assessing how many kicks a player has racked up, to things like contested possessions - but there's even more room to make more assessments.

Champion Data might already assess these figures - I'm not sure.

I'll try and find out.

My father in law does work for SEN in Perth on match days and has access to champion data live streams...

 

[Selggad]

Things like score involvements (Barlow had huge numbers in this regard) is one. Effective spoils

One day, when the resources are there, serious attention will be thrown at highly detailed stats. We currently have crude positioning stats like I50s, but stats like effective passes into the corridor, running into open position, etc. All depends on the granularity - we've gone a huge way from simply assessing how many kicks a player has racked up, to things like contested possessions - but there's even more room to make more assessments.

Champion Data might already assess these figures - I'm not sure.

I agree that more detailed analysis like you describe could have a huge impact but also think the evaluation as 'contribution to winning' is poorly termed. It needs to be more around efectiveness of the players in their role. A fullback in a team that loses could play a great game but has no chance of stopping the avalanche caused by poor midfield work. His contribution to winning is zip because the team lost, but his contribution to preventing a blow out may be enormous.

There also needs to be some sort of measurements on contribution to team structures and game plan - individual efforts, which is primarily what we currently measure don't really reflect these things. I see how running to position and passes to the corridor, using your examples, might be going in that direction. Other things would be measuring positioning in set plays and how often we use set plays.

 

[LemmingMaster]

I'll try and find out.

My father in law does work for SEN in Perth on match days and has access to champion data live streams...

I think Champion Data only did Gold Coast games at the request by either GC or other AFL teams.

Not sure Champion data collects enough information on the Colts players.

 

[Arcassius]

Post draft - there seems to be a slight trend with the club looking at good disposers by foot and 'footy smarts' above other attributes.

Not necessarily players who are 'ready to play' but those that have the right mindset and good skills.

Was reading the Eagles board and the thread they have there about how disposal efficiency is overrated as a metric, and yet our focus seems to be leaning towards smart, skilled players.

http://www.bigfooty.com/forum/showthread.php?t=777892

Some people seem to think that loading up on players that can kick exceptionally will undoubtedly make our side better.

Thoughts?

 

[Clay Davis]

Post draft - there seems to be a slight trend with the club looking at good disposers by foot and 'footy smarts' above other attributes.

Not necessarily players who are 'ready to play' but those that have the right mindset and good skills.

Was reading the Eagles board and the thread they have there about how disposal efficiency is overrated as a metric, and yet our focus seems to be leaning towards smart, skilled players.

http://www.bigfooty.com/forum/showthread.php?t=777892



Thoughts?

That thread makes little sense.

How many goals are generated these days from poor disposal by the opposition? It's why Schammer and Dodd are such liabilities.

Other skills like tackling pressure and workrate can be taught if a player has the correct attitude.

But decision making and general ball use is very difficult to teach.

Comparing to West Coast's 2006 side as the benchmark of good footy seems an error. Yes, they won the premiership, but the era in which they and Sydney were dominant hardly measured up to the teams that came before and after that era.

West Coast had a far better overall team in the early 90s. Greater spread and depth of talent - quite similar to what Collingwood have at the moment.

 

[Arcassius]

Yeah I was scratching my head over their arguments. All the more amusing given their season last season where skills were an area of concern, and the difference in values highlighted further by our drafting approach.

 

[Alfonz]

WC's problem isn't not having enough elite kicks, its having too many players with below average to atrocious disposal (and decision making). You need to at least do something with the ball. Guys like Bartel or Barlow aren't elite kicks, but they are smart footballers who generally don't make mistakes with the ball in their possession.

 

[Arcassius]

Incidentally - if we're recruiting skilled players who are versatile (reading the Sheehan comments on our new recruits) are we heading towards the Freo version of 'Total Football'

 

[Scham]

WC's problem isn't not having enough elite kicks, its having too many players with below average to atrocious disposal (and decision making). You need to at least do something with the ball. Guys like Bartel or Barlow aren't elite kicks, but they are smart footballers who generally don't make mistakes with the ball in their possession.

Yep, They've got a spattering of really bad users of the ball which is enough to bring the whole game plan down.

The one that surprised me how bad he was last year was Beau Watters. His decision making and disposal coming out of the backline was attrocious! No one seems to highlight him when discussing the WC skills but he was one of the worst last year IMO.

On the other hand, we've concentrated on versatile (good height), agile, footy smart, skillful players in this years draft which makes real sense given that possession time and efficiency is gold in the modern game.

 

[Nickos66]

Post draft - there seems to be a slight trend with the club looking at good disposers by foot and 'footy smarts' above other attributes.

Not necessarily players who are 'ready to play' but those that have the right mindset and good skills.

Was reading the Eagles board and the thread they have there about how disposal efficiency is overrated as a metric, and yet our focus seems to be leaning towards smart, skilled players.

http://www.bigfooty.com/forum/showthread.php?t=777892



Thoughts?

the results really speak for themselves, don't they? One thing's for certain, it's darn well more important than what the eagles recruiters & coaches believe.

 

[Scham]

Incidentally - if we're recruiting skilled players who are versatile (reading the Sheehan comments on our new recruits) are we heading towards the Freo version of 'Total Football'

It seems that way doesn't it. Apart from the forward pocket, the ruck and the spine we're mostly filling the rest of the side with a bunch of agile 6' 1'' - 6' 2" players with elite skills.