Moneyball II

[Whispering_Jack 30,103]

I read what he's saying but I'm not that sure that its "new" in the sense that all they're doing is putting on a computer programme something that many already know.

The new Moneyball: scouting's next step

[Moonshadow 17,676]

"On the other hand, Ward and Deeble have gone back through recent years of trade periods, finding that when one particular club brought in a number of recruits at once, they totalled to just 0.4 wins."

Hmmmmm, sounds familiar......

[Deeman 1,497]

"On the other hand, Ward and Deeble have gone back through recent years of trade periods, finding that when one particular club brought in a number of recruits at once, they totalled to just 0.4 wins."

Hmmmmm, sounds familiar......

Couldn't be us.

We won 2.0 games last year.

[DeeSpencer 25,697]

It's interesting, but the clubs have to tweak what they are putting in to get relevant results, therefore somewhat making it no different to just scouting a player.

Baseball is much better suited because it's just pitcher v batter over and again. Obviously different pitches pitch differently and different fields have different dimensions but really it's a one on one battle time and time again. Football is such a 360 degree game with so many variables I don't know how useful it will be.

I like the idea of comparing draftees to current players to get a guide on them, that is interesting. And I wonder if there is an overall stat that they can find that is of more value than any other.

[rpfc 28,417]

Look, I am happy to see people trying to bring a SABR-like focus on AFL analysis, but Baseball has such defined roles that this is much easier to do.

When it comes to Wins-Above-Replacement (player), the roles are comparable ie. a closer to another closer, slugger to another slugger...

With the AFL the lines of demarcation for player roles are blurry, and to say that removing one player for another because they both play 'midfield' can bring a certain probability of wins increasing is a bit rich...

Love data, but keeping it team based is probably best for its analysis in our game.

[Moonshadow 17,676]

Couldn't be us.

We won 2.0 games last year.

That many?!!!

[Return to Glory 8,441]

Two games. Any way you look at it, the only way is up. Unless, of course, we win one. Or none. Im trying to work out whether I'd be angry or bemused or suicidal.

[rjay 24,924]

I think some of our recruiting under Neeld was based on getting a spread in age topping up the mid to late 20's and I think it was a big mistake.

The mistake to me was that it seemed like age first, ability second or not at all.

I like how we approached it this year in picking up a bunch of good kids and balancing the list with some quality early 20's (Tyson, Michie) to right the past mistakes and real quality at the older end in Vince and Cross.

A feel for the player and the game will always be a recruiters best asset.

[Jarka 767]

Look, I am happy to see people trying to bring a SABR-like focus on AFL analysis, but Baseball has such defined roles that this is much easier to do.

When it comes to Wins-Above-Replacement (player), the roles are comparable ie. a closer to another closer, slugger to another slugger...

With the AFL the lines of demarcation for player roles are blurry, and to say that removing one player for another because they both play 'midfield' can bring a certain probability of wins increasing is a bit rich...

Love data, but keeping it team based is probably best for its analysis in our game.

I was going to post something similar but you've nailed it perfectly

[Little Goffy 14,542]

Despite the massive accumulation of a great variety of data, and conversation about football using statistics more and more as dream team / supercoach become ubiquitous, Australian football remains fairly resistant to analysis or prediction by statistics. Statistics can confirm or deny, inform, point in interesting directions, but never 'answer the question'.

Three cheers for that!

[Guest]

Upon seeing the title of this thread, I was lead to believe this was a thread about the sequel to the Brad Pitt Movie 'Moneyball 2: Billy's Beane's Revenge'.

[Little Goffy 14,542]

Moneyball Z?

In which Brad Pitt plays a data analyst who has to run from screeching hordes of dream-team players afflicted with an insatiable urge to eat differentials.

"Civilization has never been tested by a regression analysis like this before"

In the end, we are all saved by a secret society known only as the 'Chi Square'.

[rpfc 28,417]

FCS Lil Goffy, spoiler alert!

[old dee 23,790]

Two games. Any way you look at it, the only way is up. Unless, of course, we win one. Or none. Im trying to work out whether I'd be angry or bemused or suicidal.

I suspect accepting

[pitchfork 181]

Once again, Australian media has failed to understand what - 'moneyball' is/was.

Simple: 'Identifying undervalued assets in order to make the most out of limited resources'.

So moneyball is not recruiting Cam Pedersen.

If statistics and analytics can help us work out who is undervalued - this will be of great benefit to the club.

Anyone with an understanding of WAR knows that it is a useful tool to compare the overall output player vs player, but says little about what they actually did. Eg. It's great if you're saying "Gary Ablett had an awesome 2013 and here is how we can see how much better he was then any other player", but limited in identifying where a team/individual's shortcomings are.

Postscript:
The AFL (and champion data)'s game rating system is a joke...The system has an in built weighting. Now here is the philosophical argument in some ways. The system weights importance of impact depending on when in the game you do what you do. i.e. if there is a blowout in the first quarter and you get ten possessions in the second quarter, but the team still loses by the same margin, that ten possies will be rated less than if you got them when the 'game was on the line'. Now this may be meaningful in 'junk time', but how is a player to know that there impact at a time in the game such as this example is for nil? Footy analytics has to be more mature than Mark Robinson's weekly column

Edited January 28, 2014 by pitchfork

[MO FINE 170]

I came across this interesting article about the NAB Challenge - 10 things we learnt from Melbourne’s NAB Challenge win against Richmond

MONEYBALL 2

Like the Godfather 2, the sequel looks better than the original. A year ago Mark Neeld picked up discards from other clubs under what some dubbed a “moneyball” recruiting strategy. Twelve months on David Rodan and Tom Gillies are gone and the other recruits (Shannon Byrnes, Cam Pedersen, Chris Dawes) are far from locks in the best 22. The Demons went headhunting again after Roos arrived but this time with a clear purpose — to give the besieged Nathan Jones a hand in the guts. Bernie Vince, Daniel Cross and Dom Tyson were acquired and all showed their worth last night, picking up 32, 26 and 16 (in a half) disposals respectively, while a relieved Jones helped himself to 28 and a rejuvenated Jack Trengove had 29. Richmond’s top ball-winner was Dustin Martin with 21 touches — nine Dees had more including defenders James Frawley and Dean Terlich.

[rpfc 28,417]

As a baseball fan, I now wish they never released that damn movie in Australia...

Sigh.

[rjay 24,924]

I came across this interesting article about the NAB Challenge - 10 things we learnt from Melbourne’s NAB Challenge win against Richmond

Quote

MONEYBALL 2

Like the Godfather 2, the sequel looks better than the original. A year ago Mark Neeld picked up discards from other clubs under what some dubbed a “moneyball” recruiting strategy. Twelve months on David Rodan and Tom Gillies are gone and the other recruits (Shannon Byrnes, Cam Pedersen, Chris Dawes) are far from locks in the best 22. The Demons went headhunting again after Roos arrived but this time with a clear purpose — to give the besieged Nathan Jones a hand in the guts. Bernie Vince, Daniel Cross and Dom Tyson were acquired and all showed their worth last night, picking up 32, 26 and 16 (in a half) disposals respectively, while a relieved Jones helped himself to 28 and a rejuvenated Jack Trengove had 29. Richmond’s top ball-winner was Dustin Martin with 21 touches — nine Dees had more including defenders James Frawley and Dean Terlich.

Nothing to do with Moneyball but everything to do with picking up players that can play. You don't need a statistical approach to tell you that Vince, Cross and Tyson are better footballers than Rodan, Gillies and Byrnes, you just need to understand the game.

[puntkick 325]

Despite the massive accumulation of a great variety of data, and conversation about football using statistics more and more as dream team / supercoach become ubiquitous, Australian football remains fairly resistant to analysis or prediction by statistics. Statistics can confirm or deny, inform, point in interesting directions, but never 'answer the question'.

Three cheers for that!

Yet again you confirm your public servant mentality

Just try watching TV or reading the papers and enjoy !

[dpositive 1,665]

Despite the massive accumulation of a great variety of data, and conversation about football using statistics more and more as dream team / supercoach become ubiquitous, Australian football remains fairly resistant to analysis or prediction by statistics. Statistics can confirm or deny, inform, point in interesting directions, but never 'answer the question'.

Three cheers for that!

Haven't heard of the Chaos Theory obviously.