Opinion "Help me out where I need faith!" - The Statistical Data Thread

Do you believe?


  • Total voters
    67

Remove this Banner Ad

"I looked, and there before me was a white horse! Its rider held a bow, and he was given a crown, and he rode out as a conqueror bent on conquest." - Revelation 6:2

Today there was a knock at the door, and finally, FINALLY I got my hands on the 2017 AFL Prospectus. Naturally, straight to the Port Adelaide section. The opinion of Champion Data, based purely on the stats, is that we believed that we were a contender in 2014 - and justifiably so because we were according to the stats - and so recruited accordingly. But it is of their opinion that getting Dixon was a mistake because 'gorilla forwards are on the way out' (sounds familiar) and the fact that we gave up so much to get both he and Ryder meant that our list profile suffered when we should have been continuing to build through the draft.

The question Champion Data pose throughout their write up is a simple one: "Can these players get back to the level they were at in 2013/14 where their transition was elite? Or did they simply overachieve and 2015/16 is their true level?" Now, they don't really have the space to do a really in depth analysis of the data they are presented - but I do. So...let's begin. The first post will be about our list profile. All stats are per AFL 2017 Prospectus.

List Profile (Percentage better/worse than players the same age)

Elite Players - Robbie Gray (+53%), Justin Westhoff (+32%), Chad Wingard (+28%)

What will surprise no one is that both Chad Wingard and Robbie Gray are ranked Elite for 2017. Nothing needs to be said about these guys. They will perform week in, week out.

What will surprise everyone is that Justin Westhoff is ranked as an Elite KPF for 2017. Ranked elite for disposals and tackles and above average for marks, inside 50s and score assists. He only spent 74% of his time in the forward line - consider this to be the bench mark of key forwards in the future. Since 2010, he has averaged 16 disposals, 6 marks and 1 goal per game - the only other two to do this in the competition are Nick Reiwoldt and Tom Lynch from Adelaide.

As I've said before: 'gorilla forwards are on the way out'. That's why we got rid of Butcher and drafted Todd Marshall - who will be a better version of Westhoff. Westhoff's ability to fill multiple roles - second tall, third tall, crumber, leading forward - in the forward line allows us to play a smaller, more nippier set up rotating around Dixon.

These guys will stay elite in 2017.

Above Average Players - Jasper Pittard (+31%), Ollie Wines (+25%), Jack Hombsch (+24%), Nathan Krakouer (+ 20%), Matthew Broadbent (+20%), Travis Boak (+13%), Hamish Hartlett (-13%), Patrick Ryder (-13%)

These are the guys we need to step up and go to the next level if we are to be a contender this year. And honestly, it's not really hard for them to do so. If Pittard keeps up his form from last season, he'll join Gray, Westhoff and Wingard as being elite. Same with Wines - there were only three categories he didn't rate as elite or above average last season - marks, uncontested possessions and kick rating.

Boak fell down quite a bit last year - his pressure was elite, but his tackles, contested and uncontested possessions - and therefore disposals - dropped to average levels. But this will improve with the return of Ryder, who is listed as above average in both ruck win and hitout to advantage percentage - the new rule changes limiting third man up will enormously benefit him as an athletic ruck. These two are linked together like Batman and Robin.

Hombsch needs to improve their intercept marking this year. It's fine playing the holistic defensive structure that Bassett wants us to play, with repeat spoil attempts, but unless you're adding the other side of defending through intercept marking to generate quick rebound instead of a stoppage, any side is going to struggle. Again, I believe this is due to the lack of pressure from the midfield, which needs to be addressed if we want to get back where we belong.

Broadbent - did someone say he's not very good? He ranks above average for meters gained, disposals and score involvements. A prolific ball winner, elite for contested possessions and score launches per game. If you don't think we'll be better putting him at half back instead of having him try to fill holes on the wing, you're crazy.

Nathan Krakouer improved his one on one defending this year, moving from 36% of contests lost to 22%. Expect that to drop further as he gets another pre-season under his belt. Above average for contested possessions, intercept marks, intercept possessions and spoils. He is what you might think Byrne-Jones is, but actually isn't, and it's why he was elevated to the senior list.

Hartlett couldn't be any worse than he was last year. Well, he could, but that would make him a scrub, and he's not a scrub. Maybe he was mentally affected by his hamstring injury - who knows? But a new dual role (defender/midfield) and added responsibility on the field instead of off it (making him vice captain was a mistake) could be the making of him.

Average Players - Aaron Young (+10%), Jackson Trengove (+8%), Jarman Impey (+8%), Tom Jonas (+5%), Jared Polec (+1%), Karl Amon (-1%), Jake Neade (-6%), Brendon Ah Chee (-6%), Charlie Dixon (-7%), Brad Ebert (-15%), Sam Gray (-16%), Matthew White (-16%)

If you really want to know where the problem with our season was last year, it's in this list - and for various reasons. Aaron Young should improve, but one of the slights is his kicking, though his ball winning was above average. With a better build and more focused commitment, expect him to star in a forward pocket where his kicking actually improves.

Jackson Trengove suffered this year by playing as a primary ruck instead of a secondary one. He ranked elite for disposals, clearances and centre clearances, and above average for tackles. He is a competitive beast and will be a blessing swapping with Westhoff and Ryder through defence, ruck and forward. Expect him to improve massively.

Ditto Jarman Impey, who slot into the defensive forward role and ranked elite for score assists. His pressure acts were also elite for his position, but the one thing he needs to improve on is his goal scoring, since he provides a good target up forward.

Tom Jonas is an interesting one. Apparently he's above average as a general defender and above average for intercept marking (which is his primary role in the side) but his kicking and decision making lets him down because he was poor for generating plays from rebound. But that's why we have guys like Hartlett, Pittard and Broadbent in defence. This is why I have him in the back pocket over someone like Byrne-Jones, who we will look at later.

Jared Polec is a lynchpin of what makes us tick as a side. 3rd for meters gained at the club, sixth in the competition for intercept marking (3.5 per game), his performance or lack there of in 2016 was directly responsible for our lack of ball movement. He wasn't dropped just because he wasn't performing, but because when he doesn't perform his role, the entire side suffers. I'm expecting him to bounce back in a massive way in 2017.

If Karl Amon is having a good a pre-season as people suggest, it bodes well for the performance of the senior side. He is steadily improving every year and will push a number of players to perform to their maximum potential. Will end up being on the emergency list quite a bit if he isn't playing.

Jake Neade. If you don't rate him, you're clueless about football. Elite in forward 50 tackles and pressure acts, elite kick and the most post-clearance tackles at the club. His issue isn't his want or skill, it's his ability to win the ball. But Monfries may edge him out of the other defensive forward role (Impey has the other spot locked down) purely due to this fact. Along with Amon, he's good pressure on the first 22.

Brendon Ah Chee should be looking at Powell-Pepper, Atley and Drew and saying to himself 'There's no way I'm letting these guys jump the queue ahead of me'. In the SANFL, he ranks elite for disposals, contested possessions, clearances and score involvements. The issue is that his defensive side isn't what you need from an inside mid - but as a forward he still rated elite in the AFL for score assists and above average for tackles when playing forward. More good depth to the squad.

And now we come to the gorilla in the room. Charlie Dixon. For someone who was targeted 34% of the time by our midfield, he only managed to win the ball 27% of the time. So while 1 in 3 kicks were heading his way, he only won around 3 in 10. "That's because he had 4 guys hanging off of him!" Well, he was involved in 88 one on one contests but only won 30 - which was down 8% on 2015. So not only was he a poor target up forward, even when he got himself into a good position he struggled. However, I believe this was due to two things. 1) Dixon needed time to adjust to being the primary target of a forward line 2) Dixon needed to get his fitness base up to cope with our gameplan.

Oh, and get Tredrea to work with him on his positioning, FFS.

I'm not worried in the slightest about Brad Ebert, since the very decline in output that he suffered in 2016 can be directly attributed to the loss of a decent ruck in Ryder. Expect his clearance and ball winning numbers to improve this season and to push into the above average category.

Sam Gray is more depth. He was number one for chain involvements, playing more in an outside role. However, with Powell-Pepper most likely coming into the side, you will probably find that guys like Ebert and Boak will fill these roles.

Poor Matthew White had a horrible season in terms of luck, but his contribution as a half forward/wing cannot be underestimated. If we aren't playing Monfries in the side as a forward target, playing White would be almost a necessity, since we score 64% of the time when he is the sole target inside 50.

Below Average - Darcy Byrne-Jones (-5%), Logan Austin (-8%), Matthew Lobbe (-17%), Tom Clurey (-25%), Angus Monfries (-25%)

And now we reach end game. For all the people who waxed lyrical about Darcy Byrne-Jones, consider this - he is a defensive Jake Neade for pressure acts, and that's it. Elite for pressure in defensive 50, but below average or average in all other key areas. That's not what you want from a player in line for a pivotal role in defence. Can he improve and make that spot his own? Sure. But a lay down lock in everyone's best 22? Only if you're seduced by actual output over expected output. Or it could be that the way Port Adelaide defends is going to make the statistics they measure irrelevant.

Logan Austin is a case in point. Elite in spoils, but poor or below average in every other metric...yet managed to keep Tom Lynch (GC), Jesse Hogan, Drew Petrie, Ty Vickery and Travis Cloke goalless. Don't be surprised if he does the old Hinkley special and starts back in the SANFL to work on his intercept marking.

If Matthew Lobbe comes anywhere near the side in 2017, you know one of two things. He's either had a massive reversal of form or we are tanking to improve the free agency compensation we'll get for losing Pittard and Trengove when they leave to find a club that actually has a clue. Below average to poor in every single metric you can measure for a ruck. Needs to play permanent forward in the SANFL in the Dixon role (with Marshall playing the Westhoff role) to force him to learn accountability and contested marking.

Lack of offensive firepower is what is holding Tom Clurey back. Great at stopping, but poor in every offensive metric. Now, this could be a similar situation to Logan Austin, and it's just that he is playing a dedicated role and letting guys like Pittard do the heavy lifting of attack, but if he doesn't learn to 'join in', as Hinkley says, and push up the ground instead of just worrying about his man and that's it, we're never going to get anywhere. But that comes from experience, so give him time.

It's no secret that I'm a fan of what Angus Monfries brings to the side, and whether he spends his time in the SANFL or the AFL he will bring that defensive forward pressure on the opposition's key playmaker while kicking a goal a game. If the year off has done him good, he'll be an asset to the AFL side. If it hasn't, he'll still be serviceable depth.

Poor - Jimmy Toumpas (-24%)

Remember how there was an outcry about Dimitri being dropped? Well, being below average in every single area except meters gained and kick rating might have had something to do with it. It's good that he's having a good pre-season, because he needs one. That being said...he still improved on his rating that he had at Melbourne, so maybe there's hope.

The Unknowns

Of the rest, Riley Bonner will be a star, Jesse Palmer needs to work on his consistency at SANFL level which is why he didn't get a call up sooner, Dougal Howard will probably end up a victim of the move away from big key forwards/defenders, Billy Frampton is steadily improving every year but still has a bit more improvement to go before he earns a call up and Aidyn Johnson just needs to get on the park and find the pill. I'd expect Bonner and Palmer to get games at some point this season.

The Draftees

If you want to know what we drafted Sam Powell-Pepper, it's this: He is elite in delivery, goals, intercept possessions and score assists. The fact that Champion Data didn't even have him in their top 25 isn't a slight on him as a player - it just showed the strength of the draft pool.

Todd Marshall went about as good as you can expect a forward who hadn't played a lot of footy to go - elite tackler, above average at hitting the scoreboard, contested marks and forward 50 ground-ball gets. He will be a better version of Westhoff - which is why we drafted him.

Joe Atley can play inside or a more balanced inside/outside role and his performance doesn't dip. In TAC Cup, 55% of his possessions were contested, whereas at the U18 Championships, it was only 40% - yet he still racked up 22 and 20 disposals per game in each competition.

But Willem Drew is the standout for what we are looking for. missionpossible rated him as the player Port Adelaide needed, and when you consider that he ranked elite in clearances, contested possessions, tackles and score involvements, and above average for disposals...it wouldn't surprise me if all three of Powell-Pepper, Atley and Drew get games ahead of the likes of Sam Gray, Jesse Palmer and Brendon Ah Chee.

The Rookies

Pretty much every rookie on our list is rated by Champion Data except Will Snelling, who wins the ball a lot but has poor kicking. The only one worth looking at is Brett Eddy for 2017 - whether he plays or not is entirely determined by the ability of our midfield to find a target other than Charlie Dixon. Which is why it's good that Dixon hasn't been involved in training because quite frankly...we are our own worst enemy when it comes to the forward line.

But that's for the next post.
 
Dixon is far from a 'gorilla forward'. Stopped reading there. If there's anything else of value in that post, somebody please quote it for me.

Champion Data referred to him as such, not me.

But the way we were using him in 2016, he most certainly was, which is why he was a failure. He was our forward target 33.6% of the time - the third highest in the league. When we were actually successful is when he didn't try to monster defenders and played the role we actually want him to play - the target up forward that brings everyone else into the play like he did against Collingwood.

People like yourself have a cry about how he needs help when he needs to help himself and by extension the team by actually drawing players away from the ball if he's getting double/triple teamed.
 
Last edited:
From the Alberton Facebook page regarding Hinkley's comments at the AGM. "Hinkley said that any player not playing to the maximum of their ability will certainly be replaced. The draftees and rookies are of high quality."
Let's hope this gets the best out of every player. Please no golden tickets this year Kenny.
 

Log in to remove this ad.

I'm not sure how the statistics work, are they averaged over a four year period? I find your positive analyses interesting although I believe there's an element to reality that statics will never be able to measure. As for Dixon not being a gorilla forward, then what is he? I remember reading an article at the beginning of the season showing he is both the tallest and heaviest actual forward...
 
Introducing The Arc’s system for rating AFL teams
December 29, 2016
We’ve started rating the strength of each AFL team and using those ratings to predict the results of upcoming games and simulate the remainder of the season. If you’re anything like me, you’ll have a few questions about all that, namely ‘why?’, ‘how?’, and ‘what is the name of the Hungarian physicist who created this ratings system?’. This post answers those questions......

......

WHICH HUNGARIAN PHYSICIST IS YOUR RATINGS SYSTEM NAMED AFTER?
Good question! Our ratings system is based on the Elo system, which was devised by old mate Arpad Elo to rate chess players. Thanks, Arpad. The Elo system is widely used to rate sports teams, notably by FiveThirtyEight for baseball, basketball and American football team ratings.

The use of an Elo ratings system in AFL isn’t novel. Matter of Stats has several different ratings systems, many of which are Elo variants of one kind or another; Figuring Footy has a very interesting Elo-based system; and other sites like PlusSixOne Blog and FootyMaths have created their own Elo systems. Someone created an AFL Elo system for their PhD thesis. We’re not treading new ground here. Our system takes bits of inspiration from all the other systems that already exist.......
https://thearcfooty.com/2016/12/29/introducing-the-arcs-ratings-system/


The Arc reckons our 2015 team was one of the best to miss a final
https://thearcfooty.com/2017/02/02/best-teams-missed-finals/

bestnonfinalists.png


https://thearcfooty.com/2017/02/02/best-teams-missed-finals/

But 2016 wasn't so good

https://thearcfooty.com/2017/01/28/a-complete-history-of-the-afl/
upload_2017-2-4_18-56-33.png

https://thearcfooty.com/2017/01/28/a-complete-history-of-the-afl/
 

Attachments

  • upload_2017-2-4_18-54-43.png
    upload_2017-2-4_18-54-43.png
    51.9 KB · Views: 12
I'm not sure how the statistics work, are they averaged over a four year period? I find your positive analyses interesting although I believe there's an element to reality that statics will never be able to measure. As for Dixon not being a gorilla forward, then what is he? I remember reading an article at the beginning of the season showing he is both the tallest and heaviest actual forward...

The ratings are averaged over two years.

Dixon's primary role in the side is akin to the target man in association football. He is our long forward target when he's in the forward line, designed to pull defenders back into their back 50 so our midfield can push up. That's the 'gorilla forward' part. When they do that, Dixon is supposed to drag the defenders who play on him up the field and create space in behind - which is the 'lead up forward' part that Champion Data didn't make mention of because apart from the Collingwood game, it never happened - primarily because our midfield was so slow and stagnant, playing 'Jabba the Hutt' (ugly) rather than 'Princess Leia' (sexy) football (credit to Champion Data for that) so it just became the 'boot it up the field to Dixon' show.

Champion Data looks at the stats and says 'Hey, Port went to Dixon 1 in 3 times, they must be looking to play that way!' but the reality is that we played that way because we didn't have the ability to move the ball out of congestion. We looked our best when our midfield was scoring the majority of goals because the days of one lone key forward target doing the business are long gone.
 
Introducing The Arc’s system for rating AFL teams
December 29, 2016
We’ve started rating the strength of each AFL team and using those ratings to predict the results of upcoming games and simulate the remainder of the season. If you’re anything like me, you’ll have a few questions about all that, namely ‘why?’, ‘how?’, and ‘what is the name of the Hungarian physicist who created this ratings system?’. This post answers those questions......

......

WHICH HUNGARIAN PHYSICIST IS YOUR RATINGS SYSTEM NAMED AFTER?
Good question! Our ratings system is based on the Elo system, which was devised by old mate Arpad Elo to rate chess players. Thanks, Arpad. The Elo system is widely used to rate sports teams, notably by FiveThirtyEight for baseball, basketball and American football team ratings.

The use of an Elo ratings system in AFL isn’t novel. Matter of Stats has several different ratings systems, many of which are Elo variants of one kind or another; Figuring Footy has a very interesting Elo-based system; and other sites like PlusSixOne Blog and FootyMaths have created their own Elo systems. Someone created an AFL Elo system for their PhD thesis. We’re not treading new ground here. Our system takes bits of inspiration from all the other systems that already exist.......
https://thearcfooty.com/2016/12/29/introducing-the-arcs-ratings-system/


The Arc reckons our 2015 team was one of the best to miss a final
https://thearcfooty.com/2017/02/02/best-teams-missed-finals/

bestnonfinalists.png


https://thearcfooty.com/2017/02/02/best-teams-missed-finals/

But 2016 wasn't so good

https://thearcfooty.com/2017/01/28/a-complete-history-of-the-afl/
View attachment 333267

https://thearcfooty.com/2017/01/28/a-complete-history-of-the-afl/

Any system that ranks Adelaide (1640) over Geelong (1635) is fundamentally flawed. Also, how the **** did we drop 5 points against the Dogs, who were top four at the time and flying? They must really over rate the home field advantage we have in their model, or underrate the Dogs as a side.
 
I understand this is a statistics thread so my opinion is literally the opposite of what this is about but I have an issue with some of the analysis.

Does anyone think Westhoff is an elite KPF, even within the scope of the new type of KPF?

And now we reach end game. For all the people who waxed lyrical about Darcy Byrne-Jones, consider this - he is a defensive Jake Neade for pressure acts, and that's it. Elite for pressure in defensive 50, but below average or average in all other key areas. That's not what you want from a player in line for a pivotal role in defence. Can he improve and make that spot his own? Sure. But a lay down lock in everyone's best 22? Only if you're seduced by actual output over expected output. Or it could be that the way Port Adelaide defends is going to make the statistics they measure irrelevant.

“He’s played eight of the best first games you’ll ever see” - Ken Hinkley
“He’s a signficicant player in our team right now" - Ken Hinkley
"He gets big match-ups, we look for the most damaging small forwards and say ‘Darc, your turn’, and he’s been able to get every one of them" - Ken Hinkley
“I don’t think anyone would say ‘Darce’ hasn’t earned his place in our team this year, but that comes from two years of really strong, hard work in the SANFL and during pre-season." - Ken Hinkley

I'd say our Head Coach disagrees with you and Champion Data.

Poor - Jimmy Toumpas (-24%)

Remember how there was an outcry about Dimitri being dropped? Well, being below average in every single area except meters gained and kick rating might have had something to do with it. It's good that he's having a good pre-season, because he needs one. That being said...he still improved on his rating that he had at Melbourne, so maybe there's hope.

I'm fairly certain the thoughts were that he got dropped after his best game of the season, not that his average champion data mega points were better another player of a similar age. I think the crux of the argument came down to "You can't just drop him after his best game of the season, he's improving, getting more confidence and could be a weapon if we invest in him" vs "We're a footy team at the highest level, pick the best team on the day, no carrying players, no golden passes."
 
Westhoff is rated statistically as an 'elite' 'KPF' because he doesn't actually play as a KPF. He's a tall wing/flanker utility who keeps getting KPF eligibility because CD obviously feels too weird about not naming a 199cm player as a KPP. The SC requirements to be an 'elite' KPF are a lot lower than they are to be an elite midfielder/flanker.
 
I understand this is a statistics thread so my opinion is literally the opposite of what this is about but I have an issue with some of the analysis.

Does anyone think Westhoff is an elite KPF, even within the scope of the new type of KPF?

Depends on what your definition of a key forward is. As they say, the only two players to statistically do what Westhoff has done since 2010 are Reiwoldt and Adelaide's Tom Lynch. Anyone who spends 74% of their time in the forward line is either going to be a general forward or a key forward, and he's not a general forward - so he's a key forward by definition. But he's really a unique animal all on his own.

“He’s played eight of the best first games you’ll ever see” - Ken Hinkley
“He’s a signficicant player in our team right now" - Ken Hinkley
"He gets big match-ups, we look for the most damaging small forwards and say ‘Darc, your turn’, and he’s been able to get every one of them" - Ken Hinkley
“I don’t think anyone would say ‘Darce’ hasn’t earned his place in our team this year, but that comes from two years of really strong, hard work in the SANFL and during pre-season." - Ken Hinkley

I'd say our Head Coach disagrees with you and Champion Data.

That's a coach giving his player the support needed to keep him improving as a player (and get him a rising star nomination). His first four games were great statistically - he rated 12.7 AFL Rating points per game. But after R8 - his sixth game, he dropped to 5.7 points per game. So when Hinkley says he's played eight of the best first games you'll ever see, he's right, because it was only after that time that coaches started to pay him respect.

Byrne-Jones was significant because he was able to do a role that enabled Broadbent to push up onto the wing and fill that role. He also allowed Impey to play forward. He has a place as a lock down defender, but my point wasn't that he isn't good - it's that we need him to get better, just like the rest of our defence. And he will.

I'm fairly certain the thoughts were that he got dropped after his best game of the season, not that his average champion data mega points were better another player of a similar age. I think the crux of the argument came down to "You can't just drop him after his best game of the season, he's improving, getting more confidence and could be a weapon if we invest in him" vs "We're a footy team at the highest level, pick the best team on the day, no carrying players, no golden passes."

I know what the argument was. I'm just saying that even in his best game of the season, he wasn't that great - so it was important for him to go back into the SANFL and find his purpose in the side. Even at SANFL level, he was poor at contested possessions, averaging four per game. It was absolutely the right call to drop him so he could build up some confidence at a lower level. We played him in reserves because we are investing in him.
 
Ball Movement

16th for moving the ball in defensive 50 (this is where intercept marking comes into effect), 8th in defensive midfield, 17th in centre/bounce, 6th in attacking midfield at 15th in forward 50. Again, you can see exactly where opposition coaches line up their zones - in defensive 50, in centre midfield and in forward 50.

Defensive Ball Movement

6th for defending the opposition's forward 50, 5th for defending their defensive midfield, 2nd for defending centre/bounce, 9th for defending attacking midfield and 1st in attacking 50.

What does this tell us? It tells us that we had a lot of players that were so tired chasing jumpers all season doing their defensive assignments that they didn't have the explosive power necessary to break through zones. In fact, the entire year was pretty much a holding pattern - we were pretty much Richmond in everything - finishing around 9th/10th in most categories. Which isn't bad when you consider that we are on record as saying that we got our preseason wrong and missed Ryder and a whole bunch of other players through injury and suspension.
 
I know what the argument was. I'm just saying that even in his best game of the season, he wasn't that great - so it was important for him to go back into the SANFL and find his purpose in the side. Even at SANFL level, he was poor at contested possessions, averaging four per game. It was absolutely the right call to drop him so he could build up some confidence at a lower level. We played him in reserves because we are investing in him.

In what way is your analysis on Toumpas saying his last game wasn't that great? You've literally just ripped his average figures for the year and and said that's why he got dropped.
 
In what way is your analysis on Toumpas saying his last game wasn't that great? You've literally just ripped his average figures for the year and and said that's why he got dropped.

Let's say you run a barbershop, and a close friend comes in to help support your business. $20 is the going rate for a haircut, but your friend only has $10 on him. So you let it slide. Every six weeks, he comes in and it's the same story - $10. Finally, you've had enough - it's actually costing you money to cut your friend's hair. So you reason with him and say 'Look, I want to cut your hair, but I can't do it for $10. The going rate is $20.' Your friend agrees, and says he will make it up to you on the next visit...then when it comes to pay up, he gives you $15. Are you going to be happy because he's given more than he did before? Or pissed off because he still can't meet the standard?

Toumpas gave us $15. And that's why he was dropped.
 

(Log in to remove this ad.)

Let's say you run a barbershop, and a close friend comes in to help support your business. $20 is the going rate for a haircut, but your friend only has $10 on him. So you let it slide. Every six weeks, he comes in and it's the same story - $10. Finally, you've had enough - it's actually costing you money to cut your friend's hair. So you reason with him and say 'Look, I want to cut your hair, but I can't do it for $10. The going rate is $20.' Your friend agrees, and says he will make it up to you on the next visit...then when it comes to pay up, he gives you $15. Are you going to be happy because he's given more than he did before? Or pissed off because he still can't meet the standard?

Toumpas gave us $15. And that's why he was dropped.
Interesting analogy, id say that's fair except the mate could pay you $30-40 in the future compared to say the other bloke who comes in and gives you $15 one week and $25 the next.

Also the fact that you're giving the bloke hair cuts means he is more likely to pay more in the future.

Also by the time he pays you $15 you realise your going to have to take a bath financially so investing in future cash flows isn't the worst idea.

And a couple others that I can't make fit into this hair dresser scenario.




On XT1635-02 using BigFooty.com mobile app
 
Here's some interesting stats on meters gained - a source of much derision over at the aviary. In italics are their plus or minus on their previous four year average.

AFL Average - Midfielder - 330 M

Elite: Robbie Gray (402m +104),

Above Average: Jared Polec (363m +39), Ollie Wines (367m +70)

Average: Travis Boak (319m -54)

Below Average: Brad Ebert (288m -97), Sam Gray (280m +51), Jimmy Toumpas (257m +111)

AFL Average - Mid-Forward - 255 M

Above Average: Karl Amon (296m +135)

Below Average: Brendon Ah Chee (193m +9)

AFL Average - General Defender - 297 M

Elite: Riley Bonner (625m), Matthew Broadbent (376m -3), Jasper Pittard (554m +190)

Above Average: Hamish Hartlett (324m -59)

Average: Darcy Byrne-Jones (290m), Nathan Krakouer (277m +4),

Below Average: Tom Jonas (187m +21)

AFL Average - Key Defender - 194 M

Average: Jack Hombsch (203m -12)

Below Average: Logan Austin (131m), Tom Clurey (126m +12)


We improved 500m in 2016, registering 5858m gained over the season, which is better than the league average of 5481m. How much better? A Matthew Broadbent better (377m).
 
Also, how the **** did we drop 5 points against the Dogs, who were top four at the time and flying? They must really over rate the home field advantage we have in their model, or underrate the Dogs as a side.

The Dogs were pretty shithouse in 2014, and North's utter capitulation last year is a good idea of why you can't just blindly trust the last two months of results.

Arc's model proved its success with a relatively small annual regression (10%), and because of the Elo system's natural requirement to actually win (or lose!) a decent number of games to substantially change your ranking - e.g., a 1200 Elo chess player somehow fluking a victory against a Grand Master wouldn't suddenly give him a 2800 Elo - it ends up having about a 2 year rolling window of analysis.
 
Last edited:
Westhoff isn't an elite forward if we're measuring by any metrics that are important for a forward.

Ebert plays as more of a defence stretching link man than a centre square mid, so i'd say what's hurt him the most is that teams set up specifically to counter his work in transition, because if he's allowed to play his game he's devestating. Ryder coming back will help, but White coming back will help even more because he'll take some attention away from Ebert on the rebound. White isn't a forward target and shouldn't be though wtf. If he's scoring because he's a solo forward target it's surely because he's sprinted into the forward 50 first when we're rebounding.

Dixon was the only target in a forwardline always clogged with players, being delivered the ball by the worst 50 entry team in the league. If we can sort out our tactics going forward and actually make space instead of bombing away and relying on the individual brilliance of the likes of Wingard and Young, Dixon will find himself in a lot more space a lot more of the time.

Pressure acts is a junk stat and if you're hanging your hat on pressure acts and getting a lot of games, you probably play for a bad team. Neade has a lot of ability but needs to read the play better to get in better positions to actually impact the game more often. He's been given so much opportunity and has passed 50 games. Some return on investment would be nice.

Byrne-Jones is going to have some lower than average stats because he takes the game on. We dealt with Pittard turning the ball over a lot in his formative years and we'll do the same with DBJ because, like Pittard, even with the turnovers he makes us a better side. He's best 22.
 
Westhoff isn't an elite forward if we're measuring by any metrics that are important for a forward.

Westhoff vs (AFL 2016 average key forward)

Disposals: 16.5 (12.2)

Kicks: 9.9 (7.8)

Handballs: 6.8 (4.4)

Kick/Handball Ratio: 1.50 (1.88)

Retention Rate: 72.7% (69.1%)

Inside 50s: 2.9 (2.1)

Contested Possessions: 6.4 (5.6)

Uncontested Possessions: 10.1 (6.8)

Marks: 5.7 (5.2)

Contested Marks: 1.3 (1.6)

Ground Ball Gets: 3.3 (2.9)

1-on-1 win percentage: 31.5 (30.5)

Marks inside 50: 1.2 (2.0)

Ball Gets Inside 50: 0,6 (1.2)

Goals: 1.1 (1.7)

Behinds: 0.7 (1.0)

Accuracy %: 53.5 (55.6)

Score Assists: 1.1 (0.9)

Score Involvements: 5.4 (5.8)

Tackles: 3.2 (2.1)

Ebert plays as more of a defence stretching link man than a centre square mid, so i'd say what's hurt him the most is that teams set up specifically to counter his work in transition, because if he's allowed to play his game he's devestating. Ryder coming back will help, but White coming back will help even more because he'll take some attention away from Ebert on the rebound. White isn't a forward target and shouldn't be though wtf. If he's scoring because he's a solo forward target it's surely because he's sprinted into the forward 50 first when we're rebounding.

Agreed, and it's why White going down was such a huge blow to our fortunes in 2016. You could see it with Sydney - as soon as Rohan went down against Adelaide, the Crows came back, not because they are actually any good, but because without the dual threat of transition it's too easy to focus on limiting and corralling one player. But Ebert still suffered a decline in the clearance numbers when he does rotate through the middle. As for White, that's exactly what they are referring to - a target is just who you direct your kick to, nothing more, nothing less.

Dixon was the only target in a forwardline always clogged with players, being delivered the ball by the worst 50 entry team in the league. If we can sort out our tactics going forward and actually make space instead of bombing away and relying on the individual brilliance of the likes of Wingard and Young, Dixon will find himself in a lot more space a lot more of the time.

Dixon won 34% of his one on ones, which is still 15th for key forwards, but he won 42% in 2015. That's to do with fitness IMO. He needs to work harder up the ground to create space and pump up his marking numbers. If he was the only target, that came from the inability of our players to push up the ground. In 2014 we had 4 targets that would get the ball over 12% of the time - it dropped to two in 2016 (Dixon and Wingard), with Dixon getting it 33.6% of the time.

I know if we use him properly, which would be more of the Collingwood game, he'll be a great asset. It's why I'm not worried that he hasn't be on the track for much of preseason, because he doesn't need to work on contested stuff, but uncontested stuff.

Pressure acts is a junk stat and if you're hanging your hat on pressure acts and getting a lot of games, you probably play for a bad team. Neade has a lot of ability but needs to read the play better to get in better positions to actually impact the game more often. He's been given so much opportunity and has passed 50 games. Some return on investment would be nice.

If it was a junk stat they wouldn't measure it. But he's elite at forward 50 tackles too. He needs to up his possessions more, but that comes from Dixon being targeted as much as he was. If Charlie is double teamed, he should be leading away from where our other forwards are and dragging his defenders with him so guys like Neade can do their work in the space he creates.

Byrne-Jones is going to have some lower than average stats because he takes the game on. We dealt with Pittard turning the ball over a lot in his formative years and we'll do the same with DBJ because, like Pittard, even with the turnovers he makes us a better side. He's best 22.

Actually, Byrne-Jones was pretty much average for a general defender in 2016 across every stat, so he takes the game on no more than any other defender in the competition. We don't want to be average, we want to be elite. so he needs to go up another level this year. And he will.
 
Boak was crap the first 8 weeks because he has playing injured. Hartlett was crsp when he played injured.
Hinkley needs to learn if injured players put their hand up to play and don't perform they are an automatic rest the next week.
You are better off getting them right than accepting an ouput that is equivelant to a fringe player. This way when they come back you get the performance you want.
 
The Dogs were pretty shithouse in 2014, and North's utter capitulation last year is a good idea of why you can't just blindly trust the last two months of results.

Arc's model proved its success with a relatively small annual regression (10%), and because of the Elo system's natural requirement to actually win (or lose!) a decent number of games to substantially change your ranking - e.g., a 1200 Elo chess player somehow fluking a victory against a Grand Master wouldn't suddenly give him a 2800 Elo - it ends up having about a 2 year rolling window of analysis.

As a chess ranking system, Elo is great, because there's no variables in how a player can perform due to gamestyle - you can only ever do one of 32 moves at one particular moment. In association football, in baseball - in quantitative, rigidly regulated sports where there are only a finite number of options for each player, it works.

In a game like Australian Rules where it's so fluid and dynamic and fortunes change on who gets injured/suspended because the fluctuation of performance is so high between players in the squad, I just don't think it works. Adelaide was rated so highly because they had their best 22 on the park more than any other side - the garbage that Champion Data wrote in their writeup about how Pyke changed their gamestyle to be more about controlling territory and less about scoring efficiency during the second half of the year doesn't take into account the sides that they played during that time, which were all poor. It's easy to control territory when you're playing against the likes of Essendon, Brisbane, Fremantle etc.

Boak was crap the first 8 weeks because he has playing injured. Hartlett was crsp when he played injured.
Hinkley needs to learn if injured players put their hand up to play and don't perform they are an automatic rest the next week.
You are better off getting them right than accepting an ouput that is equivelant to a fringe player. This way when they come back you get the performance you want.

I reckon this is the reason why we went all in into the draft. While Boak and Hartlett had poor performances, who was going to replace them? No one that was actually deserving. But now we have Powell-Pepper, Drew and Atley in the squad, there's no reason to think that '80% of Boak' is better than '100% of Atley'.
 
As a chess ranking system, Elo is great, because there's no variables in how a player can perform due to gamestyle - you can only ever do one of 32 moves at one particular moment.

You should probably have a coffee and come back to that one.

Adelaide was rated so highly because they had their best 22 on the park more than any other side

This is mostly correct! As it turns out being less injury-prone tends to be fairly positive for sports teams. Broadly speaking I'm going to expect the team consistently fielding their best 22 to one that has to ask other teams on Facebook to lend them players.

the garbage that Champion Data wrote in their writeup about how Pyke changed their gamestyle to be more about controlling territory and less about scoring efficiency during the second half of the year doesn't take into account the sides that they played during that time, which were all poor. It's easy to control territory when you're playing against the likes of Essendon, Brisbane, Fremantle etc.

Can't speak for Champion Data's methodology and analysis, haven't found a copy of the Prospectus yet.
 
Boak was crap the first 8 weeks because he has playing injured. Hartlett was crsp when he played injured.
Hinkley needs to learn if injured players put their hand up to play and don't perform they are an automatic rest the next week.
You are better off getting them right than accepting an ouput that is equivelant to a fringe player. This way when they come back you get the performance you want.
So our coach does lie. He kept telling us Boak wasn't injured he was just out of form. No wonder people lost faith in Hinkley after being seen as the new messiah.
 
You should probably have a coffee and come back to that one

16 pieces per side, 2 sides = 32 possible piece movements to consider (because if you're a true chess player you're considering what your opponent is going to do as well). I know there are variations on where those pieces move, but those movements are fixed - a pawn can only move a maximum of one square forward (two initially), a bishop can only move diagonally, a rook can only move horizontally and vertically etc. The positions they can possible move to is infinite during the course of a game, but that's not what I was referring to. The fact is, everyone knows how to play chess and their is a rigid game structure of movement governing it. But when you get to a sport that is so variable like Australian Rules football, you just can't use the same system.

For example - there is no way that Adelaide 2016 was, at any stage, a better side than Port Adelaide 2004. But Elo rankings say they were. Why? Because the system doesn't account for injuries to players, which doesn't affect things so much in other sports because the variables in squad sizes aren't as large due to having a greater talent pool from which to draw from.

There are way too many non-mathematical variables in this game to do some sort of statistical prediction of outcomes. It's cool to look at retrospectively, but what they should also do is give weighting to the strength of the league itself for that year. But then it becomes subjective rather than objective.
 

Remove this Banner Ad

Opinion "Help me out where I need faith!" - The Statistical Data Thread

Back
Top