Changes v Hawthorn

[Demonsone 2,042]

On 31/07/2016 at 5:12 PM, demon heart said:

oscar mcdonald. turn over king

frost. turn over queen. carnt kick.

wagner. turn over prince. misses easy targets

tom mcdonald. turn over princess. no nouse

 

surly dunn, garland, pedoson and grimes arnt all behind these guys. coaching staff have put a line through there names next year.

wow thats our back line if their kicking dosent improve....will never win a flag

top teams defenders hit targets, question is though are the mids etc going to right positions?

[Nasher 33,616]

3 hours ago, mo64 said:

Not sure what the point of your post is? Steve is entitled to dismiss the quantitative argument of Disposal Efficiency, because like me, he sees it as a meaningless stat.

 

Why do you believe it is meaningless?  Tell me what's wrong with it - what doesn't it capture that you think it should?  What is it about the way it is measured that brings you to the conclusion that it is misleading?

I find flippant dismissal of data at hand to be a frustrating way of debating a topic, because usually those doing it do so purely because it doesn't support whatever over the top point they're trying to support is.  You guys use your eyes - great, so do the guys counting the stats for a living.  I do understand that statistics aren't everything, that they can be manipulated, and used to support pretty much any point, and so on.  But just once, I'd love to see a good poster like Steve go "wow, the quantitative data doesn't support my point at all, perhaps I should re-evaluate?" but nah, it's always "stats are for stupid people, try opening your eyes".  Booooor-ing.

[binman 43,138]

3 minutes ago, Nasher said:

Why do you believe it is meaningless?  Tell me what's wrong with it - what doesn't it capture that you think it should?  What is it about the way it is measured that brings you to the conclusion that it is misleading?

I find flippant dismissal of data at hand to be a frustrating way of debating a topic, because usually those doing it do so purely because it doesn't support whatever over the top point they're trying to support is.  You guys use your eyes - great, so do the guys counting the stats for a living.  I do understand that statistics aren't everything, that they can be manipulated, and used to support pretty much any point, and so on.  

Of course stats are useful and taken together are an important tool to understand the game. All coaches rely on them. And all clubs employ people to colelct the and analyse them in and out of games. 

Whilst individual coaches emphasis the importance particular stats (eg Clarkson being keen on score involvements) all are important. I would argue DE is one of the more important as it so often is the difference between winning and losing. The Hawks have won three flags on the back on their outstanding DE and accuracy at goals. 

You could see with your eyes that both GC and the dees were poor with their disposal, but the stats supported it. We won the game with our second quarter when our DE was 72% to their 50%. Coincidence? Of course not. As if roos would not have had the DE stats at hand at the quarter time huddle. 

[Nasher 33,616]

33 minutes ago, binman said:

Of course stats are useful and taken together are an important tool to understand the game. All coaches rely on them. And all clubs employ people to colelct the and analyse them in and out of games. 

Whilst individual coaches emphasis the importance particular stats (eg Clarkson being keen on score involvements) all are important. I would argue DE is one of the more important as it so often is the difference between winning and losing. The Hawks have won three flags on the back on their outstanding DE and accuracy at goals. 

You could see with your eyes that both GC and the dees were poor with their disposal, but the stats supported it. We won the game with our second quarter when our DE was 72% to their 50%. Coincidence? Of course not. As if roos would not have had the DE stats at hand at the quarter time huddle. 

Exactly.  

The reason I do like to fall back on stats, and the coaches do too, is because at times my eyes miss things, and other times my eyes blatantly lie.  So do everyone else's.  You only need to look at the Tom McDonald thread at how utterly polarized we all get, having watched exactly the same game(s).  

It's just a fact of life that you can't possibly take it all in, and we *all* watch footy with some predisposition, so confirmation bias results.  If we already hold the view that Tom McDonald or James Harmes is going turn it over with that next kick, then they do, that poor kick holds far more weight than the 5 handballs in 5 chains of play that all hit targets, then suddenly we're claiming that 21 of Harmes' 23 disposals went straight to the opposition.  Or that Disposal Efficiency is a stupid stat - after all, it never reflects what our eyes saw.

Stats offer information that is independent and not subject to bias, so they are used to support - or refute - what our own slow, biased eyes missed.  People who flippantly dismiss them miss an opportunity to critically reflect and re-evaluate their own opinion.  Their own loss.

Edit: Sorry if this post is probably just rambling gibberish.  My brain struggled to find the words numerous times throughout - usually a sign I badly need sleep.  Time to go find it I think!

[Demons11 7,032]

12 hours ago, DemonWA said:

Out: Garlett, Wagner, M. Jones

In: Terlich, Grimes, Trengove

I kid, the Terlich call is a troll, I'd bring in Grimes, Trengove and Oliver

If Hogan is injured I'd being in Pedo not Weed.

 

So let's bring in the 3 slowest players on our list for 3 of the quickest

[stuie 7,374]

16 minutes ago, Nasher said:

Exactly.  

The reason I do like to fall back on stats, and the coaches do too, is because at times my eyes miss things, and other times my eyes blatantly lie.  So do everyone else's.  You only need to look at the Tom McDonald thread at how utterly polarized we all get, having watched exactly the same game(s).  

It's just a fact of life that you can't possibly take it all in, and we *all* watch footy with some predisposition, so confirmation bias results.  If we already hold the view that Tom McDonald or James Harmes is going turn it over with that next kick, then they do, that poor kick holds far more weight than the 5 handballs in 5 chains of play that all hit targets, then suddenly we're claiming that 21 of Harmes' 23 disposals went straight to the opposition.  Or that Disposal Efficiency is a stupid stat - after all, it never reflects what our eyes saw.

Stats offer information that is independent and not subject to bias, so they are used to support - or refute - what our own slow, biased eyes missed.  People who flippantly dismiss them miss an opportunity to critically reflect and re-evaluate their own opinion.  Their own loss.

Edit: Sorry if this post is probably just rambling gibberish.  My brain struggled to find the words numerous times throughout - usually a sign I badly need sleep.  Time to go find it I think!

Disposal efficiency is kind of the odd one out though. I was shocked when I looked up how it was worked out. One example is if someone kicks a horrible kick to a 2 on 1 in our favour but the defender wins it, it's counted as an "effective disposal". That's just one example, but you can see how with things like that it wouldn't be a useful stat.

Also, I would argue coaches don't put as much weight into the same stats that we look at. Recently Chris Scott was talking about how he doesn't even bother with contested possessions as he considers it meaningless. Coaches would likely have KPI's for individual players, but they talk about things like hitouts to advantage and effective tackles, which aren't considered "proper" stats. This would also be why coaches conduct extensive reviews of match footage rather than just look up footywire.

 

Edited August 1, 2016 by stuie

[titan_uranus 25,020]

Haven't read the whole thread, got tired of reading people saying Dawes, OMac or M Jones should be dropped.

Dawes is improving and took some contested marks for the first time in who knows how long. Not the time to be dropped.

OMac played his best defensive game for the club against quality forwards in Lynch, Day and Wright. Ridiculous call.

M Jones is having disposal issues but so are about 10 other players. Meanwhile M Jones continues to run all day, use his pace, and provide the key outside run that only he and Stretch are consistently providing.

IMO Harmes is the first one out, he gets a tonne of it but adds very little to the side offensively with far too many turnovers. Garlett surely is in trouble too, the more so if Kennedy is in form in the VFL.

55 minutes ago, Nasher said:

Why do you believe it is meaningless?  Tell me what's wrong with it - what doesn't it capture that you think it should?  What is it about the way it is measured that brings you to the conclusion that it is misleading?

I find flippant dismissal of data at hand to be a frustrating way of debating a topic, because usually those doing it do so purely because it doesn't support whatever over the top point they're trying to support is.  You guys use your eyes - great, so do the guys counting the stats for a living.  I do understand that statistics aren't everything, that they can be manipulated, and used to support pretty much any point, and so on.  But just once, I'd love to see a good poster like Steve go "wow, the quantitative data doesn't support my point at all, perhaps I should re-evaluate?" but nah, it's always "stats are for stupid people, try opening your eyes".  Booooor-ing.

I agree with you on using stats and I haven't read the thread fully, but disposal efficiency is a flawed measurement.

A kick that goes 40 metres or longer to a 50/50 is counted as an effective disposal. A handpass to a stationery player is effective even if that player is about to get crunched, or even if the handpass is a hospital handpass. A short kick that ends up with the intended recipient is effective even if it misses that player in open space, but he is able to run onto it.

I don't like the way it tries to measure the thing it's aimed at (quality of disposal).

[mo64 5,875]

16 minutes ago, titan_uranus said:

I agree with you on using stats and I haven't read the thread fully, but disposal efficiency is a flawed measurement.

A kick that goes 40 metres or longer to a 50/50 is counted as an effective disposal. A handpass to a stationery player is effective even if that player is about to get crunched, or even if the handpass is a hospital handpass. A short kick that ends up with the intended recipient is effective even if it misses that player in open space, but he is able to run onto it.

I don't like the way it tries to measure the thing it's aimed at (quality of disposal).

That's exactly my point. 

[Deestroy All 14,266]

Haven't read the thread but I noticed Tyson having good high up feel of his hammy after the game. Was giving it a deep fingering. So keep an eye on if he's in full training this week. 

 

Wouldn't mind if Hogan misses this week. Have a more defensively mobile forward line to stop the Hawks waltzing it out. 

[Rod Grinter Riot Squad 5,654]

1 hour ago, Demons11 said:

So let's bring in the 3 slowest players on our list for 3 of the quickest

Grimes is not slow.

 

i would have him before M.Jones every day of the week (though after many others), both have their issues with disposal but Grimes has a flatter kick when passing and can use both feet.

[binman 43,138]

1 hour ago, titan_uranus said:

 

I agree with you on using stats and I haven't read the thread fully, but disposal efficiency is a flawed measurement.

A kick that goes 40 metres or longer to a 50/50 is counted as an effective disposal. A handpass to a stationery player is effective even if that player is about to get crunched, or even if the handpass is a hospital handpass. A short kick that ends up with the intended recipient is effective even if it misses that player in open space, but he is able to run onto it.

I don't like the way it tries to measure the thing it's aimed at (quality of disposal).

I have read this argument many times here on dl. You say DE is a flawed measurement and spell out your reasons why. I say that is a flawed argument.

It is useful as a comparitive stat when aggregated. Why? Because all clubs now use variants of the hawks template of keepings off. The difference between the best teams and the teams down the ladder is the best teams do the simple things well and consistenltly hit targets - both the easy ones chipping around the balk half, the 20 metre chips to forwards and the high risk ones to the corridor. 

As evidence that it is a good measure is the best sides have the best de. Again it is one of the key elements of the hawks dominance

[DeeSpencer 25,723]

1 hour ago, binman said:

I have read this argument many times here on dl. You say DE is a flawed measurement and spell out your reasons why. I say that is a flawed argument.

It is useful as a comparitive stat when aggregated. Why? Because all clubs now use variants of the hawks template of keepings off. The difference between the best teams and the teams down the ladder is the best teams do the simple things well and consistenltly hit targets - both the easy ones chipping around the balk half, the 20 metre chips to forwards and the high risk ones to the corridor. 

As evidence that it is a good measure is the best sides have the best de. Again it is one of the key elements of the hawks dominance

We've lost and nearly lost the last two weeks due to overusing the ball without getting momentum to the build up. The Hawks almost always are going backwards or sideways to then go forwards. And a lot of that is because of the movement off the ball which DE can't factor in.

Zac Dawson has one of the best DE in the league. That's how you know it's a bogus stat. The guy couldn't kick a drop punt if his life depended on it.

Then we get the issue that Champion Data might not even code each play properly. If a >40m kick has to go to a 50/50 then this is wrong:

 

[Samael 560]

11 hours ago, Nasher said:

Why do you believe it is meaningless?  Tell me what's wrong with it - what doesn't it capture that you think it should?  What is it about the way it is measured that brings you to the conclusion that it is misleading?

I find flippant dismissal of data at hand to be a frustrating way of debating a topic, because usually those doing it do so purely because it doesn't support whatever over the top point they're trying to support is.  You guys use your eyes - great, so do the guys counting the stats for a living.  I do understand that statistics aren't everything, that they can be manipulated, and used to support pretty much any point, and so on.  But just once, I'd love to see a good poster like Steve go "wow, the quantitative data doesn't support my point at all, perhaps I should re-evaluate?" but nah, it's always "stats are for stupid people, try opening your eyes".  Booooor-ing.

Stats like the DE% are important in my opinion, because they are measured the same way for all players. It is a valuable comparison, regardless of what they mean. Because they are measured the same way for each player they have meaning. A player with 90% DE is making better use of the ball than a player with 60% or 50%. We can debate the way it is measured but surely the aim would be to get as near 100% and a player approaching that level is more valuable than one who isn't. The emphasis placed on DE% may vary between coaching philosophies but that is another issue. Personally I don't rate tackles because it means your team does not have the ball but I am sure that tackles are important to say a defence coach. 

[stuie 7,374]

9 hours ago, binman said:

I have read this argument many times here on dl. You say DE is a flawed measurement and spell out your reasons why. I say that is a flawed argument.

It is useful as a comparitive stat when aggregated. Why? Because all clubs now use variants of the hawks template of keepings off. The difference between the best teams and the teams down the ladder is the best teams do the simple things well and consistenltly hit targets - both the easy ones chipping around the balk half, the 20 metre chips to forwards and the high risk ones to the corridor. 

As evidence that it is a good measure is the best sides have the best de. Again it is one of the key elements of the hawks dominance

As a team, Sydney have the worst disposal efficiency in the league, St Kilda have the best.

It's clearly a flawed stat.

 

[La Dee-vina Comedia 16,816]

13 hours ago, hardtack said:

Ok, I promise to pay more a10tion next time.

While some might quibble and point out that you should have gone with "2 pay more attention" I'd suggest it's time to extend yourself and go cryptic. Why not "toupee" or, even better, go fully cryptic with "a patch of false hair for covering a bald spot"?

[old dee 23,801]

23 minutes ago, La Dee-vina Comedia said:

While some might quibble and point out that you should have gone with "2 pay more attention" I'd suggest it's time to extend yourself and go cryptic. Why not "toupee" or, even better, go fully cryptic with "a patch of false hair for covering a bald spot"?

Now you are being silly.

But why not it is Tuesday after a win!

[La Dee-vina Comedia 16,816]

33 minutes ago, old dee said:

Now you are being silly.

But why not it is Tuesday after a win!

I can be silly 7 days a week, win, lose or draw. 

[old dee 23,801]

Just now, La Dee-vina Comedia said:

I can be silly 7 days a week, win, lose or draw. 

A wise state of mind LDC being a MFC supporter is no walk in the park it pays to keep ones sense of humour.

[Fat Tony 5,253]

12 hours ago, titan_uranus said:

Dawes is improving and took some contested marks for the first time in who knows how long. Not the time to be dropped.

Let's not get carried away with Dawes' game. He had 8 possessions and zero contested marks according to the stats sheet. http://afltables.com/afl/stats/players/C/Chris_Dawes.html (He did take two clean marks in the hands which drew Bronx cheers from the crowd.) Dawes is a key forward who is poor both in the air and on the ground and is not going to be part of the future. We should give Weideman the last four games of the season now that we cannot make finals.

 

Edited August 2, 2016 by Fat Tony
spelling

[ArtificialWisdom 4,045]

1 hour ago, stuie said:

As a team, Sydney have the worst disposal efficiency in the league, St Kilda have the best.

It's clearly a flawed stat.

 

I think the stat we want to see is TO + TO assists. Who starts chains that result in turnovers or who gives it to a player who is out of position or under significant pressure. Maybe TO assists only where the payer who turns it over is under pressure. Not sure, just wish there was someway we could access Champion Data's full database(s) of stats. I could spend hours digging through looking for stuff.

DE is a far more usefull stat than you make it out to be. But yes it does have obvious flaws. Game style for 1... St Kilda are an outside running team so of course their DE is higher than Sydney. Wonder what happens if you compare DE while under pressure between sydney and stkilda it's probably not that far appart. Im also a bigger fan of +/- compared to opponent in each game. Comparing sydney DE on a wet day and St Kilda under the roof is unfair. However comparing them to the same game is far more fair. 
 

[stuie 7,374]

4 minutes ago, ArtificialWisdom said:

I think the stat we want to see is TO + TO assists. Who starts chains that result in turnovers or who gives it to a player who is out of position or under significant pressure. Maybe TO assists only where the payer who turns it over is under pressure. Not sure, just wish there was someway we could access Champion Data's full database(s) of stats. I could spend hours digging through looking for stuff.

DE is a far more usefull stat than you make it out to be. But yes it does have obvious flaws. Game style for 1... St Kilda are an outside running team so of course their DE is higher than Sydney. Wonder what happens if you compare DE while under pressure between sydney and stkilda it's probably not that far appart. Im also a bigger fan of +/- compared to opponent in each game. Comparing sydney DE on a wet day and St Kilda under the roof is unfair. However comparing them to the same game is far more fair. 
 

If they could simplify it and just make it say "useful disposals" that would be good. Disposals that were to a team mates advantage, or were part of an effective chain, or ended in a mark to the same team... It's just the criteria for disposal efficiency that I have a problem with.

 

[ArtificialWisdom 4,045]

3 minutes ago, stuie said:

If they could simplify it and just make it say "useful disposals" that would be good. Disposals that were to a team mates advantage, or were part of an effective chain, or ended in a mark to the same team... It's just the criteria for disposal efficiency that I have a problem with.

Yeah agreed, I'm sure that stat exsists, its just we have little to no access to any usefull stats that champion data provides the clubs. Only stats we get to see are the super dumbed down ones so that a 5yo can understand them. We need a man on the inside to collect the usefull stats for us! I think if we had full access to those types more complicated stats we might begin to see the areas that players like M. Jones, OMac and Dawes excel which keep them in the team despite many on here kicking and screaming about it.

[binman 43,138]

8 hours ago, DeeSpencer said:

We've lost and nearly lost the last two weeks due to overusing the ball without getting momentum to the build up. The Hawks almost always are going backwards or sideways to then go forwards. And a lot of that is because of the movement off the ball which DE can't factor in.

Zac Dawson has one of the best DE in the league. That's how you know it's a bogus stat. The guy couldn't kick a drop punt if his life depended on it.

Then we get the issue that Champion Data might not even code each play properly. If a >40m kick has to go to a 50/50 then this is wrong:

 

Again this is a spurious argument. of course if DE is taken in isolation, for one player it can throw up misleading examples, such as with Dawson and as some here on DL have said Grimes. 

Its usefulness as a stat is aggregating it across the team and comparing sides and setting benchmarks.

Clubs also no doubt use it conjunction with other stats not in isolation to provide analysis. For example looking at a combination of DE and score involvements would highlight how effective jack Watts is and reflect his value to the team in a way that looking at those two stats in isolation. 

Further they would never look at players DE without also looking at the data they have on where disposals occur, in what circumstances and what the outcome of the disposal is. in addition the footy department and coaches also 'use their eyes' both during the game and in the in the incredibly forensic post game analysis where every disposal is analysed.

So looking at the DE in isolation is something that only happens on footy forums and by dullards in the media.

However even looking at DE in isolation reflects that for all the outliers like Dawson the opposite is never true - that's to say the bets users of the ball (think Watts, Gunston, etc etc) have high DE across their careers. 

[sue 9,107]

10 minutes ago, stuie said:

If they could simplify it and just make it say "useful disposals" that would be good. Disposals that were to a team mates advantage, or were part of an effective chain, or ended in a mark to the same team... It's just the criteria for disposal efficiency that I have a problem with.

 

or reverse the definition. Seem to me that if you consider the stat to mean 'percentage of not complete stuff-ups' you get a meaningful measure. An example is that Burgoyne example earlier in this thread. Sure it didn't go the forward's advantage, but equally it didn't go straight to an oppo on his own or OOB 15 yards from where he kicked it.

[old55 23,625]

I'm a Weideman fan but I don't think he's ready for AFL, especially not against Hawthorn.  Maybe he could get a taste so he knows what he's up for, that's the only purpose that would be served.