Analysis Good, Bad and Not Much Ugly

[rabidcrow]

Lions had 59 kicks inside 50 and took 3 marks. That’s a combination of good pressure and the Lions wasting a few entries, but as a few have mentioned already, many of their kicks inside 50 were shallow or wide. Sure the Lions had periods of dominance, but as the disposals showed, this game was played on our terms.

Was interesting to watch the styles of play as well. From the start Brisbane we’re hacking the ball forward - lot of knocks and particularly soccer kicks. We didn’t seem to do either, especially first half - we hand balled and looked to use our more precise kicking game. So we racked up disposals and had much higher efficiency


Sent from my iPhone using BigFooty.com

 

[rabidcrow]

Surprised by some of the negative Dawson comment by supporters. The guy is clearly our only A grader midfielder and is a Brownlow contender. He is proving to be the perfect captain.

We have a team and a rebuild that is being lauded all over the country - and yet we find a group on here that will nitpick over pretty much anything that doesn’t go perfect

Dawson getting criticism is just another example of how ridiculous some people are


Sent from my iPhone using BigFooty.com

 

[FR0GGY]

We have a team and a rebuild that is being lauded all over the country - and yet we find a group on here that will nitpick over pretty much anything that doesn’t go perfect

Dawson getting criticism is just another example of how ridiculous some people are


Sent from my iPhone using BigFooty.com

I haven’t seen any criticism other than his disposal has been a bit off over the last 6 weeks and it has so not really sure the issue.
Even his “bad” games are still good.

 

[BRL121]

I haven’t seen any criticism other than his disposal has been a bit off over the last 6 weeks and it has so not really sure the issue.
Even his “bad” games are still good.

His disposal was shocking on Sunday, too many straight to oppo chests. I attribute this to the pressure of playing in the midfield and close attention. When he left the midfield and played his old game, his kicks were in the open and much closer to his normal quality. I also think he is showing a few signs of being worn down by his midfield time. At the risk of activating Elite, maybe he should have managed time in the midfield to preserve his overall influence?

 

[NSolis]

The stats are not trying to give you some kind of alternative reality (despite the person above saying ‘we still would have won’). They are showing you the difficulty of the shots each side took, and how their conversion compared to the expected result for the difficulty of each shot. Then they add it all up to show the aggregate result so you can see how well or badly each side kicked across the game, and how easy or hard their shots were. That’s it, that’s all!

Your point - if they kicked this one or that one it would have gone back to the middle and everything afterwards would be different is obviously true, but it has nothing to do with how hard each subsequent actual shot was.

The stats absolutely ‘mean’ something, they measure the difficulty of shots taken, and performance against expectation for that difficulty.

Weird that you are getting so much pushback on this, it's incredibly straightforward and is way more useful than the usual stuff you see with fans bemoaning 'poor goalkicking' leading to losses.

A team could kick 1.9 in a quarter where every single shot was a snap from the boundary under pressure or a set shot from 55 or whatever. This would probably not be particularly bad goalkicking as the likelyhood of kicking those goals is very low. A fan of that team who claimed they lost the game due to poor goalkicking would be wrong - they lost the game due to being unable to take better shots on goal, for whatever reason.

On the other hand if a team kicked 1.9 from set shots directly in front at a modest distance, this would be extremely poor, and the majority of the time they'd do better than that. Expected score tracks these outcomes and gives a view as to how things would have gone with statistically average performance.

In terms of useful information, one thing that xscore communicates is how many high quality goalscoring opportunities each side was able to generate. Sure sometimes you'll fluke a win by kicking a bunch of amazing goals, and sometimes you will blow easy chances and lose, but if you're generating lots of good looks you are likely to score plenty of goals. People already make these claims as fans, coaches and pundits - this is attempting to use data to test the claims.

 

[NSolis]

His disposal was shocking on Sunday, too many straight to oppo chests. I attribute this to the pressure of playing in the midfield and close attention. When he left the midfield and played his old game, his kicks were in the open and much closer to his normal quality. I also think he is showing a few signs of being worn down by his midfield time. At the risk of activating Elite, maybe he should have managed time in the midfield to preserve his overall influence?

Conditions too. Lots of good kicks were butchering the ball on Sunday.

 

[baldandbroke]

The stats are not trying to give you some kind of alternative reality (despite the person above saying ‘we still would have won’). They are showing you the difficulty of the shots each side took, and how their conversion compared to the expected result for the difficulty of each shot. Then they add it all up to show the aggregate result so you can see how well or badly each side kicked across the game, and how easy or hard their shots were. That’s it, that’s all!

Your point - if they kicked this one or that one it would have gone back to the middle and everything afterwards would be different is obviously true, but it has nothing to do with how hard each subsequent actual shot was.

The stats absolutely ‘mean’ something, they measure the difficulty of shots taken, and performance against expectation for that difficulty.

How do the stats measure the difficulty of shots taken?

Let's get into some detail.

• state of the game
• period in the game (time in quarter, time in game)
• level of physical fatigue
• injury status
• time since rest / time on ground
• mental state from previous shots
• mental state from crowd interaction (home vs away)
• wind
• rain
• man on the mark

All we have is the angle and the distance.

It's like amateur golf stats. I can track numbers using my GPS trackers, but unless I add in the details like the specifics of the lie, the hazards I'm avoiding, the weather, the distractions I'm dealing with, the perceived pressure, the practice I've done, what importance does the shot have to my score ... it's pretty basic and hard to get anything from.

So yeah, in my opinion - these predictive scoring stats are not really relevant or useful.

 

[NSolis]

It's like amateur golf stats. I can track numbers using my GPS trackers, but unless I add in the details like the specifics of the lie, the hazards I'm avoiding, the weather, the distractions I'm dealing with, the perceived pressure, the practice I've done, what importance does the shot have to my score ... it's pretty basic and hard to get anything from.

So yeah, in my opinion - these predictive scoring stats are not really relevant or useful.

This is expecting too much form a single stat, I would say. You're right that many factors go into the difficulty of an attempt on goal. You can't take into account everything, but taking into account angle and distance provides an analysis of how many high quality goalscoring opportunities were generated. If a team doesn't take them, they will probably underperform their expected score. That doesn't invalidate the analysis or mean that it would be better to just not have the information at all.

Similarly, disposal count doesn't tell you how well a player uses the ball. It tells you something else, which is also useful to know.

 

[baldandbroke]

This is expecting too much form a single stat, I would say. You're right that many factors go into the difficulty of an attempt on goal. You can't take into account everything, but taking into account angle and distance provides an analysis of how many high quality goalscoring opportunities were generated. If a team doesn't take them, they will probably underperform their expected score. That doesn't invalidate the analysis or mean that it would be better to just not have the information at all.

Similarly, disposal count doesn't tell you how well a player uses the ball. It tells you something else, which is also useful to know.

Well now we are getting somewhere ...

What does the disposal count tell you exactly? Because I think that is an irrelevant stat too.

 

[FR0GGY]

His disposal was shocking on Sunday, too many straight to oppo chests. I attribute this to the pressure of playing in the midfield and close attention. When he left the midfield and played his old game, his kicks were in the open and much closer to his normal quality. I also think he is showing a few signs of being worn down by his midfield time. At the risk of activating Elite, maybe he should have managed time in the midfield to preserve his overall influence?

Not sure about that, his disposal seems to be worse early in the game.
He’s still getting a tonne of it, so I don’t think fitness is an issue at all.
I think he’s just having more time put into him so he’s having less time to dispose of it.

 

[jackster83]

Not sure about that, his disposal seems to be worse early in the game.
He’s still getting a tonne of it, so I don’t think fitness is an issue at all.
I think he’s just having more time put into him so he’s having less time to dispose of it.

A few of our guys are like that. Worrell, McHenry and Parnell tend to turn it over a bit early - even in the SANFL.

Sholl is the other way round - usually starts well but loses his polish as he gets tired.

 

[Rory Walkerfield]

A few of our guys are like that. Worrell, McHenry and Parnell tend to turn it over a bit early - even in the SANFL.

Sholl is the other way round - usually starts well but loses his polish as he gets tired.

You can always bank on Worrell kicking one out on the full with his first or second touch.

 

[NSolis]

Well now we are getting somewhere ...

What does the disposal count tell you exactly? Because I think that is an irrelevant stat too.

It tells you how many times a player collected and disposed of the ball via kick or handpass.

 

[Crow_Magnon]

I agree with a lot of the sentiments above after watching the replay. Whilst Brisbane supposedly dominated I50's in the latter part of the 2nd qtr and for much of the 3rd quarter it didn't feel like they had a real ascendancy, and I felt we were still right in the game. Really pleased with the way the younger players such as Rachele stood up too. All in all, a pretty darn good team effort across the board. We'll miss Laird this week though...anyone in the SNAFL ready? Otherwise Smith / Hinge come in and we rotate a few through there to cover.

On the umpiring, however, staggering to me to see the missed mark for Hipwood, the hold on Tex wrongly paid to his opponent with four umpires on the field. I don't understand why we have extra umps if these sorts of decisions can't be challenged by another ump in a better position than the decision making ump. It wasn't just these decisions either, there were a bunch of mystifying ones - as there are most weeks.

 

[Stephen2]

One day, some dickhead like Zorko is gonna do a non-football incident like eye-gouge (Barry Hall broken jaw coward-punch comes to mind too).

And the victim will get the police involved for assault. And I'll be cheering for it.

That kind of stuff is not just a "dog act", it's a criminal act, and these scummy pricks should stop being allowed to get away with it just because people are playing a game together.

 

[MarcusP2]

How do the stats measure the difficulty of shots taken?

Let's get into some detail.

• state of the game
• period in the game (time in quarter, time in game)
• level of physical fatigue
• injury status
• time since rest / time on ground
• mental state from previous shots
• mental state from crowd interaction (home vs away)
• wind
• rain
• man on the mark

All we have is the angle and the distance.

It's like amateur golf stats. I can track numbers using my GPS trackers, but unless I add in the details like the specifics of the lie, the hazards I'm avoiding, the weather, the distractions I'm dealing with, the perceived pressure, the practice I've done, what importance does the shot have to my score ... it's pretty basic and hard to get anything from.

So yeah, in my opinion - these predictive scoring stats are not really relevant or useful.

No stats claim to account for everything. This one measures - across every game - how many points will on average result from a shot at that location.

From that point you can analyse the above factors to determine why you over performed or under performed, if you want. 'Oh we overperformed' 'well we kept getting kicks in the pocket but they were all Rankine so his individual xScore is much higher' 'Fog was shanking everything and scored well below' 'All his shots came late in quarters and so maybe we need to look at his rotations?' - you won't get that from basic stats. Otherwise why would we have coaches?

This is basic data analytics stuff. Nothing tells you everything, you will always have to apply some skill to get what you need.

The xScore may actually be lower as the data set available does not include shots that don't score. Which is why this one didn't make it in but was hilarious:

 

[baldandbroke]

This is basic data analytics stuff. Nothing tells you everything, you will always have to apply some skill to get what you need.

And there it is. Apply some “skill” to get what you need.

Lies, damned lies, and statistics.

 

[MarcusP2]

And there it is. Apply some “skill” to get what you need.

Lies, damned lies, and statistics.

You're right, the eye test is the only objective way to assess.

 

[jamesclean]

I’m just going to go ahead and lower my EXPECTED IQ from the average bigfooty poster…

 

[Interloperer]

How do the stats measure the difficulty of shots taken?

If its the xscore described in the book Footballistics (people have developed their own) there are ~6 different types of shot (set,snap,pressured,etc...) + angle and distance that are considered.

The xScore may actually be lower as the data set available does not include shots that don't score. Which is why this one didn't make it in but was hilarious:

I think the guy that runs that account has some non-scoring shot data, but it's incomplete. Maybe inferred from other stuff?

Champion data have been sending their official expected score to Chip Le Grand recently for some reason, probably in reaction to fan created xscore. But it doesn't disagree with other xscores by more than a few points normally.

 

[Hobbes]

Darren Jarman is pretty tough to ignore.
Those finals performances are hard to better.
Yes Dawson is a great recruit.

Difficult to discount McLeod too.

 

[MRB37]

I agree with a lot of the sentiments above after watching the replay. Whilst Brisbane supposedly dominated I50's in the latter part of the 2nd qtr and for much of the 3rd quarter it didn't feel like they had a real ascendancy, and I felt we were still right in the game. Really pleased with the way the younger players such as Rachele stood up too. All in all, a pretty darn good team effort across the board. We'll miss Laird this week though...anyone in the SNAFL ready? Otherwise Smith / Hinge come in and we rotate a few through there to cover.

On the umpiring, however, staggering to me to see the missed mark for Hipwood, the hold on Tex wrongly paid to his opponent with four umpires on the field. I don't understand why we have extra umps if these sorts of decisions can't be challenged by another ump in a better position than the decision making ump. It wasn't just these decisions either, there were a bunch of mystifying ones - as there are most weeks.

Lions were usually leaving 3 of Andrews, Starcevich, Lester and Rich outside their F50 while we kept only 2 of Thilthorpe, Fogarty and Walker out, so they had set up pretty well for repeat entries, but most of 'em weren't really quality entries. I think Thilthorpe getting hot early really threw them off because they had to switch Andrews onto him and he's usually the one they want being free.

 

[NTRabbit]

Lions spent the first quarter trying to fight instead of play football, then the next two quarters turning goals into points, never going to win like that.

Defence was good in general, but some concerns over sloppiness 5m from goal and less, not enough expertise in killing the ball dead.

 

[Ham]

I don’t know if I’ve only just started to notice it - but I liked seeing ROB attack a lot more centre bounces from the side of the circle, take body contact and drop it down to his feet. There were a couple of really clean ones last weekend. Reminded me abit of what the Geelong rucks did to him.


On iPhone using BigFooty.com mobile app

 

[cleanhands]

I don’t know if I’ve only just started to notice it - but I liked seeing ROB attack a lot more centre bounces from the side of the circle, take body contact and drop it down to his feet. There were a couple of really clean ones last weekend. Reminded me abit of what the Geelong rucks did to him.


On iPhone using BigFooty.com mobile app

I’d prefer to see him tap it to one of our onballers