Analysis Crows Player Ratings

[Interloperer]

The AFL Player Ratings aren't easily available currently. And the official Player Rating twitter account isn't running this year.

However the data is still available via assorted public API. So I've taken the liberty of collating the rating's of Crows players here. Sorted by Average score, "Fantasy Points" in brackets for comparison.

I used the FitzRoy R package to grab this. Normally I pull from Fryzigg's data source as I find that a lot cleaner, but thats been a bit slow to update lately so I just used the actual AFL data. I'll try to keep it updated as the season progresses for information's sake.



(Data going in an image while BF sorts itself out)

 

[adelaidecrows]

The AFL Player Ratings aren't easily available currently. And the official Player Rating twitter account isn't running this year.

However the data is still available via assorted public API. So I've taken the liberty of collating the rating's of Crows players here. Sorted by Average score, "Fantasy Points" in brackets for comparison.

I used the FitzRoy R package to grab this. Normally I pull from Fryzigg's data source as I find that a lot cleaner, but thats been a bit slow to update lately so I just used the actual AFL data. I'll try to keep it updated as the season progresses for information's sake.

View attachment 1644521

(Data going in an image while BF sorts itself out)

Thanks.

No real surprises.

 

[WetCrow1991]

Take these 2 out and we are probably 1-1.

 

[Golumless]

The AFL Player Ratings aren't easily available currently. And the official Player Rating twitter account isn't running this year.

However the data is still available via assorted public API. So I've taken the liberty of collating the rating's of Crows players here. Sorted by Average score, "Fantasy Points" in brackets for comparison.

I used the FitzRoy R package to grab this. Normally I pull from Fryzigg's data source as I find that a lot cleaner, but thats been a bit slow to update lately so I just used the actual AFL data. I'll try to keep it updated as the season progresses for information's sake.

View attachment 1644521

(Data going in an image while BF sorts itself out)

It's kind of intriguing how Murphy and Thilthorpe are top 5. Surely it's biasing towards certain stats.

 

[Carmo]

The AFL Player Ratings aren't easily available currently. And the official Player Rating twitter account isn't running this year.

However the data is still available via assorted public API. So I've taken the liberty of collating the rating's of Crows players here. Sorted by Average score, "Fantasy Points" in brackets for comparison.

I used the FitzRoy R package to grab this. Normally I pull from Fryzigg's data source as I find that a lot cleaner, but thats been a bit slow to update lately so I just used the actual AFL data. I'll try to keep it updated as the season progresses for information's sake.

View attachment 1644521

(Data going in an image while BF sorts itself out)

It would be interesting to track these ratings versus coaches votes for every round through the year and see if they correlate much.

 

[Interloperer]

It's kind of intriguing how Murphy and Thilthorpe are top 5. Surely it's biasing towards certain stats.

I think largely it's because the team played much worse in round 1 and almost everybody scored poorly, dragging down averages. Murphy and Thilthorpe only played in the higher rating game.

That said, they did play well at times in the second half, and do things that led to goals. I find the main difference between Player Ratings and the Fantasy Score you see on the AFL website, is that the player ratings punish you much more for poor play. Fantasy Score doesn't differentiate between a "quiet game" and an "actively bad game".

The player ratings twitter that broke it down quarter by quarter makes this score much more useful IMHO.

It would be interesting to track these ratings versus coaches votes for every round through the year and see if they correlate much.

I did a project where I used these ratings to pick an All-Australian team and they correlate with the Brownlow very well, with the exception that it favours Key Forwards and Rucks as much as midfielders, and really hates shut down defenders. The only anomaly was Cripps, who was significantly down the list.

However, I do generally think that over the course of several matches this is the least worst statistic for comparing players that is easily available

 

[Kane McGoodwin]

The AFL Player Ratings aren't easily available currently. And the official Player Rating twitter account isn't running this year.

However the data is still available via assorted public API. So I've taken the liberty of collating the rating's of Crows players here. Sorted by Average score, "Fantasy Points" in brackets for comparison.

I used the FitzRoy R package to grab this. Normally I pull from Fryzigg's data source as I find that a lot cleaner, but thats been a bit slow to update lately so I just used the actual AFL data. I'll try to keep it updated as the season progresses for information's sake.

View attachment 1644521

(Data going in an image while BF sorts itself out)

dogs105

 

[dogs105]

dogs105

Stats have Hinge as only 14 worst in round 2
I thought he made too many mistakes. I have only watched it once though. I may have missed some good stuff.

They have him 2nd worst after Brown in Rd 1?

 

[Kane McGoodwin]

Stats have Hinge as only 14 worst in round 2
I thought he made too many mistakes. I have only watched it once though. I may have missed some good stuff.

They have him 2nd worst after Brown in Rd 1?

All but 1 player (Berry) I named off the top of my head are ranked worse.

No doubt Hinge needs to play better, but he's not alone as same can be said for nearly the whole side!

 

[dogs105]

All but 1 player (Berry) I named off the top of my head are ranked worse.

No doubt Hinge needs to play better, but he's not alone as same can be said for nearly the whole side!

He seems to have been rated much higher than I did for round 2.
I hope his use of the ball improves, because he looked very good in the preseason. Just cutting out the really stupid stuff would be a great start.

And he's only a young kid still - just a few months older than Murphy. I wouldn't have dropped him yet, but he does need to lift

 

[Kane McGoodwin]

He seems to have been rated much higher than I did for round 2.
I hope his use of the ball improves, because he looked very good in the preseason. Just cutting out the really stupid stuff would be a great start.

He had to go back & player on taller players. Was on Riewoldt for a while I think. Thought he did OK.

Sloane & Smith on the other hand...

 

[dogs105]

He had to go back & player on taller players. Was on Riewoldt for a while I think. Thought he did OK.

Sloane & Smith on the other hand...

He does some good things, but then has some awful brain fades. His consistency was much better in the preseason games, so hopefully he returns to that.

And he's still young.

 

[Interloperer]

Tried an experiment to see if I could get quarter by quarter Match Ratings by pulling data at each break. Very disappointed or pleased to inform you that all Crows players have turned in a rating of NaN ('Not a Number') in the first half. They need to keep it up or pull their head in.

 

[Interloperer]

After Round 3

 

[Interloperer]

After Round 4

 

[WeeBlake]

Great work, keep going

 

[Interloperer]

After Round 5. A few surprises this week. Some players hit hard by the algorithm, others seemed to be given a pass for their errors.

If I get time this week I'll tidy things up and add some more info.

 

[Interloperer]

Round Awards (WIP)
Pure Class Ratio: Nick Murray (0.41)
Quantity Over Quality: Riley Thilthorpe (0.00)

 

[Interloperer]

This week was a great game to demonstrate the utility of the player ratings. A lot of players accumulated a lot of ball, but had horrible scrappy games and their player rating reflects this where the AFL fantasy points do not.

I also added a summary of the latest game (including opposition players) and some player awards. I have one more dumb gimmick planned that should be ready soon.

 

[Golumless]

View attachment 1668078

View attachment 1668087

Round Awards (WIP)
Pure Class Ratio: Nick Murray (0.41)
Quantity Over Quality: Riley Thilthorpe (0.00)

Chayce Jones being the eight best Adelaide player this week kind of stands out as a tad confusing, but I imagine that's probably due to him not wasting a disposal in a very scrappy game.

 

[dogs105]

Chayce Jones being the eight best Adelaide player this week kind of stands out as a tad confusing, but I imagine that's probably due to him not wasting a disposal in a very scrappy game.

Very efficient.

 

[Interloperer]

Chayce Jones being the eight best Adelaide player this week kind of stands out as a tad confusing, but I imagine that's probably due to him not wasting a disposal in a very scrappy game.

Someone on the AFL subreddit recently was kind enough to provide the PhD thesis for the rating system that became the AFL player ratings. It is 257 pages, but there are 3 good examples of how it works in the executive summary right at the start.

The tl;dr version - The ratings work around an "expected next score" (a scale from '+6: certain goal scored' to '-6: certain goal for opposition') and if you do something that increases that number then your ranking go up, if you do something that makes it go down so does your ranking.

We don't have access to the raw data. But in the case of Jones, he had 100% disposal efficiency (5 of 9 contested) and 3 of his disposals were inside 50s, so I'm guessing all his disposals were very damaging.

 

[spexau]

Chayce Jones being the eight best Adelaide player this week kind of stands out as a tad confusing, but I imagine that's probably due to him not wasting a disposal in a very scrappy game.

He setup a couple goals iirc?

 

[FR0GGY]

Was Michalanney on Bruest all game?
Is that another scalp for him?

 

[dogs105]

Was Michalanney on Bruest all game?
Is that another scalp for him?

9 tackles, the lad!