Pro Stats Player Rankings -- after ten rounds

[NMWBloods]

Originally posted by Vivien
Out of interest, who were the top ranking players?

L. Gilbee 74.4
P. Everitt 74.4
B. Cousins 74.0
S. Mitchell 72.4
N. Brown 72.2
L. Hodge 70.7
B. Johnson 67.0
D. Cox 66.9
S. Crawford 65.9
R. Murphy 64.2

[barry]

Originally posted by NMWBloods

Fourthly, resorting to calling people nerds because they understand numbers is really very tragic and more than anything else illustrates a lack of understanding.

- By your own words, you dont fully understand the ratings.
- A non-nerd can understand statistics. A nerd submerges himself in them.

[NMWBloods]

Originally posted by barry
- By your own words, you dont fully understand the ratings.

As usual twists the words - I don't know how they calculate that particular stat.

- A non-nerd can understand statistics. A nerd submerges himself in them.

Ah, such brilliant insight...

[barry]

Originally posted by Charlie
Play the ball, thanks folks. There's simply no need for name-calling.

I was name-calling the people who came up with the stats in the first place, not anyone here.

Is mild name calling outsiders acceptable?

[Sean]

Originally posted by NMWBloods
L. Gilbee 74.4
P. Everitt 74.4
B. Cousins 74.0
S. Mitchell 72.4
N. Brown 72.2
L. Hodge 70.7
B. Johnson 67.0
D. Cox 66.9
S. Crawford 65.9
R. Murphy 64.2

I thought that those stats looked quite accurate for our players but looking at that maybe they aren't. 4 Hawks & 3 Bulldogs in the top 10 seems to suggest that they aren't quite right - and surely Crawford wouldn't be anywhere near the best 10 players this season.

[Ruck'n'Roll]

Originally posted by NMWBloods
Isn't Rocca stilll injured? Cloke should be a good matchup for LRT.

Rocca's got 2 weeks, also I wonder whether Saddington will have a shot at Tarrent, he once did quite well against him. I'd really like to see Leo stay away for the last line.

[liz]

Originally posted by Sean
I thought that those stats looked quite accurate for our players but looking at that maybe they aren't. 4 Hawks & 3 Bulldogs in the top 10 seems to suggest that they aren't quite right - and surely Crawford wouldn't be anywhere near the best 10 players this season.

My thoughts exactly.

In a similar vein, one of the journos (can't remember who but IIRC it was one of the better ones) published his list of the team of the season thus far last week. It included 5 Hawthorn players.

On the face of it, is was hard to argue with four of the Hawks inclusions - Mark Williams, Everitt, Hodge and Mitchell (Trent Croad was the fifth). One is leading the Coleman and the other three are, by common consensus of the media, having fantastic seasons.

Yet the Hawks are languishing in 15th spot and right now its hard to see where their next win will come from.

Is it just that these three are shining lights in an otherwise bleak outfit? Possibly, though Crawford and Bateman have also been playing well, Tim Clarke has improved significantly, Miller, Ladson, Brown and Osbourne have been "going OK", and they have had two RS nominations.

What this says to me is that many of the facets that are indicated by stats, and which are noticed by TV and media commentators, have very little to do with what actually wins a game of football.

Everitt was asked on WLF last night about Hodge's relatively quiet game at the weekend after averaging around 30 possessions a game for the last month or so. Everitt pointed out that Hodge had been tagged for pretty much the first time. This puts suggestions that Hodge is "catching up" with Judd as a top-line player somewhat into perspective. Wandering around as a spare man in defence and notching up possessions is a pretty easy thing to do for a half-decent footballer.

And while there is no doubt that Mitchell's ability to win the ball at stoppages is extraordinary, his ability to use the ball really constructively when he gets it must be questioned.

This isn't meant to be a Hawthorn bash - I'm just using them as an illustration of how stats don't really measure what wins games. I think they have some place in analysis of the game and certainly add a level of objectively that pure observation can't provide, but I reckon you still have to firstly trust what you see with your own eyes.*

* And I am a true "stats nerd", having an undergrad degree in maths!

[liz]

Originally posted by Ruckman
Rocca's got 2 weeks, also I wonder whether Saddington will have a shot at Tarrent, he once did quite well against him. I'd really like to see Leo stay away for the last line.

No way will Saddo be given the job on Tarrant. He hasn't played in defence for well over a year and can't see this changing now.

I'd say that Leo is a probability for Tarrant, while LRT will probably be given a tough assignment on the very very promising third Cloke.

Didak and Caracella are the real danger men for the Pies up forward at the moment and it's hard to see the match-ups here. Kennelly has played on Caracella before (against the Lions) so could well get the job but Didak is a tough one. I don't think B2 has the agility. Maybe Crouch or Bevan will get the job.

Dempster on Rusling seems a pretty obvious choice, assuming both play.

[NMWBloods]

Originally posted by Sean
I thought that those stats looked quite accurate for our players but looking at that maybe they aren't. 4 Hawks & 3 Bulldogs in the top 10 seems to suggest that they aren't quite right - and surely Crawford wouldn't be anywhere near the best 10 players this season.

I think all of the players in that top 10 are having great seasons, and this is generally recognised by most observers of the game.

Definitely though there is more to the game than stats, as I've noted many times before.

An easy simple example could be a fantastic passage of play involving four players and then the fifth turns it over or missed the goal or whatever. The four players will get good stats but there will be no goal.

This type of thing occurs when your leading half dozen or so players are great but your second tier players are below competition average.

There is a large degree of relevance in this for Hawthorn and the Bulldogs. Neither have a major presenting big key forward and both struggle in defence (certainly BUlldogs do anyway). Of those two sides, up to round 10 only one player in total had more than 20 goals.

Additionally you need to be very careful about possession stats these days as many teams play this keepings off style or they do dinky little kicks and handballs and get it back and can rack up 3 or 4 possessions very quickly. Hawthorn and the Bulldogs are 2 of the 3 leading possession teams in the competition.

You need to blend stats and observation. Either by itself is likely to produce misleading conclusions.

[Sean]

Originally posted by NMWBloods
I think all of the players in that top 10 are having great seasons, and this is generally recognised by most observers of the game.

True but so are Yze, Johnstone, Miller, Green, Bruce etc and yet the second place team has no-one in the top 10. It just seems as though the stats are missing an important ingredient. Players like Bartel (and about 9 other Cats midfielders) have also been better than most of those listed IMHO.

Additionally you need to be very careful about possession stats these days as many teams play this keepings off style or they do dinky little kicks and handballs and get it back and can rack up 3 or 4 possessions very quickly. Hawthorn and the Bulldogs are 2 of the 3 leading possession teams in the competition.

True. Looks like they need to start ignoring possessions that aren't actually useful.

[Wil]

Originally posted by Sean
True but so are Yze, Johnstone, Miller, Green, Bruce etc and yet the second place team has no-one in the top 10. It just seems as though the stats are missing an important ingredient.

Yes they are, they are missing your bias.

[Sean]

Originally posted by Wil
Yes they are, they are missing your bias.

I suggested that the best players in the team that is 2nd are better than the best players in the team coming 15th because I'm biased? What a bizarre comment.

[Wil]

Originally posted by Sean
I suggested that the best players in the team that is 2nd are better than the best players in the team coming 15th because I'm biased? What a bizarre comment.

There you go, that is your bias.

[NMWBloods]

Originally posted by Sean
True but so are Yze, Johnstone, Miller, Green, Bruce etc and yet the second place team has no-one in the top 10. It just seems as though the stats are missing an important ingredient. Players like Bartel (and about 9 other Cats midfielders) have also been better than most of those listed IMHO.

True. Looks like they need to start ignoring possessions that aren't actually useful.

This is why on some of these types of stats I don't put so much weighting on the actual number, but more the relativities. For example, much of the Geelong midfield and most of those Melbourne players score more than 50, and I noted before ">50 excellent" rather than paying too much attention to the actual score.

Another way of looking at it is to stratify each team into how many at each category. Additionally, some players, such as J Brown or Darcy, have high average scores but low overall scores as they haven't played many games.

You need to take into account the total team scores but who played and how many games, etc.

[barry]

Originally posted by liz
My thoughts exactly.
< snip >

What this says to me is that many of the facets that are indicated by stats, and which are noticed by TV and media commentators, have very little to do with what actually wins a game of football.

Everitt was asked on WLF last night about Hodge's relatively quiet game at the weekend after averaging around 30 possessions a game for the last month or so. Everitt pointed out that Hodge had been tagged for pretty much the first time. This puts suggestions that Hodge is "catching up" with Judd as a top-line player somewhat into perspective. Wandering around as a spare man in defence and notching up possessions is a pretty easy thing to do for a half-decent footballer.

< snip >

My thoughts exactly.

The only stats that matter is the ladder. Although I do allow myself one digression, and thats to rank the teams on percentage as a better indicator of where they are travelling. FWIW, This puts the swans down to 8th.