The Statistics Thread

[BEaston]

It's just based on the last 40 games regardless of position I believe. The graphic below is from when AFL Player Ratings first launched so ignore the "last two seasons" as the initial ratings were based off the 2010-2011 seasons.

View attachment 1098134

Yeah, that's what I mean. His ranking points gained as a forward aren't really relevant to him as a backman.

 

[Boyd oh Boyd]

As of right now here's the average difference between us and our opponents

• Disposals: +77.6 (1st)
• Kicks: +7.2 (8th)
• Handballs: +70.4 (1st)
• Marks: -16.6 (16th)
• Goals: +6.8 (1st)
• Behinds: +5.6 (2nd)
• Inside 50s: +19.2 (1st)
• Tackles: +5.6 (4th)
• Frees: +4.4 (2nd)
• Clearances: +7.6 (1st)
• Centre clearances: +2.8 (1st)
• Stoppage clearances: +4.8 (1st)
• Clangers: -8.4 (17th)
• Rebound 50s: -13.4 (18th)
• Hitouts: -8.8 (14th)
• Contested possesions: +23.8 (1st)
• Uncontested possessions: +52.4 (2nd)
• Effective disposals: +72.0 (1st)
• Contested marks: -0.8 (12th)
• Marks inside 50: +7 (2nd)
• 1%: -0.4 (11th)
• Bounces: +5.6 (1st)
• Meters gained: +695.8 (1st)
• Turnovers: -2.8 (16th)
• Intercepts: +2.8 (4th)
• Tackles inside 50: +5.2 (1st)

I've bolded the ones I consider to be really important. We're owning the footy by a pretty big margin but despite that we're still out-tackling our opponents. We're pumping the ball inside 50, they're rebounding it a fair bit but it's not getting them a lot of meters gained. Our one problem is we're letting them mark a lot, and losing the contested marks. We're actually taking a heap of CM, 2nd most in the league, we're just not very good at stopping them doing it as well. Lucky for us Richmond and Port aren't marking sides. The Eagles are the biggest contender that could use this to hurt us and we're already proved we can beat them despite this so not all bad news there. The Demons are the other team to watch here.

This will be one I'll revisit at the bye and see if we're brought a bit closer to the pack but for now we're doing well in a lot of different areas.

 

[dogwatch]

As of right now here's the average difference between us and our opponents

• Disposals: +77.6 (1st)
• Kicks: +7.2 (8th)
• Handballs: +70.4 (1st)
• Marks: -16.6 (16th)
• Goals: +6.8 (1st)
• Behinds: +5.6 (2nd)
• Inside 50s: +19.2 (1st)
• Tackles: +5.6 (4th)
• Frees: +4.4 (2nd)
• Clearances: +7.6 (1st)
• Centre clearances: +2.8 (1st)
• Stoppage clearances: +4.8 (1st)
• Clangers: -8.4 (17th)
• Rebound 50s: -13.4 (18th)
• Hitouts: -8.8 (14th)
• Contested possesions: +23.8 (1st)
• Uncontested possessions: +52.4 (2nd)
• Effective disposals: +72.0 (1st)
• Contested marks: -0.8 (12th)
• Marks inside 50: +7 (2nd)
• 1%: -0.4 (11th)
• Bounces: +5.6 (1st)
• Meters gained: +695.8 (1st)
• Turnovers: -2.8 (16th)
• Intercepts: +2.8 (4th)
• Tackles inside 50: +5.2 (1st)

I've bolded the ones I consider to be really important. We're owning the footy by a pretty big margin but despite that we're still out-tackling our opponents. We're pumping the ball inside 50, they're rebounding it a fair bit but it's not getting them a lot of meters gained. Our one problem is we're letting them mark a lot, and losing the contested marks. We're actually taking a heap of CM, 2nd most in the league, we're just not very good at stopping them doing it as well. Lucky for us Richmond and Port aren't marking sides. The Eagles are the biggest contender that could use this to hurt us and we're already proved we can beat them despite this so not all bad news there. The Demons are the other team to watch here.

This will be one I'll revisit at the bye and see if we're brought a bit closer to the pack but for now we're doing well in a lot of different areas.

Very impressive stats.

I wouldn't worry about the marking. Gold Coast spent the first half chipping the ball sideways and gaining no metres. Those sort of kicks are junk stats and they distorted the picture of what's going on.

CMs - as you say we are taking lots but so are our opponents. The West Coast factor might have skewed that a bit. My guess is we are about middle of the league or just above on that.

Also being 18th on rebounds may be just a reflection of how many times we enter the 50 and how few times we concede f50 entries to the opposition. The eye test certainly tells us we are going OK.

However for all that ...

Past performance does not guarantee future results. We have to make sure we are switched on. If not the stats over the first 5 rounds will mean nothing.

 

[BEaston]

As of right now here's the average difference between us and our opponents

• Disposals: +77.6 (1st)
• Kicks: +7.2 (8th)
• Handballs: +70.4 (1st)
• Marks: -16.6 (16th)
• Goals: +6.8 (1st)
• Behinds: +5.6 (2nd)
• Inside 50s: +19.2 (1st)
• Tackles: +5.6 (4th)
• Frees: +4.4 (2nd)
• Clearances: +7.6 (1st)
• Centre clearances: +2.8 (1st)
• Stoppage clearances: +4.8 (1st)
• Clangers: -8.4 (17th)
• Rebound 50s: -13.4 (18th)
• Hitouts: -8.8 (14th)
• Contested possesions: +23.8 (1st)
• Uncontested possessions: +52.4 (2nd)
• Effective disposals: +72.0 (1st)
• Contested marks: -0.8 (12th)
• Marks inside 50: +7 (2nd)
• 1%: -0.4 (11th)
• Bounces: +5.6 (1st)
• Meters gained: +695.8 (1st)
• Turnovers: -2.8 (16th)
• Intercepts: +2.8 (4th)
• Tackles inside 50: +5.2 (1st)

I've bolded the ones I consider to be really important. We're owning the footy by a pretty big margin but despite that we're still out-tackling our opponents. We're pumping the ball inside 50, they're rebounding it a fair bit but it's not getting them a lot of meters gained. Our one problem is we're letting them mark a lot, and losing the contested marks. We're actually taking a heap of CM, 2nd most in the league, we're just not very good at stopping them doing it as well. Lucky for us Richmond and Port aren't marking sides. The Eagles are the biggest contender that could use this to hurt us and we're already proved we can beat them despite this so not all bad news there. The Demons are the other team to watch here.

This will be one I'll revisit at the bye and see if we're brought a bit closer to the pack but for now we're doing well in a lot of different areas.

It's a bloody impressive listing.

18th in r50 differential is a good position to be in, for the reasons dogwatch mentioned (more i50s to us giving more opportunity to opposition to have r50s combined with denial of oppo i50s and therefore less opportunity for us to r50). Not worried about that stat at all.

If we're considering forward line efficiency, scores per i50 is a much better measure which we're currently sitting 1st in the league in that metric.

We can also flip this to look at opposition scores per i50 to determine at "back line efficiency" (for lack of a better term) to look at who has the stingiest defence. We sit 5th in this measure.

These measures can then be combined to find a "net efficiency rating" similar to metrics that are commonly used in basketball.

Interesting that the Dees come out number 1 in this measure based on their uber stingy defence which is way out in front of the league at 0.31 scores conceded per i50. Large gap to next best at 0.38 per i50 (Pies).

We sit at number two overall with the best attacking efficiency and a top 5 defensive efficiency.

Interesting to note that Magpies, Giants and Lions all have top 5 defensive efficiency ratings but all have bottom 5 attacking efficiencies with poor to middling results so far. They will need to rebalance I would think to have success this year. Also goes some way to explaining our grind it out type wins we had against the Pies and Lions.

Crows, Eagles and Swans have the opposite problem while the Dockers look to be quite a balanced side.

 

[BEaston]

Further to the above, that the teams with stronger attacking efficiency (Crows, Eagles, Swans) are all much better placed on the ladder than those with the opposite problem. Hopefully a good sign for those that worry about the "state of the game" !

 

[X_box_X]

Ladder sorted by quarters and halves:

1ST QTR:


2ND QTR:


3RD QTR:


4TH QTR:


1ST HALF:


2ND HALF:

 

[Boyd oh Boyd]

Something I just realised. I love looking through Statspro to check specialist stats and what level they're average, elite, etc. I just checked Rhylee West

14

Rhylee West

Age

23

Ht

183cm

Wt

82kg

Pos.

Fwd

Career

Season

Last 5

D

12.3

3star

K

5.7

2star

HB

6.7

4star

M

2.5

3star

T

1.5

3star

G

0.5

3star

D

12.0

3star

K

6.0

2star

HB

6.0

3star

M

2.7

3star

T

2.0

3star

G

0.3

3star

D

13.0

4star

K

5.6

3star

HB

7.4

5star

M

2.2

3star

T

1.4

4star

G

0.6

3star

's page and realised they have him averaging 10.8 disposals a game in his career but that includes his one 'game' this year where he was a sub with 0% time on ground. It's a minor inconvenience but for people like me who love to use this stuff it really *s the whole thing up. I can correct the averages but the whole rating scale of average/above average/elite is thrown out of wack. Not only for the sub player but every other player in the league, as the average for every position will be pushed down by 10+ subs a week who don't play so the players who do get a game get an inflated rating.

I know most people won't care but I like this stuff so I'm disappointed.

 

[BEaston]

Something I just realised. I love looking through Statspro to check specialist stats and what level they're average, elite, etc. I just checked Rhylee West's page and realised they have him averaging 10.8 disposals a game in his career but that includes his one 'game' this year where he was a sub with 0% time on ground. It's a minor inconvenience but for people like me who love to use this stuff it really fu**s the whole thing up. I can correct the averages but the whole rating scale of average/above average/elite is thrown out of wack. Not only for the sub player but every other player in the league, as the average for every position will be pushed down by 10+ subs a week who don't play so the players who do get a game get an inflated rating.

I know most people won't care but I like this stuff so I'm disappointed.

Just another reason the sub sucks.

 

[Oliver G]

Something I just realised. I love looking through Statspro to check specialist stats and what level they're average, elite, etc. I just checked Rhylee West's page and realised they have him averaging 10.8 disposals a game in his career but that includes his one 'game' this year where he was a sub with 0% time on ground. It's a minor inconvenience but for people like me who love to use this stuff it really fu**s the whole thing up. I can correct the averages but the whole rating scale of average/above average/elite is thrown out of wack. Not only for the sub player but every other player in the league, as the average for every position will be pushed down by 10+ subs a week who don't play so the players who do get a game get an inflated rating.

I know most people won't care but I like this stuff so I'm disappointed.

The benchmarking on StatsPro has always been strange and uses a different formula to what Champion Data actually use to calculate ratings which...



As for counting games as played for an unplayed sub, this has always been the case even going back to the 19th and 20th man.

 

[Libba39]

As of right now here's the average difference between us and our opponents

• Disposals: +77.6 (1st)
• Kicks: +7.2 (8th)
• Handballs: +70.4 (1st)
• Marks: -16.6 (16th)
• Goals: +6.8 (1st)
• Behinds: +5.6 (2nd)
• Inside 50s: +19.2 (1st)
• Tackles: +5.6 (4th)
• Frees: +4.4 (2nd)
• Clearances: +7.6 (1st)
• Centre clearances: +2.8 (1st)
• Stoppage clearances: +4.8 (1st)
• Clangers: -8.4 (17th)
• Rebound 50s: -13.4 (18th)
• Hitouts: -8.8 (14th)
• Contested possesions: +23.8 (1st)
• Uncontested possessions: +52.4 (2nd)
• Effective disposals: +72.0 (1st)
• Contested marks: -0.8 (12th)
• Marks inside 50: +7 (2nd)
• 1%: -0.4 (11th)
• Bounces: +5.6 (1st)
• Meters gained: +695.8 (1st)
• Turnovers: -2.8 (16th)
• Intercepts: +2.8 (4th)
• Tackles inside 50: +5.2 (1st)

I've bolded the ones I consider to be really important. We're owning the footy by a pretty big margin but despite that we're still out-tackling our opponents. We're pumping the ball inside 50, they're rebounding it a fair bit but it's not getting them a lot of meters gained. Our one problem is we're letting them mark a lot, and losing the contested marks. We're actually taking a heap of CM, 2nd most in the league, we're just not very good at stopping them doing it as well. Lucky for us Richmond and Port aren't marking sides. The Eagles are the biggest contender that could use this to hurt us and we're already proved we can beat them despite this so not all bad news there. The Demons are the other team to watch here.

This will be one I'll revisit at the bye and see if we're brought a bit closer to the pack but for now we're doing well in a lot of different areas.

Stop the presses!!!

Are you telling me we are ONLY SECOND for free kick differential!!!!????

Who is this usurper who dares take our mantle???

 

[Boyd oh Boyd]

The benchmarking on StatsPro has always been strange and uses a different formula to what Champion Data actually use to calculate ratings which...

View attachment 1112349

As for counting games as played for an unplayed sub, this has always been the case even going back to the 19th and 20th man.

It's not really the counting of games that annoys me, it's how it messes with the averages and rankings of other players in comparison. I'd love to just use Champion Data but I have to work with what I can access.

That actually pretty much aligns with Statspro ranking of Rhylee so good that both seem to be on similar wavelengths with different formulas. You'd know better than anyone - with the CD rankings has there been any significant change in what level you need to reach to be considered say average in disposals? Just having a quick look it seems to be somewhere between 12 and 14 as a forward for Statspro, is that the same as Champion Data and is that different to the last few years? In my mind there'd have to be if they're counting players who get 0 disposals as a sub every week but maybe CD are correcting for that or just not a significant amount of players?

 

[threenewpadlocks]

Something I just realised. I love looking through Statspro to check specialist stats and what level they're average, elite, etc. I just checked Rhylee West's page and realised they have him averaging 10.8 disposals a game in his career but that includes his one 'game' this year where he was a sub with 0% time on ground. It's a minor inconvenience but for people like me who love to use this stuff it really fu**s the whole thing up. I can correct the averages but the whole rating scale of average/above average/elite is thrown out of wack. Not only for the sub player but every other player in the league, as the average for every position will be pushed down by 10+ subs a week who don't play so the players who do get a game get an inflated rating.

I know most people won't care but I like this stuff so I'm disappointed.

For historical purposes, blokes who never came off the bench in the old days (like the 19th and 20th men in our '54 flag) have a game counted. If the player is ready to play (Tom Lynch shenanigans aside), it should rightfully count as a game - should players miss out on the various qualifications like the AFL 200 club, Father/Son, etc. because they were the sub? Of course not.

So then, by strict definition of a disposals per game average, it counts the substitute game.

Might be annoying for you and me but as seen by the above CD dashboard and obviously the coaches that use the stats, they're obviously not silly enough to fail to calculate with that sub game.

 

[Werewolf]

The sub shouldn't be recorded as having played a game unless they take to the field.

 

[Mutt]

 

[VL3OJ]

 

[Oliver G]

One positive tonight was Naughton's accuracy, 3 goals from 4 shots. The one miss was horrendous though.

 

[BEaston]

It's a bloody impressive listing.

18th in r50 differential is a good position to be in, for the reasons dogwatch mentioned (more i50s to us giving more opportunity to opposition to have r50s combined with denial of oppo i50s and therefore less opportunity for us to r50). Not worried about that stat at all.

If we're considering forward line efficiency, scores per i50 is a much better measure which we're currently sitting 1st in the league in that metric.

We can also flip this to look at opposition scores per i50 to determine at "back line efficiency" (for lack of a better term) to look at who has the stingiest defence. We sit 5th in this measure.

These measures can then be combined to find a "net efficiency rating" similar to metrics that are commonly used in basketball.

Interesting that the Dees come out number 1 in this measure based on their uber stingy defence which is way out in front of the league at 0.31 scores conceded per i50. Large gap to next best at 0.38 per i50 (Pies).

We sit at number two overall with the best attacking efficiency and a top 5 defensive efficiency.

Interesting to note that Magpies, Giants and Lions all have top 5 defensive efficiency ratings but all have bottom 5 attacking efficiencies with poor to middling results so far. They will need to rebalance I would think to have success this year. Also goes some way to explaining our grind it out type wins we had against the Pies and Lions.

Crows, Eagles and Swans have the opposite problem while the Dockers look to be quite a balanced side.

View attachment 1108828

Wanted a 20 minute distraction so thought I'd look at this again after round 8.

This time I included a midfield measure as well to cover all 3 'realms' of the field.

Defence - opposition scores per i50 conceded
Attack - scores per i50
Midfield - ratio of i50s to opposition i50s

I then standardised these measures to find the z-score on each. This is just a statistical method to enable comparison of different measures. Turns everything into apples (or oranges if you prefer). These z-scores can then be summed to find the overall rating of a team. I then did a comparison to current ladder position just as a point of interest.



Note: z-score usually results in roughly ±2. A result of 0 would mean exactly neutral. Anything in the ±3 range is extremely high/low, an anomaly compared to the data set. (This applies to the z-scores for atk, def, mid only, the Total column is a sum which is not a true z-score)

Anyway, I would say this measure roughly correlates to ladder position, with a few outliers notably Carlton, Sydney, Saints, Dons. Those outliers could be due to over/under performance of opposition, or strength of fixture so far, or just variance in chance within those games.

As you'd probably expect, we sit well in attack and midfield being ranked the highest in both areas. Our defence is slightly below par.
Demons are the only team in the league without a negative z-score in any of the 3 areas, however the strength of their team is heavily tilted towards their defence which is ridiculously strong.
Us and Demons are well ahead of the league in total z-sore, which backs up both the eye test and W/L column.
Geelong, Lions, Power are in the next rung down.
Richmond, despite being capable to play at a very high level (as we saw last week), overall have been inconsistent and this is shown in their Total rating.

Anyway, somewhat interesting to look at. Might update again further along the season.

 

[Boristown]

Wanted a 20 minute distraction so thought I'd look at this again after round 8.

This time I included a midfield measure as well to cover all 3 'realms' of the field.

Defence - opposition scores per i50 conceded
Attack - scores per i50
Midfield - ratio of i50s to opposition i50s

I then standardised these measures to find the z-score on each. This is just a statistical method to enable comparison of different measures. Turns everything into apples (or oranges if you prefer). These z-scores can then be summed to find the overall rating of a team. I then did a comparison to current ladder position just as a point of interest.

View attachment 1124391

Note: z-score usually results in roughly ±2. A result of 0 would mean exactly neutral. Anything in the ±3 range is extremely high/low, an anomaly compared to the data set. (This applies to the z-scores for atk, def, mid only, the Total column is a sum which is not a true z-score)

Anyway, I would say this measure roughly correlates to ladder position, with a few outliers notably Carlton, Sydney, Saints, Dons. Those outliers could be due to over/under performance of opposition, or strength of fixture so far, or just variance in chance within those games.

As you'd probably expect, we sit well in attack and midfield being ranked the highest in both areas. Our defence is slightly below par.
Demons are the only team in the league without a negative z-score in any of the 3 areas, however the strength of their team is heavily tilted towards their defence which is ridiculously strong.
Us and Demons are well ahead of the league in total z-sore, which backs up both the eye test and W/L column.
Geelong, Lions, Power are in the next rung down.
Richmond, despite being capable to play at a very high level (as we saw last week), overall have been inconsistent and this is shown in their Total rating.

Anyway, somewhat interesting to look at. Might update again further along the season.

Really interesting stuff. It's going to take an incredible effort to break down Melbourne's defence, especially at the MCG.

 

[X_box_X]

Our third quarters haven't been impressive. We've won 2/8. In contrast, Port are 7-1 in third quarters.



It was last year's third quarter that decided the result of our match against Port. They went nuts, kicking the only four goals of the quarter. If we are to beat Port, we can't have any lapse in concentration like last year. Maybe you can get away with having a momentary lapse against a bottom eight team, but doing that against a contender is going tobe costly.

 

[Oliver G]

Most accurate shot for goal (35+ shots)...

 

[ScottyDogg]

Our third quarters haven't been impressive. We've won 2/8. In contrast, Port are 7-1 in third quarters.

View attachment 1124422

It was last year's third quarter that decided the result of our match against Port. They went nuts, kicking the only four goals of the quarter. If we are to beat Port, we can't have any lapse in concentration like last year. Maybe you can get away with having a momentary lapse against a bottom eight team, but doing that against a contender is going tobe costly.

View attachment 1124423

Really interesting stuff. What do we think is behind our third quarter lethargy?

 

[X_box_X]

Really interesting stuff. What do we think is behind our third quarter lethargy?

I'm not sure, but it's an ongoing trend.

Last year our best quarters were also our 4th (12-5 WL) and 1st quarters (11-6 WL). We were 8-8-1 in both the second and third quarters. This doesn't include our EF loss to the Saints where the two quarters we won were the 1st and 4th quarters.

This year it's hard to argue against our first quarters (8-0 WL) being our strongest performing quarters, even though our fourth quarters (5-3 WL) have produced our best F+A percentage. In the five fourth quarters we have won, we have smashed the opposition. Smashed them. This is evidenced by our whopping percentage of 210!

I also recall at the end of 2019 we were really hot in second halves. I think prior to losing to GWS in the EF, we had won our previous eight second halves.

 

[dogwatch]

Really interesting stuff. What do we think is behind our third quarter lethargy?

They should stop passing the bong around during the HT break.

 

[bobs head soup]

Really interesting stuff. What do we think is behind our third quarter lethargy?

Part of the reason I reckon is that Bevo will often preserve our best players in their best positions for Q4. That means giving the 2nd Ruck and others extra minutes in the 3rd.

Depending on the match situation, he will sometimes try to slow the game down and play a bit of tempo whilst the 'b grade'' combos are rolling.

 

[ScottyDogg]

Whether it be bongs or preserving players, I’m glad to hear that it’s within our control.