Comparing Sharks' Forward Output in 2021

[BurgoShark]

@bort posted some raw "average metres per game" for Sharks forwards in 2021 in the Player Movement Thread.

This got me thinking. You guys must be sick of me banging on about "metres" or "metres per game" being a poor stat for comparing forward output - so I came up with a slightly better way of looking at it. This is still flawed in many ways - but I would argue that this gives you more information and a better basis for comparison.

**
The old way: NRL.com average metres per game

This is a really simple stat, which takes total metres and divides it by games played. it doesn't factor in whether the player went off injured after a minute, what position he played, or whether his team had a fair share of the ball. Here is how Sharks players stacked up in 2021.

1. Willams / Rudolf - 120m
3. Wilton / Talakai - 119m
5. Woods - 111m
6. Tolman - 106m
7. Hamlin-Uele - 89m
8. Graham 88m
9. Nikora 80m

**
The new way: only count hit-ups - and factor in both minutes played and share of possession

Only counting hit-ups
You may think this is a little unfair. After all - the guy did actually run the ball for those metres. However, when you include kick return metres is massively skews a player's totals and averages, because often they will have a couple of 20+ metre runs in a game. Almost 1/4 of Toby Rudolf's total running metres for the season were kick returns, and other players hardly did it at all. Obviously that tells us that Toby Rudolf is the coach's preferred choice to return kicks, but it doesn't help us know who is doing more in the other 99% of the footy game. Kick returns are out.

Factoring in Minutes played
This one is a no-brainer. One bloke plays half a game for 100m and another bloke plays a full game for 200m. Obviously one may be fitter and be able to play longer stretches - which is again a coaching decision - but if you are comparing their "output in metres" while they are on the field, these two players are equivalent.

Factoring in share of possession
Something else to consider is which team has the ball. The NRL commentary team has told us for years that making more metres helps you win - but that's not really an accurate way of looking at it. In general, the team who has the greater share of possession makes more metres simply because they have the ball more often. To determine whether a forward is doing a good job of making metres, we have to look at his stats relative to how much possession his team had. You can't run the ball without first having the ball.

**

Introducing the Burgo-Metre (which is quite different to the Burgometer, a device for measuring your Burgo-ness).

A Burgo-Metre (Bm) is a score for how many metres the player would run, excluding kick returns, in 80 minutes of game time, assuming a 50/50 share of possession. The numbers below are a season average for each player.

Calibration
First off - I looked at a few well known workhorses to see what would be a good performance.

Payne Haas --> 230 Bm's per game
Paul Gallen in 2013 --> 203
Jason Taumalolo --> 195

Comparing the Sharks' Forwards
When we remove kick returns and factor in both minutes played and possession...

1. Williams - 179 Bm's per game
2. Talakai - 176
3. Woods - 175
4. Wilton - 154
5. Hamlin-Uele - 146
6. Rudolf - 143
7. McInnes (Dragons 2020) - 138
8. Tolman - 137
9. Finucane (Storm) - 137
10. Graham - 132
11. Nikora - 96

Just like normal metres, a score of 100 in a game is about average and anything above above 160-170 is quite good. In an individual game, for a player to play more than 40 minutes and score over 220 is very rare. It only happened 4 times for the Sharks in 2021. 3 of them were Aaron Woods and the 4th was Talakai. Williams scored over 200 in almost half his games, but only when he played limited minutes.

**

So after all of this, what did I learn? That unless he is returning a kick Toby Rudolf is about as effective as Aiden Tolman, that Aaron Woods was on the same level as Williams and Talakai in terms of go-forward, that Wade Graham is still rolling up his sleeves, that if metres is what you want you would pick Wilton over Nikora easily, and that Jack Williams is a weapon if you can keep him under 40 minutes of game time.

A lot of this I knew already, and tbh it probably could all have been shown by just excluding kick return stats and any games where guys went off injured early - but sometimes it can be interesting to learn this through investigation.

Again - this is a flawed model with lots of holes in it... but I reckon it gives us a better picture than the raw stats.

Thanks again @bort for sending me down this rabbit hole

 

[Sparkles]

@bort posted some raw "average metres per game" for Sharks forwards in 2021 in the Player Movement Thread.

This got me thinking. You guys must be sick of me banging on about "metres" or "metres per game" being a poor stat for comparing forward output - so I came up with a slightly better way of looking at it. This is still flawed in many ways - but I would argue that this gives you more information and a better basis for comparison.

**
The old way: NRL.com average metres per game

This is a really simple stat, which takes total metres and divides it by games played. it doesn't factor in whether the player went off injured after a minute, what position he played, or whether his team had a fair share of the ball. Here is how Sharks players stacked up in 2021.

1. Willams / Rudolf - 120m
3. Wilton / Talakai - 119m
5. Woods - 111m
6. Tolman - 106m
7. Hamlin-Uele - 89m
8. Graham 88m
9. Nikora 80m

**
The new way: only count hit-ups - and factor in both minutes played and share of possession

Only counting hit-ups
You may think this is a little unfair. After all - the guy did actually run the ball for those metres. However, when you include kick return metres is massively skews a player's totals and averages, because often they will have a couple of 20+ metre runs in a game. Almost 1/4 of Toby Rudolf's total running metres for the season were kick returns, and other players hardly did it at all. Obviously that tells us that Toby Rudolf is the coach's preferred choice to return kicks, but it doesn't help us know who is doing more in the other 99% of the footy game. Kick returns are out.

Factoring in Minutes played
This one is a no-brainer. One bloke plays half a game for 100m and another bloke plays a full game for 200m. Obviously one may be fitter and be able to play longer stretches - which is again a coaching decision - but if you are comparing their "output in metres" while they are on the field, these two players are equivalent.

Factoring in share of possession
Something else to consider is which team has the ball. The NRL commentary team has told us for years that making more metres helps you win - but that's not really an accurate way of looking at it. In general, the team who has the greater share of possession makes more metres simply because they have the ball more often. To determine whether a forward is doing a good job of making metres, we have to look at his stats relative to how much possession his team had. You can't run the ball without first having the ball.

**

Introducing the Burgo-Metre (which is quite different to the Burgometer, a device for measuring your Burgo-ness).

A Burgo-Metre (Bm) is a score for how many metres the player would run, excluding kick returns, in 80 minutes of game time, assuming a 50/50 share of possession. The numbers below are a season average for each player.

Calibration
First off - I looked at a few well known workhorses to see what would be a good performance.

Payne Haas --> 230 Bm's per game
Paul Gallen in 2013 --> 203
Jason Taumalolo --> 195

Comparing the Sharks' Forwards
When we remove kick returns and factor in both minutes played and possession...

1. Williams - 179 Bm's per game
2. Talakai - 176
3. Woods - 175
4. Wilton - 154
5. Hamlin-Uele - 146
6. Rudolf - 143
7. Tolman - 137
8. Graham - 132
9. Nikora - 96

Just like normal metres, a score of 100 in a game is about average and anything above above 160-170 is quite good. In an individual game, for a player to play more than 40 minutes and score over 220 is very rare. It only happened 4 times for the Sharks in 2021. 3 of them were Aaron Woods and the 4th was Talakai. Williams scored over 200 in almost half his games, but only when he played limited minutes.

**

So after all of this, what did I learn? That unless he is returning a kick Toby Rudolf is about as effective as Aiden Tolman, that Aaron Woods was on the same level as Williams and Talakai in terms of go-forward, that Wade Graham is still rolling up his sleeves, that if metres is what you want you would pick Wilton over Nikora easily, and that Jack Williams is a weapon if you can keep him under 40 minutes of game time.

A lot of this I knew already, and tbh it probably could all have been shown by just excluding kick return stats and any games where guys went off injured early - but sometimes it can be interesting to learn this through investigation.

Again - this is a flawed model with lots of holes in it... but I reckon it gives us a better picture than the raw stats.

Thanks again @bort for sending me down this rabbit hole

Realyl interesting Burgo, cheers. A bit depressing, but interesting...!

Don't suppose you'd want to slot Dale and Cam in there too, just for comparison? I'm really hoping we can find one or both are a step above Woods!

 

[stormshark]

Well done very thorough. A Metres conceded to opposition, via dropped balls, from individual forwards, may take the Sheen off one or two?

 

[bort]

good work Burgo - agree these are more interesting stats but I just cbf'd crunching the numbers.

Obviously that tells us that Toby Rudolf is the coach's preferred choice to return kicks

Does it tell is that, or does it suggest that out of our props (usually Toby and Woods to start games) other teams coaches preferred to kick to Toby?
I honestly don't know how much coaches would kick to a preferred side of theirs (where we could put our preferred returner) or to the side with the prop they'd rather tackle with a run up.

Out of curiosity did you account for position (only meters made as a forward)?

Don't suppose you'd want to slot Dale and Cam in there too, just for comparison? I'm really hoping we can find one or both are a step above Woods!

Finucane could presumably go in easy enough, Cam from 2021 wouldn't rank well...

 

[Rob]

Great stuff, I personally would include 50% of the kick return meters because they would likely have done a different hit up in that set if they hadn’t already taken one.

Other big factor that effects the whole team so would balance out is working it out of our own 20 which is a lot harder than other meters and without looking I’d dare say we seemed to be starting our sets in our own 20 more than anyone else with our poor defence and poor kick meters.

 

[bort]

Great stuff, I personally would include 50% of the kick return meters because they would likely have done a different hit up in that set if they hadn’t already taken one.

Other big factor that effects the whole team so would balance out is working it out of our own 20 which is a lot harder than other meters and without looking I’d dare say we seemed to be starting our sets in our own 20 more than anyone else with our poor defence and poor kick meters.

That is an interesting point, doing a large bulk of the returns would impact Toby's 'measured' output. You can always look deeper!

 

[Proud Shark]

Great stuff, I personally would include 50% of the kick return meters because they would likely have done a different hit up in that set if they hadn’t already taken one.

Other big factor that effects the whole team so would balance out is working it out of our own 20 which is a lot harder than other meters and without looking I’d dare say we seemed to be starting our sets in our own 20 more than anyone else with our poor defence and poor kick meters.

Yeah I was going to say this. You can't completely exclude kick returns, 50% would be much better. Appreciate the effort that went into this Burgo.

 

[BurgoShark]

Realyl interesting Burgo, cheers. A bit depressing, but interesting...!

Don't suppose you'd want to slot Dale and Cam in there too, just for comparison? I'm really hoping we can find one or both are a step above Woods!

Didn’t think to do Cam - but I did Dale. 174.

Will add Cam (2020) when I have time.

** edit ** whoops - misremembered that. Just checked my numbers. Dale Finucane = 137

Cam McInnes played 7 games at lock in 2020 = 138

I'll add both to the first post.

 

[HaroldBishop]

Can someone please summarise this thread for egg?

 

[egg]

Can someone please summarise this thread for egg?

I read it as swap Toby for Hunt in the 17

Edit : And Wilton for Nikora

 

[BurgoShark]

good work Burgo - agree these are more interesting stats but I just cbf'd crunching the numbers.



Does it tell is that, or does it suggest that out of our props (usually Toby and Woods to start games) other teams coaches preferred to kick to Toby?
I honestly don't know how much coaches would kick to a preferred side of theirs (where we could put our preferred returner) or to the side with the prop they'd rather tackle with a run up.

Out of curiosity did you account for position (only meters made as a forward)?

I’d say it has to be at least in-part deliberate. The difference between some players is massive. Toby over 500m. Woods 26m. Preferred returner stands in middle, and the catcher either side passes to him as a priority. Set starts mid-field.

On position… just metres in games where they played as a forward at some stage. I excluded the games where Wilton/Talakai played 80 in the centres, but in games where they swapped mid-game there is no way of differentiating from nrl.com stats which metres came when. I’d need to watch every game.

 

[BurgoShark]

Great stuff, I personally would include 50% of the kick return meters because they would likely have done a different hit up in that set if they hadn’t already taken one.

Other big factor that effects the whole team so would balance out is working it out of our own 20 which is a lot harder than other meters and without looking I’d dare say we seemed to be starting our sets in our own 20 more than anyone else with our poor defence and poor kick meters.

I definitely agree on point #1. I just don’t think including them fully yields a fair comparison. Toby gets to run 20m with air in front on him but the next bloke has 3 blokes in his face by the time he has taken four steps.

On point #2, that is why this thread exists. Using just metres gives you zero context. Unfortunately there is no way to get this type of field-position based data (for the average punter anyway). I’m sure the teams have this sort of granularity.

Anyone who has taken a hit up off his own tryline will tell you that a 1m run there is tougher than a 10m run off a quick play the ball 60m upfield.

Ultimately this is all just supporting the argument that unless someone is pumping out numbers like Paul Gallen, Payne Haas or JT3 - you can't use "metres" as a justification of them doing a better/worse job than anyone else. It's not really indicative of anything other than the player having touched the ball at least 10-15 times.

 

[bort]

My main takeaway has been we need to cut back on scoring tries as we are making Toby take too many kick returns

 

[BurgoShark]

My main takeaway has been we need to cut back on scoring tries as we are making Toby take too many kick returns

Good call. Probably should deliberately kick out on the full too so we don’t accidentally jag a dropout.

BHU, Williams and Tolman were all pretty high on the kick returner metres too. Maybe the rest of the NRL has done video and are deliberately avoiding kicking anywhere near Woods.

 

[HaroldBishop]

My main takeaway has been we need to cut back on scoring tries as we are making Toby take too many kick returns

There was a call in here last week or so to do away with goalkicking, maybe we scrap tries as well.

 

[Rob]

Is say they don’t kick to woods so he is forced to get dominated on tackle 2

 

[Born&bred]

Another factor to add to the mix, but no stats would ever be available, is how far back the defence is - 10 metres? or as in many poorly officiated games, nowhere near it...

I've only ever played under the 5 metre rule, so hit-ups metres were tough to get.

 

[BurgoShark]

Another factor to add to the mix, but no stats would ever be available, is how far back the defence is - 10 metres? or as in many poorly officiated games, nowhere near it...

I've only ever played under the 5 metre rule, so hit-ups metres were tough to get.

So exclude games against Penrith and Melbourne?

 

[BurgoShark]

Based on the feedback and what I have learned, I am going to make some more adjustments and re-post the results.

- For kick return metres, rather than saying 50% of all metres, I'll say that kick return metres are capped at 20% of total output for any one game. E.g. A player has 100m from hitups and 30m from kick returns he only gets 20m added to the total (120m). Seems fairer. If his total for any game is less than 20% he gets all of it.
- I'll adjust with respect to where they played for the majority of any one game. E.g. 30 minutes in the middle and 15 on the edge gets recorded as having played a game in the middle
- I'll separate results based on length of game time. E.g. Williams is a beast when he plays less than 40 minutes, but not so much when he plays 45+
- I'll also make some changes to how the averages are done, to better weigh high versus low minute performances (boring maths stuff)

 

[Sparkles]

Didn’t think to do Cam - but I did Dale. 174.

Will add Cam (2020) when I have time.

That's a good number

Based on the feedback and what I have learned, I am going to make some more adjustments and re-post the results.

- For kick return metres, rather than saying 50% of all metres, I'll say that kick return metres are capped at 20% of total output for any one game. E.g. A player has 100m from hitups and 30m from kick returns he only gets 20m added to the total (120m). Seems fairer. If his total for any game is less than 20% he gets all of it.
- I'll adjust with respect to where they played for the majority of any one game. E.g. 30 minutes in the middle and 15 on the edge gets recorded as having played a game in the middle
- I'll separate results based on length of game time. E.g. Williams is a beast when he plays less than 40 minutes, but not so much when he plays 45+
- I'll also make some changes to how the averages are done, to better weigh high versus low minute performances (boring maths stuff)

Could you also count the metres Fifita runs to join a melee? Maybe at 20%, it'd still be a great bump