The Impact Scores - Who REALLY did play well

[DownAtTheJunction]

This is the second match for those interested in finding players that impact, rather than raw statistics. For those unsure of how this works I’ll quickly revisit. I allocate points (and half points) to actions that impacted our performance. What constitutes impact? This is quite subjective - it might be a tap on, a spoil, a handpass, a kick, a block etc. Some actions are worthy of higher points. These include more difficult goals, multiple efforts and higher level skills. Conversely deductions were made for ‘negative impacts’ - usually clangers. Not every kick (and mark) receives a score. A player has to show something or do something to score.
This form of rating is fallible as it’s reliant on what a camera shows (which in the main is significant).
There is no account for the time the ball is in the player’s zone, nor is there any consideration for the time the player was off the field, nor the quality of the player’s opponent This is only the final TOTAL. I do have all four quarters but I don't have Sainter_Dad and his expertise.
If he can come to my rescue (again) I will publish the full details (Q1 - Q4)
Please understand that I endeavour to exclude bias. Some may think I have treated some players harshly (an example may be Paton). However I deal strictly in the 'facts' of the vision and I replay all actions multiple times.
24 Steele
21.5 Jones
20 Gresham
19 Hind
18 Sinclair
16.5 Clark
15.5 Coffield
14 Butler
13.5 Marshall
13 Savage
11.5 Howard
11.5 Billings
11.5 Ryder
11 King
10 Hill
9.5 Carlisle
9.5 Wilkie
9 Battle
9 Membrey
9 Parker
8.5 Paton
7.5 Geary

[Toy Saint]

Another outstanding assessment down at the Junction.

Last week your analysis identified Hind which surprised many, but he delivered again this week, and even blind Freddy could see he was outstanding.

This week you have rated Sinclair higher than most. After watching the replay it's clear that he does a lot of things right, his decision making and his disposal are excellent. Players with Sinclair's characteristics tend to be better in good teams, whereas they are often criticized in battling teams for being soft.

Really appreciate your numerical analysis, and believe it's a more accurate reflection of the game than any other structured analysis. Great work.

[Mr Magic]

Great effort Junction

[CURLY]

Paton at second last shows this system has major faults.

[StPeter]

Another outstanding assessment down at the Junction.

Last week your analysis identified Hind which surprised many, but he delivered again this week, and even blind Freddy could see he was outstanding.

This week you have rated Sinclair higher than most. After watching the replay it's clear that he does a lot of things right, his decision making and his disposal are excellent. Players with Sinclair's characteristics tend to be better in good teams, whereas they are often criticized in battling teams for being soft.

Really appreciate your numerical analysis, and believe it's a more accurate reflection of the game than any other structured analysis. Great work.

Sinclair has been unappreciated by many fans for a long time.
He has long been close to my favourite player since I first saw him play while still a rookie and if memory serves me correctly, he was the only player on our list rated as elite by the stats guys a couple of years ago.

The selectors thought enough of him to play every game last year but he has been unfortunately squeezed out for most of this season by our impressive list of new players brought into the team.

[DownAtTheJunction]

Another outstanding assessment down at the Junction.

Last week your analysis identified Hind which surprised many, but he delivered again this week, and even blind Freddy could see he was outstanding.

This week you have rated Sinclair higher than most. After watching the replay it's clear that he does a lot of things right, his decision making and his disposal are excellent. Players with Sinclair's characteristics tend to be better in good teams, whereas they are often criticized in battling teams for being soft.

Really appreciate your numerical analysis, and believe it's a more accurate reflection of the game than any other structured analysis. Great work.

That's very kind, Toy Saint. I must say the analysis surprises me too - I would have sold off Hind but he is proving many on here (but not all) wrong. I fear consistency may trouble him but I think he'd be feeling pretty damn confident right now.
With respect to Sinclair I think he is a fairly maligned player. His contested possession rate is much higher than many have given him credit for. He's not only a good looking kick (like Billings) but the difference is he hits targets more often. I also think he is one of our most creative footballers. I might add he puts in some ordinary games and I suspect this is where he attracts the bad press.

[DownAtTheJunction]

Paton at second last shows this system has major faults.

I would sincerely (and respectfully) ask you to check the vision again. I may well be in the minority, but I watch it several times over (often 'frame by frame' of every action) to make the best judgement I can. It would be unfair to score him on efforts that are not on the screen. Clearly he held a dangerous opponent but I did not see many significant actions where I could give him points.
However this is an opinion piece and I respect your view.

[skeptic]

Paton at second last shows this system has major faults.

I would sincerely (and respectfully) ask you to check the vision again. I may well be in the minority, but I watch it several times over (often 'frame by frame' of every action) to make the best judgement I can. It would be unfair to score him on efforts that are not on the screen. Clearly he held a dangerous opponent but I did not see many significant actions where I could give him points.
However this is an opinion piece and I respect your view.

To speak to this however... the key word is IMPACT.

It makes a lot of that to some degree it’s going to be harder for defenders or defensive minded players who tend to do this that may not directly appear to impact play (in a... wow look at that play way).

Going through the list... note that Paton, Wilkie, Geary, Carlisle... all in the second half of that list.

It doesn’t mean that they didn’t do their bit or that they didn’t do well.

I wouldn’t rate Savage’s contribution higher than most of them but am happy to concede the point that he pbly directly influenced more of the play

[DownAtTheJunction]

Sinclair has been unappreciated by many fans for a long time.
He has long been close to my favourite player since I first saw him play while still a rookie and if memory serves me correctly, he was the only player on our list rated as elite by the stats guys a couple of years ago.

The selectors thought enough of him to play every game last year but he has been unfortunately squeezed out for most of this season by our impressive list of new players brought into the team.

Yes St Peter. Sinclair has added some genuine grunt to his game. 10 contested possessions was equal 3rd for the Saints. He had the second most number of tackles (for either team) with 7. He's not afraid to get his hands dirty.

[DownAtTheJunction]

Paton at second last shows this system has major faults.

I would sincerely (and respectfully) ask you to check the vision again. I may well be in the minority, but I watch it several times over (often 'frame by frame' of every action) to make the best judgement I can. It would be unfair to score him on efforts that are not on the screen. Clearly he held a dangerous opponent but I did not see many significant actions where I could give him points.
However this is an opinion piece and I respect your view.

To speak to this however... the key word is IMPACT.

It makes a lot of that to some degree it’s going to be harder for defenders or defensive minded players who tend to do this that may not directly appear to impact play (in a... wow look at that play way).

Going through the list... note that Paton, Wilkie, Geary, Carlisle... all in the second half of that list.

It doesn’t mean that they didn’t do their bit or that they didn’t do well.

I wouldn’t rate Savage’s contribution higher than most of them but am happy to concede the point that he pbly directly influenced more of the play

Absolutely correct, Skeptic. The defenders have a role to do. And it's pretty basic and simple - stop the opposition. Some of it is sheer grit, a subtle block or positioning (often before the ball is near). And the players you mentioned - all contributed. in fact I gave Paton several minor points for his impacts in Q3. Wilkie is almost 'Mr Invisible' but when he's needed he's there and he's ready.
I might add that we won by more than 50 points so much of the time our defence was not in high need.

[CQ SAINT]

A quick question for you DATJ.
Say Hind receives a handpass, steps around an opponent and kicks a banana kick for a goal on his left foot and then later gets the ball and kickd in a similar fashion but misses everything, would the 2 scores cancel each other?

[cps]

Enjoyed your analysis but what of players who are asked to sacrifice their own game for the benefit of the team?

For example, early on in the game we brought the ball into the forward line and two of the forwards led their opponents away from the contest leaving King (I think) one out with their opponent. At the time it looked like the F50 was deserted! It was a superb example of team play - how would you rate it?

Cheers!

[Scollop]

This is the second match for those interested in finding players that impact, rather than raw statistics. For those unsure of how this works I’ll quickly revisit. I allocate points (and half points) to actions that impacted our performance. What constitutes impact? This is quite subjective - it might be a tap on, a spoil, a handpass, a kick, a block etc. Some actions are worthy of higher points. These include more difficult goals, multiple efforts and higher level skills. Conversely deductions were made for ‘negative impacts’ - usually clangers. Not every kick (and mark) receives a score. A player has to show something or do something to score.
This form of rating is fallible as it’s reliant on what a camera shows (which in the main is significant).
There is no account for the time the ball is in the player’s zone, nor is there any consideration for the time the player was off the field, nor the quality of the player’s opponent This is only the final TOTAL. I do have all four quarters but I don't have Sainter_Dad and his expertise.
If he can come to my rescue (again) I will publish the full details (Q1 - Q4)
Please understand that I endeavour to exclude bias. Some may think I have treated some players harshly (an example may be Paton). However I deal strictly in the 'facts' of the vision and I replay all actions multiple times.
24 Steele
21.5 Jones
20 Gresham
19 Hind
18 Sinclair
16.5 Clark
15.5 Coffield
14 Butler
13.5 Marshall
13 Savage
11.5 Howard
11.5 Billings
11.5 Ryder
11 King
10 Hill
9.5 Carlisle
9.5 Wilkie
9 Battle
9 Membrey
9 Parker
8.5 Paton
7.5 Geary

Gees!! blocks hey? Might have been a similar system used by Richo and ol mate stonecold when they were rating Billy Longer

Stevie Baker was accomplished at executing a few effective blocks

Great effort D a T J and I agree with most of the ratings. I didn't have Jones high up in my votes simply because I think a guy who pinpoints a pass over 30m metres should be rated higher than the bloke who gets it and kicks it long to a contest. I saw Jones kick a few deep inside 50 entries that ended up being goals, but they could have easily been turnovers as well. If a player sets us up accidentally I don't think that deserves as many points as a player deliberately looking for and finding a team mate in an attacking position.

[CURLY]

Paton at second last shows this system has major faults.

I would sincerely (and respectfully) ask you to check the vision again. I may well be in the minority, but I watch it several times over (often 'frame by frame' of every action) to make the best judgement I can. It would be unfair to score him on efforts that are not on the screen. Clearly he held a dangerous opponent but I did not see many significant actions where I could give him points.
However this is an opinion piece and I respect your view.

Don't need to. He kept their best player to one goal. He has had as equal impact to any Saints player on the ground.

[Secret Kiel]

Paton at second last shows this system has major faults.

I would sincerely (and respectfully) ask you to check the vision again. I may well be in the minority, but I watch it several times over (often 'frame by frame' of every action) to make the best judgement I can. It would be unfair to score him on efforts that are not on the screen. Clearly he held a dangerous opponent but I did not see many significant actions where I could give him points.
However this is an opinion piece and I respect your view.

Don't need to. He kept their best player to one goal. He has had as equal impact to any Saints player on the ground.

Is it possible his opponent had a very ordinary game which would mean objectively speaking, DATJuncs measure system for acts that are considered "game impacting" is an accurate (ish) measure on that stat alone? As with everything to do with analytics, you need to set up a complete measurement system of what you are trying to achieve so it would be interesting to analyse Paton's opponent to get a more fullish picture. Paton's oponent could have been having a really crap game thus in-turn making his job much easier. The scoreline suggests the backline had a very easy day at the office so not necessarily a knock on Paton.

[CURLY]

Paton at second last shows this system has major faults.

I would sincerely (and respectfully) ask you to check the vision again. I may well be in the minority, but I watch it several times over (often 'frame by frame' of every action) to make the best judgement I can. It would be unfair to score him on efforts that are not on the screen. Clearly he held a dangerous opponent but I did not see many significant actions where I could give him points.
However this is an opinion piece and I respect your view.

Don't need to. He kept their best player to one goal. He has had as equal impact to any Saints player on the ground.

Is it possible his opponent had a very ordinary game which would mean objectively speaking, DATJuncs measure system for acts that are considered "game impacting" is an accurate (ish) measure on that stat alone? As with everything to do with analytics, you need to set up a complete measurement system of what you are trying to achieve so it would be interesting to analyse Paton's opponent to get a more fullish picture. Paton's oponent could have been having a really crap game thus in-turn making his job much easier. The scoreline suggests the backline had a very easy day at the office so not necessarily a knock on Paton.

Odd observation. The scoreline was what it was because of the impact of our defence. Its like supporters saying their ruckman chose to have his worst game against Nic Nat.

[Secret Kiel]

Paton at second last shows this system has major faults.

I would sincerely (and respectfully) ask you to check the vision again. I may well be in the minority, but I watch it several times over (often 'frame by frame' of every action) to make the best judgement I can. It would be unfair to score him on efforts that are not on the screen. Clearly he held a dangerous opponent but I did not see many significant actions where I could give him points.
However this is an opinion piece and I respect your view.

Don't need to. He kept their best player to one goal. He has had as equal impact to any Saints player on the ground.

Is it possible his opponent had a very ordinary game which would mean objectively speaking, DATJuncs measure system for acts that are considered "game impacting" is an accurate (ish) measure on that stat alone? As with everything to do with analytics, you need to set up a complete measurement system of what you are trying to achieve so it would be interesting to analyse Paton's opponent to get a more fullish picture. Paton's oponent could have been having a really crap game thus in-turn making his job much easier. The scoreline suggests the backline had a very easy day at the office so not necessarily a knock on Paton.

Odd observation. The scoreline was what it was because of the impact of our defence. Its like supporters saying their ruckman chose to have his worst game against Nic Nat.

You've missed the point. Your "observation" is just that, a subjective observation that is disputing DATJuncs objective measure of a particular act in the game. Remember he is only measuring one very small aspect of a player's output and it's possible Paton a very good game in the eyes of the coach but has scored low on the measure of DATjuncs interest in the game and players.

[CQ SAINT]

Paton at second last shows this system has major faults.

I would sincerely (and respectfully) ask you to check the vision again. I may well be in the minority, but I watch it several times over (often 'frame by frame' of every action) to make the best judgement I can. It would be unfair to score him on efforts that are not on the screen. Clearly he held a dangerous opponent but I did not see many significant actions where I could give him points.
However this is an opinion piece and I respect your view.

Don't need to. He kept their best player to one goal. He has had as equal impact to any Saints player on the ground.

Is it possible his opponent had a very ordinary game which would mean objectively speaking, DATJuncs measure system for acts that are considered "game impacting" is an accurate (ish) measure on that stat alone? As with everything to do with analytics, you need to set up a complete measurement system of what you are trying to achieve so it would be interesting to analyse Paton's opponent to get a more fullish picture. Paton's oponent could have been having a really crap game thus in-turn making his job much easier. The scoreline suggests the backline had a very easy day at the office so not necessarily a knock on Paton.

In the experiences I have had. I find the data and the eye are fairly equal in terms ranking your best players for impact on any given day.
I don't believe the screen or the measures used in this case can evaluate Paton's or Papleys true impact without considering a number of other factors, the main ones being clearances, Sydney's F50 entries and St.Kilda's D50 rebounds.
Paton had no tackles, 7 marks, only 1 contested mark, 5 contested and 9 uncontested possessions and only 3 pressure acts. He did this @ 71%DE.
He didnt just beat Papley, he provided rebound and run which cut off Sydney's entries. He wasn't alone, Coffield and Clark dominated for much of the game in this area.
So if you consider the work done by these 3 in tandem with the impact of our mids and then consider time with ball in your half, you get a better picture of why Paton beats Papley. Geary could have beaten Papley but he would have needed to let Papley take him to the ball.
If Paton tracks Papleys movement, which is based on getting ahead of the ball and into one on one's, I'm more than impressed with a 9.5 from Paton going the other way compared to 24 for Steele who feeds off stoppages and a distinct ruck advantage. Both of them had a massive impact on the team's ability to win.

[Secret Kiel]

Paton at second last shows this system has major faults.

I would sincerely (and respectfully) ask you to check the vision again. I may well be in the minority, but I watch it several times over (often 'frame by frame' of every action) to make the best judgement I can. It would be unfair to score him on efforts that are not on the screen. Clearly he held a dangerous opponent but I did not see many significant actions where I could give him points.
However this is an opinion piece and I respect your view.

Don't need to. He kept their best player to one goal. He has had as equal impact to any Saints player on the ground.

Is it possible his opponent had a very ordinary game which would mean objectively speaking, DATJuncs measure system for acts that are considered "game impacting" is an accurate (ish) measure on that stat alone? As with everything to do with analytics, you need to set up a complete measurement system of what you are trying to achieve so it would be interesting to analyse Paton's opponent to get a more fullish picture. Paton's oponent could have been having a really crap game thus in-turn making his job much easier. The scoreline suggests the backline had a very easy day at the office so not necessarily a knock on Paton.

In the experiences I have had. I find the data and the eye are fairly equal in terms ranking your best players for impact on any given day.
I don't believe the screen or the measures used in this case can evaluate Paton's or Papleys true impact without considering a number of other factors, the main ones being clearances, Sydney's F50 entries and St.Kilda's D50 rebounds.
Paton had no tackles, 7 marks, only 1 contested mark, 5 contested and 9 uncontested possessions and only 3 pressure acts. He did this @ 71%DE.
He didnt just beat Papley, he provided rebound and run which cut off Sydney's entries. He wasn't alone, Coffield and Clark dominated for much of the game in this area.
So if you consider the work done by these 3 in tandem with the impact of our mids and then consider time with ball in your half, you get a better picture of why Paton beats Papley. Geary could have beaten Papley but he would have needed to let Papley take him to the ball.
If Paton tracks Papleys movement, which is based on getting ahead of the ball and into one on one's, I'm more than impressed with a 9.5 from Paton going the other way compared to 24 for Steele who feeds off stoppages and a distinct ruck advantage. Both of them had a massive impact on the team's ability to win.

Yes agree that DATJunc is measuring a very isolated part of the game but disagree with eye and objective measurement systems being equal.

[CQ SAINT]

Paton at second last shows this system has major faults.

I would sincerely (and respectfully) ask you to check the vision again. I may well be in the minority, but I watch it several times over (often 'frame by frame' of every action) to make the best judgement I can. It would be unfair to score him on efforts that are not on the screen. Clearly he held a dangerous opponent but I did not see many significant actions where I could give him points.
However this is an opinion piece and I respect your view.

Don't need to. He kept their best player to one goal. He has had as equal impact to any Saints player on the ground.

Is it possible his opponent had a very ordinary game which would mean objectively speaking, DATJuncs measure system for acts that are considered "game impacting" is an accurate (ish) measure on that stat alone? As with everything to do with analytics, you need to set up a complete measurement system of what you are trying to achieve so it would be interesting to analyse Paton's opponent to get a more fullish picture. Paton's oponent could have been having a really crap game thus in-turn making his job much easier. The scoreline suggests the backline had a very easy day at the office so not necessarily a knock on Paton.

In the experiences I have had. I find the data and the eye are fairly equal in terms ranking your best players for impact on any given day.
I don't believe the screen or the measures used in this case can evaluate Paton's or Papleys true impact without considering a number of other factors, the main ones being clearances, Sydney's F50 entries and St.Kilda's D50 rebounds.
Paton had no tackles, 7 marks, only 1 contested mark, 5 contested and 9 uncontested possessions and only 3 pressure acts. He did this @ 71%DE.
He didnt just beat Papley, he provided rebound and run which cut off Sydney's entries. He wasn't alone, Coffield and Clark dominated for much of the game in this area.
So if you consider the work done by these 3 in tandem with the impact of our mids and then consider time with ball in your half, you get a better picture of why Paton beats Papley. Geary could have beaten Papley but he would have needed to let Papley take him to the ball.
If Paton tracks Papleys movement, which is based on getting ahead of the ball and into one on one's, I'm more than impressed with a 9.5 from Paton going the other way compared to 24 for Steele who feeds off stoppages and a distinct ruck advantage. Both of them had a massive impact on the team's ability to win.

Yes agree that DATJunc is measuring a very isolated part of the game but disagree with eye and objective measurement systems being equal.

Fair enough, but Blind Freddy would have come up with the top 10 ranked players in DaTJ impact stats. But, only a keen tactical mind sees Paton as the answer to Papley. 30 games of Paton's data might give you a hint, which is very useful in fine tuning tactically but I need more data, much more, to qualify who REALLY played well. I saw Paton coming in his first year. I reckon Ratten did too.

[DownAtTheJunction]

A quick question for you DATJ.
Say Hind receives a handpass, steps around an opponent and kicks a banana kick for a goal on his left foot and then later gets the ball and kickd in a similar fashion but misses everything, would the 2 scores cancel each other?

Allocating a score to an action is undoubtedly the most difficult part of any analysis. I'm not a huge fan of 'long kick = 1pt' approach. I have tried to clearly assess 'that long kick' - to see whether it was to advantage, and by how much, the difficulty involved, and the 'stress' the player was under. As I do it by myself at least there should be some consistency (in theory anyway).
As for the specifics above - every circumstance is different, and I would try to take that into account. I am not great on penalising in this model. So I'd give points for positive acts. I tend to penalise when there is a turnover by a player who is not under pressure.

[CQ SAINT]

A quick question for you DATJ.
Say Hind receives a handpass, steps around an opponent and kicks a banana kick for a goal on his left foot and then later gets the ball and kickd in a similar fashion but misses everything, would the 2 scores cancel each other?

Allocating a score to an action is undoubtedly the most difficult part of any analysis. I'm not a huge fan of 'long kick = 1pt' approach. I have tried to clearly assess 'that long kick' - to see whether it was to advantage, and by how much, the difficulty involved, and the 'stress' the player was under. As I do it by myself at least there should be some consistency (in theory anyway).
As for the specifics above - every circumstance is different, and I would try to take that into account. I am not great on penalising in this model. So I'd give points for positive acts. I tend to penalise when there is a turnover by a player who is not under pressure.

Thanks. Makes a lot of sense. Great work.

[DownAtTheJunction]

Enjoyed your analysis but what of players who are asked to sacrifice their own game for the benefit of the team?

For example, early on in the game we brought the ball into the forward line and two of the forwards led their opponents away from the contest leaving King (I think) one out with their opponent. At the time it looked like the F50 was deserted! It was a superb example of team play - how would you rate it?

Cheers!

It's a perfectly fine point, cps which I can't definitively answer. I only give points to the vision I can see. I don't think it's appropriate for me to guess. There are many things that an observer is not privy to, and that does not mean they should not be allocated points - I'm sure the coach will though!

[DownAtTheJunction]

Great effort D a T J and I agree with most of the ratings. I didn't have Jones high up in my votes simply because I think a guy who pinpoints a pass over 30m metres should be rated higher than the bloke who gets it and kicks it long to a contest. I saw Jones kick a few deep inside 50 entries that ended up being goals, but they could have easily been turnovers as well. If a player sets us up accidentally I don't think that deserves as many points as a player deliberately looking for and finding a team mate in an attacking position.

Thank you Scollop. I definitely understand your perspective re Jones - he can be a tough one to adjudicate. The set up pass is important, for sure, as is the endeavour of a player. I am confident the players value Jones' tenacity.