Twitter

Tuesday 21 July 2015

Man Games Lost To Injury In The Premier League (2012/13 - 2014/15)

I first wrote about injury and man games lost to injury all the way back at the start of 2013. That first post on the topic of injuries was followed by this , more extensive, post on the 2012/13 season. I didn't post too much on the injury data for the 2013/14 season for I was informed, with the best of intentions, that I may want to keep any future data private for purposes that were never disclosed (money, I guess. Though I am certain there ain't no money in this!). 

Anyways, I don't want to keep data private or try and hawk any wares, be it good, bad, or any lemon in-between. Personally, I much prefer making any data, or writing, public with the intent of having folk beat the shit out of it, mock it, or make recommendations for the purposes of improving one's own work. That said, I am probably in the minority here. 

I digress, go read those linked posts for more info on method, the effect of European football on injury totals (only 1 season), and a few brisk thoughts on potential solutions to the worst of the injury problems that the league has.


Key


Man Game Lost To Injury = the number of games missed by a single injured player. For example, imagine a player, let's call him Linvoy Primus, suffers a simple calf strain and is injured for 30 days. Linvoy's club plays 4 Premier League games during those 30 days. Linvoy Primus misses 4 man games lost to injury, his club has lost 4 man games lost to injury.
I'll try to keep things short from now on. Lots of graphs, fewer words. 


Table


This sortable table lists each club and their man games lost to injury for each season.

Villa, Newcastle, Man United and Arsenal posted consistently high numbers by this metric. Chelsea, on the other hand, have posted consistently low numbers despite carrying a relatively small squad.

Sort away, if you like.


S 12/13 Man Games Lost  S 13/14 Man Games Lost  S 14/15 Man Games Lost 
Arsenal 12/13 138 Arsenal 13/14 211 Arsenal 14/15 231
Aston Villa 12/13 219 Aston Villa 13/14 182 Aston Villa 14/15 169
Chelsea 12/13 81 Cardiff 13/14 67 Burnley 14/15 113
Everton 12/13 82 Chelsea 13/14 79 Chelsea 14/15 53
Fulham 12/13 116 Palace 13/14 128 Palace 14/15 98
Liverpool 12/13 122 Everton 13/14 175 Everton 14/15 180
Man City 12/13 140 Fulham 13/14 120 Hull 14/15 130
Man United 12/13 148 Hull 13/14 112 Leicester 14/15 79
Newcastle 12/13 175 Liverpool 13/14 131 Liverpool 14/15 173
Norwich 12/13 109 Man City 13/14 134 Man City 14/15 86
QPR 12/13 119 Man United 13/14 135 Man United 14/15 169
Reading 12/13 94 Newcastle 13/14 157 Newcastle 14/15 253
Southampton 12/13 66 Norwich 13/14 123 QPR 14/15 140
Stoke 12/13 70 Southampton 13/14 92 Southampton 14/15 138
Sunderland 12/13 117 Stoke 13/14 71 Stoke 14/15 124
Swansea 12/13 91 Sunderland 13/14 99 Sunderland 14/15 114
Tottenham 12/13 128 Swansea 13/14 108 Swansea 14/15 88
West Brom 12/13 106 Tottenham 13/14 165 Tottenham 14/15 72
West Ham 12/13 115 West Brom 13/14 127 West Brom 14/15 82
Wigan 12/13 145 West Ham 13/14 143 West Ham 14/15 111

The above table tells us the raw man games lost to injury for each team, but it tells us nothing about injury trends throughout the season. 

Man Games Lost To Injury On A Monthly Basis


The chart below tells us the league total number of man games lost to injury on a monthly basis for each of the past three Premier League seasons. There's a pretty clear trend in this chart.





If we focus on the 3-year average for a moment we can see a gradual increase in total man game lost from August through to October. We then see a significant jump from 188 games lost in October to 297 in November. There is then another huge jump in December to 453 games lost. 

That number then cools to 313 games lost in January and 248 games lost in February. 

Those mid-winter months of November, December and January really deserve some further investigation. 

A simple question: Why does the league suffer so many injuries during these months? 

Games Per Month


Part of the answer to the question posed above is pretty simple: the number of (league) games played in each month is a pretty important reason as to the number of injuries that occur in certain months.




If we look once again at the 3-year average, we see that the Premier League has a light schedule in the months of August to October. November sees an average of 5 Premier League games, December 6 Premier League games and thereafter, more or less, we see an average of 4 Premier League games scheduled in the remaining months.

It's been over three years since I started collecting this data and I am still unsure as to why the Premier League continues to schedule so many games around November, December, and January (not to mention that clubs also have European football and domestic cup competitions to contend with). Why not redistribute some of those mid-winter games to the early season months thus flattening out the schedule? The international schedule is probably the reason why this 'flattening' out doesn't happen.

Anyway, back to the question I posed: Why does the league suffer so many injuries during these months? 

Part of the answer is the bottleneck of games in mid-winter months. There's also the graph below.

Man Games Lost To Injury On A Game Week Basis


The chart below is showing us man games lost to injury on a game week basis. If fatigue, recovery time, and rotation were unimportant things then maybe we should see a semi-random distribution of man games lost throughout the season.

Instead, we see a gentle increase in weekly games lost to injury, a peak of sorts from week 14 to week 21, and then a gentle decline to week 27. Then injuries slowly pick up again. 

Curious stuff.





The chart above shows the aforementioned steady increase in games lost to injury in the early weeks of the season. We then see a smallish peak in games lost to injury from about week 14 to week 21. Game week 14 is roughly the first week in December and game week 21 is roughly the first week in January. This is our peak of games lost to injury.

Why do we see a peak for those 35 or so days? Big squads and rotation are now the norm at nearly every Premier League club yet the intensity of the league, sheer number of games, insufficient recovery time, travel, games just 3 days apart at certain points in our 'peak', and players playing whilst in the red zone in terms of fatigue would be my guesses.

Despite all the work that goes into injury prevention and sports science I'm unsure how a team who is playing in all competitions is expected to cope with 8 PL games, 1 European game, 1 League Cup game and an FA cup game in a ~35 day period.

Two years ago I wrote the following: 


Looking over the injury total numbers for the November to January period it becomes difficult not to argue the case for a winter break. No club wants to pay a substantial wage to players who are not able to do their job. Injured players - especially crucial players - cost clubs in terms of performance and points, and thus money. Why clubs and the PFA don't push harder for a winter break in order to protect their players and assets is difficult to grasp.
A front and back loaded league schedule which avoids clashing with the bulk of the the domestic and European cups would surely be a better idea than the current setup. Games could be more evenly distributed across the months, injuries may decline due to this and teams would benefit in the ways already stated. Players may stay healthier, managers can field stronger teams and owners may not be wasting as much money paying injured players.
Seems sensible, so it probably won't happen.
My thoughts haven't changed much in two years. A flattening out of the schedule - be it front or back-loaded - whilst retaining the historic Boxing Day and New Years Day games, makes a little more sense to me than the current setup which crams a crazy amount of fixtures (all competitions)  into the months of November, December and January.

In short: 


  • Teams suffer too many injuries in the mid-winter months. 


  • This is due to the sheer number of games played in all competitions. 
  • The frequency of games played in those mid-winter months allows for little recovery time between fixtures. 
  • This leads to fatigue, minutes on the field when players are in the red zone re injury risk, and finally, actual injuries.


Fix the winter fixture list.

*Also worth noting that work on injury metrics could be improved upon if we could we had salary information available for all players. A metric along the lines of CHIP (cap hit of injured players) would be an improvement.



Thursday 16 July 2015

2014/15 Player Numbers By Field Position



A simple question:



Well the answer for the 13/14 Premier League is here, along with the conversion rate for defenders and midfielders. There's tons of other stuff too.

Anyway, let's answer the question (and repeat last years study) for the 14/15 Premier League season.

All players who played at least a single minute in the 14/15 Premier League are included here.


Numbers

Shooting% = goals/shots
Shot Assist Sh% - assists/shot assists
On-Field Sh% = goals+assists/shots+shot assists
Shot Contribution = shots+shot assists p90

Count Position Goals p90 Assists p90 Points p90 Shots p90 SoT p90 Shot Assist p90 Shot Contr p90 Shooting% Shot Assist Sh% On-Field Sh% Sc% SoT% Passes p90 Fouls p90 Fouled p90 Ave Mins
173 Defenders 0.03 0.04 0.07 0.48 0.12 0.48 0.96 6.25 8.56 7.40 24.73 25.28 39.95 0.85 0.70 1544.6
213 Midfielders 0.12 0.12 0.24 1.45 0.45 1.36 2.82 8.37 9.01 8.68 27.08 30.90 47.82 1.27 1.24 1368.2
101 Attackers 0.35 0.11 0.46 2.81 1.04 1.15 3.97 12.41 9.94 11.69 33.64 36.89 29.08 1.50 1.48 1195.5

Some quick thoughts below.

Goals, Assists & Points

I think the trends between defenders, midfielders and forwards is pretty intuitive. Forwards score than midfielders; midfielders score more than defenders.

There's less of a gap between the positions when looking at assists.

Add goals and assists together to get points and we see a nice uniformed decline from forward to midfielder to defender.


Shot Volume & Shot Contribution

Midfielders have a shots p90 rate that is around 3x that what defenders post. Forwards post 2x the shots p90 volume of midfielders.

Midfielders post the best shot assist p90 numbers and this is pretty intuitive: midfielders (the data I am using groups wingers and attacking midfielders in this group) are often the setup men for for the shots that forwards take.

If we add shot volume and shot assists volume together, we get shot contribution p90. Forwards are easily posting the best volume by this metric.

Shooting%

Forwards post the best number (12.4%) by this metric and distance from goal when the shot is taken is likely a big factor here.

Defenders have a shooting% that is roughly half that of forwards. Distance in open play shots may well be one factor. Shot type and how the shot is created - think of headers from set pieces and corners - may well be another.

Shot Assist sh%

Fairly small spread between the positions by this metric. I am open to thoughts as to why this is? Compared to shooting% does shot assist sh% contain more luck/non-skill in having the chances a player creates converted into goals?

Passes p90

Midfielders, naturally, post the best pass volume. Defenders pass the ball more times p90 than forwards do.

No surprise here.

Minutes

The average minutes played for defenders is 1544. For forwards this number is 1195. Forwards are often rotated from game-to-game and are more than likely to be subbed within the game, hence the lower average minutes played. Defenders aren't rotated or subbed off at anywhere near the same rate as forwards. 


All told, I don't think any of this information is new or revolutionary. It simply confirms - using data - some of the basic ideas we have about how each position on the field contributes to offense. 

Forwards>midfielders>defenders.


Thursday 9 July 2015

The Best Premier League Forwards (2010 - 2015): Top 20 Lists

Top 20 Lists

It's summer. It's hotter than hell in my part of the world. It hasn't rained for 2 months and it feels like I might never see rain again. 

Test

So, seeing as it's too hot and I'm lazy, I wanted to post something silly and fun. The meatier subjects of injury analytics, score effects and whatever the hell else is on my list of 'things to du' can wait.

It's time now to look at some of the best Premier League forwards of the last 5 years.

I've got 6 categories and a song. All stat categories will be sorted by p90 but I've also included season, age, raw counts, and minutes played to give you an idea of the size of the sample. 

40% minutes played cut-off. All penalties are removed, as are penalty shots. (man, it feels like so long since I started stripping penalty goals out of goal totals. Nobody knew what the fuck to do with it!)


Top 20's. Tables are sortable.

Goals p90

Oh, Sergio! You might never get your dues from journalists or your peers, but good lord you can score.

Agüero features 3 times on this list. As does RvP (age 27,28,29) and the now fading Edin Dzeko. Suarez features twice. Lukaku posted a 15 goal/0.67 p90 season at age 20. Quite remarkable. Ditto Kane (finishing %'s are a touch high)

I miss Gareth Bale.

Season Age Player / Season Goals p90 Goals Mins
S 13/14 25 Sergio Agüero 0.99 17 1539
S 13/14 27 Luis Suárez 0.94 31 2965
S 10/11 30 Dimitar Berbatov 0.86 20 2096
S 11/12 26 Edin Dzeko 0.84 14 1497
S 13/14 24 Daniel Sturridge 0.83 21 2277
S 14/15 26 Diego Costa 0.82 19 2085
S 10/11 27 Robin van Persie 0.81 16 1768
S 10/11 23 Javier Hernández 0.79 13 1482
S 11/12 28 Robin van Persie 0.75 28 3344
S 14/15 26 Sergio Agüero 0.74 21 2540
S 13/14 28 Edin Dzeko 0.72 16 1993
S 12/13 26 Luis Suárez 0.7 23 2955
S 11/12 23 Sergio Agüero 0.69 20 2602
S 12/13 27 Edin Dzeko 0.69 14 1816
S 11/12 26 Wayne Rooney 0.67 21 2839
S 12/13 20 Romelu Lukaku 0.67 15 2006
S 14/15 28 Olivier Giroud 0.67 14 1867
S 12/13 29 Robin van Persie 0.66 23 3123
S 14/15 21 Harry Kane 0.66 19 2582
S 12/13 23 Gareth Bale 0.65 21 2925


Assists p90

Double figures in assists, for forwards, is a pretty rare feat. 

Rooney has 3 of the top 10 seasons. Aguero is on this list twice, as is Hazard. No other player repeats.

I was surprised the corpse of Fernando Torres appeared on the top 20 assists p90 list. I can only assume those assists were mis-hit shots. 

Glad to see Jamie Vardy on here. His road to the top is as great a story as it is rare.


Season Age Player / Season Assits p90 Assists Mins
S 12/13 27 Wayne Rooney 0.45 10 2019
S 10/11 33 Didier Drogba 0.42 13 2793
S 12/13 28 Lukas Podolski 0.41 9 2000
S 12/13 22 Eden Hazard 0.38 11 2640
S 13/14 28 Wayne Rooney 0.37 10 2448
S 13/14 27 Luis Suárez 0.36 12 2965
S 10/11 27 Robin van Persie 0.36 7 1768
S 10/11 25 Wayne Rooney 0.36 11 2723
S 10/11 30 Peter Crouch 0.36 6 1481
S 13/14 25 Sergio Agüero 0.35 6 1539
S 11/12 28 Emmanuel Adebayor 0.35 11 2841
S 13/14 32 Rickie Lambert 0.32 10 2820
S 11/12 25 Gervinho 0.32 6 1688
S 14/15 28 Jamie Vardy 0.32 8 2247
S 12/13 29 Carlos Tevez 0.3 8 2410
S 13/14 30 Fernando Torres 0.29 5 1572
S 14/15 28 Shane Long 0.29 5 1542
S 14/15 24 Eden Hazard 0.29 11 3379
S 13/14 24 Daniel Sturridge 0.28 7 2277
S 14/15 26 Sergio Agüero 0.28 8 2540

Points p90

Points = goals+assists.

Aguero's injury shortened 13/14 season is the best forward season of the last 5 years (and possibly longer).

Suarez joins Aguero in being way clear of the pack. You want a quick example of how amazingly good Suarez was in 13/14? He dragged the one-legged Daniel Sturridge into his spectacular orbit and enabled the England forward to post a season good enough to be in #5 spot.

Van Persie appears 3 times on this list. Aguero x2. Rooney x2. Suarez x2. Dzeko x2.

Giroud and Kane creep into the bottom of the list. 

A quick mention for #12, the one and only Lukas Podolski. The question any man, balanced and sound of mind, may ask is: How the fuck is Podolski on the list? 

Podolski had a sh% of 20% and a shot assist% of 28% and an on-field shooting% (all goals and assists/ all shots and shot assists) of 23%. All these percentages may as well be orbiting in outer space such is how high and elevated they are. No player can repeat those percentages. Arsenal should of sold him the minute that 12/13 season finished.

I look at this list and I can't help think of all the talent that has either left or is soon to be aged beyond recognition. 10 of the players who posted top 20 points p90 seasons have already left/about to leave the league. Sturridge is broken, Rooney's best years are gone. Aguero, Lukaku and Kane are the last men standing. 


Season Age Player / Season Points p90 Points Mins
S 13/14 25 Sergio Agüero 1.35 23 1539
S 13/14 27 Luis Suárez 1.31 43 2965
S 10/11 27 Robin van Persie 1.17 23 1768
S 13/14 24 Daniel Sturridge 1.11 28 2277
S 14/15 26 Sergio Agüero 1.03 29 2540
S 10/11 30 Dimitar Berbatov 1.03 24 2096
S 11/12 26 Edin Dzeko 1.02 17 1497
S 11/12 28 Robin van Persie 1 37 3344
S 11/12 23 Sergio Agüero 0.97 28 2602
S 14/15 26 Diego Costa 0.95 22 2085
S 12/13 27 Wayne Rooney 0.94 21 2019
S 12/13 28 Lukas Podolski 0.9 20 2000
S 12/13 29 Robin van Persie 0.89 31 3123
S 13/14 28 Wayne Rooney 0.88 24 2448
S 12/13 20 Romelu Lukaku 0.85 19 2006
S 12/13 26 Luis Suárez 0.85 28 2955
S 10/11 23 Javier Hernández 0.85 14 1482
S 12/13 27 Edin Dzeko 0.84 17 1816
S 14/15 28 Olivier Giroud 0.82 17 1867
S 14/15 21 Harry Kane 0.8 23 2582

Shots p90

Moving away from goals now and focusing on all things shots.

Suarez was a shot monster in 13/14 when everybody correctly said he was amazing. Suarez was also a shot monster in his previous seasons with Liverpool. Location could be sketchy, but the chaos and high event game was evident in 12/13 - which is the best shot volume season on record (10 - 15). I think I came up with the saying 'a hub player' specifically for Suarez. 

Aguero, Suarez, Dzeko and Van Persie take up 11 spots on this list. Wilfried Bony had a nice season.

Oddballs such as Pavlyuchenko, Obinna, and Saha somehow make this list.

I miss Gareth Bale.

Season Age Player / Season Shots p90 Shots Mins
S 12/13 26 Luis Suárez 5.7 187 2955
S 11/12 26 Edin Dzeko 5.53 92 1497
S 13/14 27 Luis Suárez 5.49 181 2965
S 12/13 28 Demba Ba 5.15 134 2342
S 12/13 23 Gareth Bale 5.08 165 2925
S 14/15 26 Sergio Agüero 5.03 142 2540
S 13/14 25 Sergio Agüero 4.97 85 1539
S 14/15 26 Wilfried Bony 4.95 88 1599
S 10/11 27 Robin van Persie 4.79 94 1768
S 10/11 33 Didier Drogba 4.74 147 2793
S 11/12 26 Wayne Rooney 4.72 149 2839
S 13/14 28 Edin Dzeko 4.65 103 1993
S 12/13 27 Edin Dzeko 4.61 93 1816
S 11/12 28 Robin van Persie 4.6 171 3344
S 11/12 25 Luis Suárez 4.47 127 2557
S 10/11 24 Victor Obinna 4.4 77 1574
S 11/12 23 Sergio Agüero 4.39 127 2602
S 10/11 32 Louis Saha 4.33 66 1372
S 12/13 20 Romelu Lukaku 4.31 96 2006
S 10/11 29 Roman Pavlyuchenko 4.23 75 1595


Shot Assists p90

Suarez is here again x 3. Hazard x3. Van Persie x2. We also have the oddballs again: Lambert, Kalou, Obinna, Odemwingie, Zamora.

Season Age Player / Season Shot Assists p90 Shot Assists Mins
S 12/13 26 Luis Suárez 2.77 91 2955
S 14/15 24 Eden Hazard 2.69 101 3379
S 13/14 27 Luis Suárez 2.67 88 2965
S 13/14 23 Eden Hazard 2.58 83 2898
S 10/11 24 Victor Obinna 2.57 45 1574
S 10/11 25 Salomon Kalou 2.53 44 1563
S 14/15 34 Bobby Zamora 2.52 50 1783
S 14/15 26 Alexis Sánchez 2.5 82 2953
S 11/12 28 Robin van Persie 2.48 92 3344
S 10/11 34 Kevin Davies 2.44 92 3387
S 11/12 22 Gareth Bale 2.35 84 3217
S 12/13 31 Rickie Lambert 2.33 81 3133
S 10/11 33 Didier Drogba 2.32 72 2793
S 12/13 23 Gareth Bale 2.31 75 2925
S 12/13 29 Carlos Tevez 2.28 61 2410
S 10/11 29 Peter Odemwingie 2.28 68 2685
S 10/11 27 Robin van Persie 2.24 44 1768
S 10/11 30 Dimitar Berbatov 2.23 52 2096
S 11/12 25 Luis Suárez 2.22 63 2557
S 12/13 22 Eden Hazard 2.22 65 2640


Shot Contribution p90

Suarez takes the top 2 spots and  it's likely not a stretch to say he was the best player we have seen in the last 5 years of the Premier League. It's not just the fact that Suarez takes the top 2 spots, it's the gap between his best seasons and the rest of the pack. Volume matters. As does his durability.

Suarez x3. Aguero x3. 

Van Persie x2. Bale x2 (aged 22 & 23). Dzeko x2 (sub effects).



Season Age Player / Season Shot Cont p90 Shots + Shot Assists Mins
S 12/13 26 Luis Suárez 8.47 278 2955
S 13/14 27 Luis Suárez 8.17 269 2965
S 11/12 26 Edin Dzeko 7.52 125 1497
S 12/13 23 Gareth Bale 7.38 240 2925
S 13/14 25 Sergio Agüero 7.13 122 1539
S 11/12 28 Robin van Persie 7.08 263 3344
S 10/11 33 Didier Drogba 7.06 219 2793
S 10/11 27 Robin van Persie 7.02 138 1768
S 10/11 24 Victor Obinna 6.98 122 1574
S 11/12 25 Luis Suárez 6.69 190 2557
S 10/11 30 Dimitar Berbatov 6.44 150 2096
S 14/15 26 Wilfried Bony 6.42 114 1599
S 11/12 23 Sergio Agüero 6.36 184 2602
S 11/12 26 Wayne Rooney 6.34 200 2839
S 10/11 25 Salomon Kalou 6.28 109 1563
S 12/13 27 Edin Dzeko 6.24 126 1816
S 14/15 26 Sergio Agüero 6.2 175 2540
S 14/15 26 Alexis Sánchez 6.16 202 2953
S 11/12 22 Gareth Bale 6.15 220 3217
S 12/13 29 Carlos Tevez 6.12 164 2410

There's a few guys that consistently show up on all these lists. And I'll rank them (imo) in order of brilliance and value to their teams.

#1 Luis Suarez
#2 Sergio Aguero
#3 Gareth Bale
#4 Robin van Persie

These guys are the supermen of the Premier League 2010 - 2015. These players had tremendous impacts on their respective teams. #1 and #3 went to the best 2 clubs in the world. #2 is untouchable and #4 is a huge outlier - but what a talent - in that he was able to post incredible seasons during his 27/28/29 yo seasons - the autumn of his prime.

Again, these guys were the supermen of the league and nobody else got close to equaling their blend of excellence and consistency. Rare talents.

Music

Superman. Rare talent.

Play LOUD.

(for music nerds, I adore the My Bloody Valentine-type dense, muscular  tone to Billy's guitar during Siamese Era (Alan Mulder influence))



Wednesday 8 July 2015

The 2014/15 Premier League: Territory%

Quick post. 

What is Territory%?

For starters, it's a terrible name. 

The chart you see below is pretty simple. It takes all the completed passes (controlled possession) for the team in question and its opponents, and splits them into the 3 zones where the passes took place. 

Basically, I am using completed pass volume in each zone to create a territory proxy.

Defensive zone = defensive zone passes for + attacking zone passes against.
Defensive zone = neutral zone passes for + neutral zone passes against.
Attacking zone = attacking zone passes for + defensive zone passes against.


Total


The chart below is interactive and features the Territory% proxy for each team.



City, Chelsea, Arsenal & United are your top 4 in terms of attacking territory%. Burnley and QPR are in 5th and 6th place.

Sunderland and Swansea are well adrift by attacking territory%. 

Sunderland, West Brom, Everton & Swansea are the 4 teams who are somewhat adrift of the pack per the amount of time spent hemmed in their own defensive zone.


With & Without


Now, the chart above looked at the Territory% for each team using all completed passes. But that chart doesn't tell us what each team looks like by Territory% with and without the ball.

I grappled with the design of the chart below, and I still ain't happy with it.

The important thing to remember here is that the group of bars on the left (brighter colours) tells us about the territory% split for each team when IN possession of the ball. The group of bars on the right (faded colours) tells us what their opponents did with the ball (without possession of the ball)

A quick example: 

Arsenal, with the ball, had a territory split of 15.6/52.8/31.6 in the D zone/N zone/A zone. These are some good numbers.

Arsenal's opponents had a territory split of 20.4/55.7/23.8.

With the ball, 31.6% of Arsenal's completed passes were in the attacking zone. But when their opponents had the ball, Arsenal restricted them to posting just 23.8% of their completed passes in the attacking zone.

Arsenal were a good, attacking territory team with the ball. Without the ball, Arsenal were able to restrict their opponents to a much lower percentage of passes in the attacking zone.

Red = defensive zone
Yellow = neutral zone
Green = attacking zone
 

Manchester City were the most dominant territory% team both with and without the ball.

Sunderland's opponents posted a really high percentage of their passes in the attacking zone. Either Sunderland were unable to prevent themselves from being hemmed in, or the deep defensive strategy was deliberate. Ditto Everton.

Swansea posted pretty bad territory% numbers both with and without the ball.

Manchester United were quite the team at forcing their opponents into fairly harmless possession.

***

Some interesting pieces of information in the interactive charts above.  I'm not quite sure how usable territory% is but it gives us another layer of information about teams.

Total territory% is about as good a proxy as we can get for the percentage of time spent in each zone. Add in with and without the ball territory% and we have some cool info.



;