Football Analytics

bosnian_red

Worst scout to ever exist
Joined
Aug 13, 2011
Messages
58,132
Location
Canada
Analytics is obviously becoming more and more into football, whether some people like it or not. There are stats like xG, xA, ball progression stats, and so on, basically quantifying how a player performs individually in every aspect, but generally the data has been hard to get or inconvenient to read. This is a game changer now though, with FBref pretty much joining with Statsbomb and making all the data public in pretty much doing fifa style ratings for players and full scouting reports based on real life performances over time, compared as a percentile with every other player who plays a similar position.

Search a name, click "complete scouting report", pick your time frame and enjoy. They have similar players ranking, pretty much found by taking players who have a similar scouting report (the way they pass, the quality, quantity, distance, dribbling locations and success rates, actions leading to shots, goals, defensive actions, possession stats, etc). Literally anything anyone would want on a player. Many clubs were already partnered with statsbomb to get this info, but combined with fbrefs database this is now easily accessible. Can see a big switch happening into scouting using this data now. More than anything though it's cool to have it accessible to the public.

Will do an example with Rashford. Ranks in the 89th percentile among the big 5 leagues and European competition for xG for attacking mids and wingers, 92 percentile for non penalty xG, 86th percentile for passes into the box, 79th for shot creating actions through dribbling, 98th percentile for nutmegs, but only 63rd percentile in successful dribble percentage. Change his filters to compare with forwards and his numbers change to show much better passing (as he'll get involved a lot more in build up than a normal forward), but his shooting goes down in comparison.
Anyway, here's their tweet with the link. Personally think this is absolutely huge for scouting and comparing players.
 

Steve Bruce

Full Member
Joined
Aug 19, 2008
Messages
1,370
Tbh, I'm personally not into stats too much. All this xg stuff means nothing to me.

I done a comparison their between mbappe, salah, Rashford & Martial for this season.

I don't know outside of basic stats which players are performing the best.

If someone has the time to do that comparison & explain what everything means that would fantastic.
 

sullydnl

Ross Kemp's caf ID
Joined
Sep 13, 2012
Messages
34,063
I know just enough about stats to know what I really need are sources who interpret the stats intelligently, rather than access to that raw data.

xG conceded/prevented stats being a recent case in point. I've seen a few people recently point to that stat as proof that, say, Nick Pope is the best shot stopper in the league because he's prevented +5.90 or whatever on what he was expected to concede. But I'm nearly certain that because that's calculated as an absolute number it will favour goalkeepers who face more shots, which is why I currently see the likes of Pope and Areola heading that particular list. So goalkeeper A is expected to concede 20 goals and concedes 10, giving him a score of +10. Meanwhile Goalkeeper B is expected to concede 10 and concedes 4, giving him a score of +6. On face value Goalkeeper A has outperformed Goalkeeper B by +4. But in reality Goalkeeper B has prevented 60% of the xG he faced compared to Goalkeeper A's 50%, so is in fact performing better relative to the goal threat he's facing.

Another example: I've seen posters say (in relation to a James Ward-Prowse freeckick, I think) "the xG on that shot was 0.1, therefore it was an easy save". Again, as far as I can see that a hideous misunderstanding of the stat and what it's supposed to tell you.

Problem is I don't know quite enough about stats to be certain I'm right (I could well be wrong) or to argue my point effectively.

So basically when I see stats like this being made readily available my immediate assumption is that I'm going to see a lot more people whose knowledge I don't trust copying and pasting stats they don't understand into football conversations that will then become more tedious as a result.
 

TheMagicFoolBus

Full Member
Joined
Nov 5, 2016
Messages
6,612
Location
Lisboa, Portugal
Supports
Chelsea
bosnian_red said:
Analytics is obviously becoming more and more into football, whether some people like it or not. There are stats like xG, xA, ball progression stats, and so on, basically quantifying how a player performs individually in every aspect, but generally the data has been hard to get or inconvenient to read. This is a game changer now though, with FBref pretty much joining with Statsbomb and making all the data public in pretty much doing fifa style ratings for players and full scouting reports based on real life performances over time, compared as a percentile with every other player who plays a similar position.

Search a name, click "complete scouting report", pick your time frame and enjoy. They have similar players ranking, pretty much found by taking players who have a similar scouting report (the way they pass, the quality, quantity, distance, dribbling locations and success rates, actions leading to shots, goals, defensive actions, possession stats, etc). Literally anything anyone would want on a player. Many clubs were already partnered with statsbomb to get this info, but combined with fbrefs database this is now easily accessible. Can see a big switch happening into scouting using this data now. More than anything though it's cool to have it accessible to the public.

Will do an example with Rashford. Ranks in the 89th percentile among the big 5 leagues and European competition for xG for attacking mids and wingers, 92 percentile for non penalty xG, 86th percentile for passes into the box, 79th for shot creating actions through dribbling, 98th percentile for nutmegs, but only 63rd percentile in successful dribble percentage. Change his filters to compare with forwards and his numbers change to show much better passing (as he'll get involved a lot more in build up than a normal forward), but his shooting goes down in comparison.
Anyway, here's their tweet with the link. Personally think this is absolutely huge for scouting and comparing players.
Click to expand...
Great post to call attention to it - fbref.com have been the definitive source for ages, and today's update just further cements them ahead of the curve.

Stats will never be the end all be all with football because there is inherently more randomness compared to other sports, but anyone who dismisses a more complete understanding of the game via analytics just because they aren't correct 100% of the time is being closed-minded if you ask me.
 

bosnian_red

Worst scout to ever exist
Joined
Aug 13, 2011
Messages
58,132
Location
Canada
sullydnl said:
I know just enough about stats to know what I really need are sources who interpret the stats intelligently, rather than access to that raw data.

xG conceded/prevented stats being a recent case in point. I've seen a few people recently point to that stat as proof that, say, Nick Pope is the best shot stopper in the league because he's prevented +5.90 or whatever on what he was expected to concede. But I'm nearly certain that because that's calculated as an absolute number it will favour goalkeepers who face more shots, which is why I currently see the likes of Pope and Areola heading that particular list. So goalkeeper A is expected to concede 20 goals and concedes 10, giving him a score of +10. Meanwhile Goalkeeper B is expected to concede 10 and concedes 4, giving him a score of +6. On face value Goalkeeper A has outperformed Goalkeeper B by +4. But in reality Goalkeeper B has prevented 60% of the xG he faced compared to Goalkeeper A's 50%, so is in fact performing better relative to the goal threat he's facing.

Another example: I've seen posters say (in relation to a James Ward-Prowse freeckick, I think) "the xG on that shot was 0.1, therefore it was an easy save". Again, as far as I can see that a hideous misunderstanding of the stat and what it's supposed to tell you.

Problem is I don't know quite enough about stats to be certain I'm right (I could well be wrong) or to argue my point effectively.

So basically when I see stats like this being made readily available my immediate assumption is that I'm going to see a lot more people whose knowledge I don't trust copying and pasting stats they don't understand into football conversations that will then become more tedious as a result.
Click to expand...
You're 100% right, easy to take data out of context but there's a lot of great info there too. All about translating it and displaying what's useful. This source is huge as it's basically what professional clubs could have had access to (though many are in pretty novice stages with it still as it's newer in football), but now it's open to everyone.

For goalkeepers in particular, I feel like post shot xG - gA is the most accurate measure of their shot stopping. Basically totals the amount of post-shot xG (expected goals based on how likely the goalkeeper is to actually save the shot, normal xG doesn't take into account the actual finish/sooting), and then subtracts the actual goals conceded from that. That way every goalkeepers shot stopping is compared on a per shot basis.

Looking at that stat in particular, you can compare De Gea 17/18, 18/19, 19/20, 20/21 (all premier league stats) with the overall last 365 days comparing top 5 leagues player and european competitions.

De Gea 17/18 - 87th percentile (+0.21)
De Gea 18/19 - 60th percentile (+0.02)
De Gea 19/20 - 63rd percentile (+ 0.06)
De Gea 20/21 - 31st percentile (-0.06)
De Gea last 365 days - 76th percentile (+0.07)

Pope for example is at +0.31 over the past year, in the 98th percentile of shot stopping. Donnarumma 87th, Neuer 99th, Ter Stegen 90th, Ederson 78th, Allison 58th ( ! ), and so on.
 

VorZakone

What would Kenny G do?
Joined
May 9, 2013
Messages
33,077
How are these stats measured? Are people literally watching and doing data entry in some system? Oh look this player had a completed pass, add 1 to his passing stats.
 

bosnian_red

Worst scout to ever exist
Joined
Aug 13, 2011
Messages
58,132
Location
Canada
Steve Bruce said:
Tbh, I'm personally not into stats too much. All this xg stuff means nothing to me.

I done a comparison their between mbappe, salah, Rashford & Martial for this season.

I don't know outside of basic stats which players are performing the best.

If someone has the time to do that comparison & explain what everything means that would fantastic.
Click to expand...
Personally think the easiest way to look at the information is to just search each individual player, click on view complete scouting report and look at each stat by percentile ranking, rather than raw totals. You never have context with the raw totals, while the percentile ranking for each stat (let's say for Rashford and Salah, if you select a specific premier league season) is a direct comparison of all wingers/attacking midfielders in the premier league and ranking them. Radar charts are usually great at comparing players with sets like this, where you select the stats you want to look at (relevance to the position) and then just see how their maps overly. Rashford is generally 97th/98th percentile for nutmegs... but nobody would really use that stat as a direct comparison of how good a player is, right?

So you look at stuff like xG or non pen xG, xGassisted, progressive passes, passes into the penalty area, touches in the penalty box, dribbles completed and the dribble completion %, progressive carries, carries into the penalty area, progressive passing received, total shots per 90 as well as the non pen xG/shot attempt....
Basically this info can tell you:
  • xG: the quality of chances they find themselves in every game compared to other attackers
  • xGassisted: the quality of chances they create for others every game
  • progressive passes: Basically the usefulness of their passing, how frequently they pass the ball up the pitch accurately whether it's to advance 10 yards or to get into the box
  • Passes/carries/touches in pen box: Point of football is to make chances and score goals, being in the box gives you the best chance to do that. Hard to keep the ball in the box so generally the more touches you have there, the better positions you find yourself in and the more involved you are in the key moments. Whether it's receiving the ball in the box, dribbling it in, passing it in (those are successful totals, not just lumps to nobody)... that's what you want from your attackers
  • Dribbles completed and the percentage: Obvious taking both success rate of how good their dribbling is and bouncing it off of the frequency to see if they are dribblers or just sparingly dribble but do it well when they do
  • Non pen xG/shot attempt and total shots per 90 - Basically anyone can take 5 shots per game from 40 yards out, but that's kind of pointless, so the non-pen xG/shot basically narrows it down into decision making with their shooting, the higher the xG per shot attempt generally means they are taking good shots and not wasting time (but also important to look at quantity to make sure they are still shooting, otherwise you get a Martial scenario where he rarely shoots but always has a very high xG/shot attempt).
Basically all sorts of stats though that do an excellent job at comparing the strengths and weaknesses of players. Bruno vs Grealish is one you see a lot on this forum and this shows their comparison really well (just using the general stats). Bruno being involved a lot more in goals, similar creativity, Bruno more through receiving advanced passes and playing passes forward, Grealish through bringing the ball into the box himself. Bruno far more involved defensively than Grealish is. Breaking it down further with their complete report gives you a better comparison too.
 

bosnian_red

Worst scout to ever exist
Joined
Aug 13, 2011
Messages
58,132
Location
Canada
VorZakone said:
How are these stats measured? Are people literally watching and doing data entry in some system? Oh look this player had a completed pass, add 1 to his passing stats.
Click to expand...
Always wondered this myself. Definitely used to be like this at first with the simpler stats like passes completed/missed where it's easy to tick, but when you're now talking about every little action/type of shot/positioning of defenders/shot placement and power for a variety of stats... It'd be impossible to do without programming.
 

kundalini

Full Member
Joined
Aug 26, 2004
Messages
5,756
VorZakone said:
How are these stats measured? Are people literally watching and doing data entry in some system? Oh look this player had a completed pass, add 1 to his passing stats.
Click to expand...
Statsbomb recently bought an Egyptian company called Arqam that had been doing the raw data for them. People watch each match and count a billion different events. Where was the shot taken from ? How many defenders between the shot taker and the goal ? Where is the keeper stood ? What happened just before the shot ? What height was the ball off the ground when the shot or header happened ? Each year Statsbomb add more detail. Pressures is a big thing for them. It must take hours to do each match.

This is what their website says:

StatsBomb Data
Our unique event data collection spec has over 3,400 events per match of on and off the ball data including pressures, ball carries, possession chains and more. Data generated from a blend of Computer Vision and human driven collection with automated validation checks and a highly experienced quality assurance team, makes it the most accurate event data in the industry
 
Last edited:

Stack

Leave Women's Football Alone!!!
Joined
Sep 6, 2006
Messages
13,339
Location
Auckland New Zealand
bosnian_red said:
Analytics is obviously becoming more and more into football, whether some people like it or not. There are stats like xG, xA, ball progression stats, and so on, basically quantifying how a player performs individually in every aspect, but generally the data has been hard to get or inconvenient to read. This is a game changer now though, with FBref pretty much joining with Statsbomb and making all the data public in pretty much doing fifa style ratings for players and full scouting reports based on real life performances over time, compared as a percentile with every other player who plays a similar position.

Search a name, click "complete scouting report", pick your time frame and enjoy. They have similar players ranking, pretty much found by taking players who have a similar scouting report (the way they pass, the quality, quantity, distance, dribbling locations and success rates, actions leading to shots, goals, defensive actions, possession stats, etc). Literally anything anyone would want on a player. Many clubs were already partnered with statsbomb to get this info, but combined with fbrefs database this is now easily accessible. Can see a big switch happening into scouting using this data now. More than anything though it's cool to have it accessible to the public.

Will do an example with Rashford. Ranks in the 89th percentile among the big 5 leagues and European competition for xG for attacking mids and wingers, 92 percentile for non penalty xG, 86th percentile for passes into the box, 79th for shot creating actions through dribbling, 98th percentile for nutmegs, but only 63rd percentile in successful dribble percentage. Change his filters to compare with forwards and his numbers change to show much better passing (as he'll get involved a lot more in build up than a normal forward), but his shooting goes down in comparison.
Anyway, here's their tweet with the link. Personally think this is absolutely huge for scouting and comparing players.
Click to expand...
Do you know if there are any scouting platforms that look at body shape when receiving the ball, receiving on front foot, receiving on back foot, inside turns, outside turns, receiving on the half turn, receiving on the full turn, making space under pressure for passes, making space to receive, movement creating inside passing channels, movement creating outside passing channels, underlapping runs, overlapping runs, body shape when closing down opposition players, angled closing down runs, curved closing down runs, holding out of channels, holding out of spaces, creating space for team mates with movement, weight of passes, driven passes, delayed passes, chipped passes etc etc etc
Those are the things coaches are looking at when assessing players and the things coaches want to hear from scouts.
 

AkaAkuma

Full Member
Joined
May 15, 2012
Messages
3,203
Stack said:
Do you know if there are any scouting platforms that look at body shape when receiving the ball, receiving on front foot, receiving on back foot, inside turns, outside turns, receiving on the half turn, receiving on the full turn, making space under pressure for passes, making space to receive, movement creating inside passing channels, movement creating outside passing channels, underlapping runs, overlapping runs, body shape when closing down opposition players, angled closing down runs, curved closing down runs, holding out of channels, holding out of spaces, creating space for team mates with movement, weight of passes, driven passes, delayed passes, chipped passes etc etc etc
Those are the things coaches are looking at when assessing players and the things coaches want to hear from scouts.
Click to expand...
If it can be accessed the it can be recorded.
 

Steve Bruce

Full Member
Joined
Aug 19, 2008
Messages
1,370
bosnian_red said:
Personally think the easiest way to look at the information is to just search each individual player, click on view complete scouting report and look at each stat by percentile ranking, rather than raw totals. You never have context with the raw totals, while the percentile ranking for each stat (let's say for Rashford and Salah, if you select a specific premier league season) is a direct comparison of all wingers/attacking midfielders in the premier league and ranking them. Radar charts are usually great at comparing players with sets like this, where you select the stats you want to look at (relevance to the position) and then just see how their maps overly. Rashford is generally 97th/98th percentile for nutmegs... but nobody would really use that stat as a direct comparison of how good a player is, right?

So you look at stuff like xG or non pen xG, xGassisted, progressive passes, passes into the penalty area, touches in the penalty box, dribbles completed and the dribble completion %, progressive carries, carries into the penalty area, progressive passing received, total shots per 90 as well as the non pen xG/shot attempt....
Basically this info can tell you:
  • xG: the quality of chances they find themselves in every game compared to other attackers
  • xGassisted: the quality of chances they create for others every game
  • progressive passes: Basically the usefulness of their passing, how frequently they pass the ball up the pitch accurately whether it's to advance 10 yards or to get into the box
  • Passes/carries/touches in pen box: Point of football is to make chances and score goals, being in the box gives you the best chance to do that. Hard to keep the ball in the box so generally the more touches you have there, the better positions you find yourself in and the more involved you are in the key moments. Whether it's receiving the ball in the box, dribbling it in, passing it in (those are successful totals, not just lumps to nobody)... that's what you want from your attackers
  • Dribbles completed and the percentage: Obvious taking both success rate of how good their dribbling is and bouncing it off of the frequency to see if they are dribblers or just sparingly dribble but do it well when they do
  • Non pen xG/shot attempt and total shots per 90 - Basically anyone can take 5 shots per game from 40 yards out, but that's kind of pointless, so the non-pen xG/shot basically narrows it down into decision making with their shooting, the higher the xG per shot attempt generally means they are taking good shots and not wasting time (but also important to look at quantity to make sure they are still shooting, otherwise you get a Martial scenario where he rarely shoots but always has a very high xG/shot attempt).
Basically all sorts of stats though that do an excellent job at comparing the strengths and weaknesses of players. Bruno vs Grealish is one you see a lot on this forum and this shows their comparison really well (just using the general stats). Bruno being involved a lot more in goals, similar creativity, Bruno more through receiving advanced passes and playing passes forward, Grealish through bringing the ball into the box himself. Bruno far more involved defensively than Grealish is. Breaking it down further with their complete report gives you a better comparison too.
Click to expand...
Great Post, certainly makes more sense to me.
 

Classical Mechanic

Full Member
Joined
Aug 25, 2014
Messages
35,216
Location
xG Zombie Nation
Stack said:
Do you know if there are any scouting platforms that look at body shape when receiving the ball, receiving on front foot, receiving on back foot, inside turns, outside turns, receiving on the half turn, receiving on the full turn, making space under pressure for passes, making space to receive, movement creating inside passing channels, movement creating outside passing channels, underlapping runs, overlapping runs, body shape when closing down opposition players, angled closing down runs, curved closing down runs, holding out of channels, holding out of spaces, creating space for team mates with movement, weight of passes, driven passes, delayed passes, chipped passes etc etc etc
Those are the things coaches are looking at when assessing players and the things coaches want to hear from scouts.
Click to expand...
I think every club of any worth in the world uses data driven player recruitment now. One of the most popular platforms is Whyscout, they also use video clips of players to add context to the data. Willian's dad said his son uses the platform to assess his own performances so some players are using it too.

https://wyscout.com/

Mason Mount made their PL team of the year last season so safe to say their data model picks up on things that most casual watchers miss.
 

Borys

Statistics Wizard
Joined
May 10, 2013
Messages
9,135
Location
Bielsko Biala, Poland
sullydnl said:
I know just enough about stats to know what I really need are sources who interpret the stats intelligently, rather than access to that raw data.

xG conceded/prevented stats being a recent case in point. I've seen a few people recently point to that stat as proof that, say, Nick Pope is the best shot stopper in the league because he's prevented +5.90 or whatever on what he was expected to concede. But I'm nearly certain that because that's calculated as an absolute number it will favour goalkeepers who face more shots, which is why I currently see the likes of Pope and Areola heading that particular list. So goalkeeper A is expected to concede 20 goals and concedes 10, giving him a score of +10. Meanwhile Goalkeeper B is expected to concede 10 and concedes 4, giving him a score of +6. On face value Goalkeeper A has outperformed Goalkeeper B by +4. But in reality Goalkeeper B has prevented 60% of the xG he faced compared to Goalkeeper A's 50%, so is in fact performing better relative to the goal threat he's facing.

Another example: I've seen posters say (in relation to a James Ward-Prowse freeckick, I think) "the xG on that shot was 0.1, therefore it was an easy save". Again, as far as I can see that a hideous misunderstanding of the stat and what it's supposed to tell you.

Problem is I don't know quite enough about stats to be certain I'm right (I could well be wrong) or to argue my point effectively.

So basically when I see stats like this being made readily available my immediate assumption is that I'm going to see a lot more people whose knowledge I don't trust copying and pasting stats they don't understand into football conversations that will then become more tedious as a result.
Click to expand...
First 5 GK in terms of PSxG-GA are: 1. Pope 2. Areola 3. Martinez 4. Sanchez 5. Lloris. If you sort them by SoTA (shots on target against): 6. Pope 7. Areola 8. Martinez 11. Lloris 20. Sanchez. At the same time, 4 worst players on PSxG-GA are: 1. Patricio 2. McCarthy 3. Ryan 4. Pickford, sorted by SoTA ascending: 1. Ryan 7. Pickford 8. Patricio 13. McCarthy .
So based on that, there is no correlation between number of shots against and PSxG-GA score.

However, you can sort players by PSxG-GA/90' and you will see that actually Sanchez (4th in total PSxG-GA) tops that parameter with score of 0,34 - because he faced less shots. It doesn't change the table completely but you are right, this reflects the shot stopping ability better because it eliminates the factor of number of shots against.

EDIT: To be clear - all explained on PSxG-GA parameter rather than PSxG obviously.
 
Last edited:

Classical Mechanic

Full Member
Joined
Aug 25, 2014
Messages
35,216
Location
xG Zombie Nation

Pogue Mahone

The caf's Camus.
Joined
Feb 22, 2006
Messages
134,197
Location
"like a man in silk pyjamas shooting pigeons
Classical Mechanic said:
@Pogue Mahone

The new FBRef scouting tool allows us to easily look at our forward's defence stats (among many others stat) in terms of performance against every player in their position(s) in Europe's top 5 leagues.

https://fbref.com/en/players/a1d5bd30/Marcus-Rashford

https://fbref.com/en/players/8b788c01/Anthony-Martial

Compared to Cavani

https://fbref.com/en/players/527f063d/Edinson-Cavani
Click to expand...
Holy shit. That’s disgraceful. Even more so when you consider wide forwards are generally supposed to do a bit more defensive work. Must be an absolute fecking nightmare being our fullbacks.
 

roonster09

Hercule Poirot of the scouting world
Scout
Joined
May 10, 2009
Messages
36,789
Classical Mechanic said:
@Pogue Mahone

The new FBRef scouting tool allows us to easily look at our forward's defence stats (among many others stat) in terms of performance against every player in their position(s) in Europe's top 5 leagues.

https://fbref.com/en/players/a1d5bd30/Marcus-Rashford

https://fbref.com/en/players/8b788c01/Anthony-Martial

Compared to Cavani

https://fbref.com/en/players/527f063d/Edinson-Cavani
Click to expand...
It looks bad, i checked the same stats for Mane, Salah, Sterling. They are all similar to Rashford's and Martial's stats.
 

Raees

Pythagoras in Boots
Joined
May 16, 2009
Messages
29,469
Damn, this looks epic thanks for the heads up! Just wish the comparison tool also compared the percentiles in a 1 v 1 setting so it could be in a fancy coloured table.
 

roonster09

Hercule Poirot of the scouting world
Scout
Joined
May 10, 2009
Messages
36,789
Pogue Mahone said:
They are all better though. Not as better as you’d think. Makes Cavani’s effort look even more impressive, mind you.
Click to expand...
Yeah they are all better. I thought Mane, Salah and Sterling would be way higher considering they press a lot as a team.
 

Classical Mechanic

Full Member
Joined
Aug 25, 2014
Messages
35,216
Location
xG Zombie Nation

bosnian_red

Worst scout to ever exist
Joined
Aug 13, 2011
Messages
58,132
Location
Canada
Borys said:
First 5 GK in terms of PSxG are: 1. Pope 2. Areola 3. Martinez 4. Sanchez 5. Lloris. If you sort them by SoTA (shots on target against): 6. Pope 7. Areola 8. Martinez 11. Lloris 20. Sanchez. At the same time, 4 worst players on PSxG are: 1. Patricio 2. McCarthy 3. Ryan 4. Pickford, sorted by SoTA ascending: 1. Ryan 7. Pickford 8. Patricio 13. McCarthy .
So based on that, there is no correlation between number of shots against and PSxG score.

However, you can sort players by PSxG/90' and you will see that actually Sanchez (4th in total PSxG) tops that parameter with score of 0,34 - because he faced less shots. It doesn't change the table completely but you are right, this reflects the shot stopping ability better because it eliminates the factor of number of shots against.
Click to expand...
Take the PSxG-GA stat instead, as it'll be the figure of post shot expected goals minus the actual goals conceded. So the higher the number, the better the shot stopper. Just post shot xG would just show what goalkeeper should've conceded the most goals as they faced the best shots, nothing to do with their actual shot stopping.
 

bosnian_red

Worst scout to ever exist
Joined
Aug 13, 2011
Messages
58,132
Location
Canada
Raees said:
Damn, this looks epic thanks for the heads up! Just wish the comparison tool also compared the percentiles in a 1 v 1 setting so it could be in a fancy coloured table.
Click to expand...
Yeah! Statsbomb loves doing radar charts to compare players so to get the Grealish/Bruno comparison above I just copy/pasted the data in excel and pulled the specific stats into a radar chart. Doesnt take long. But would be nice if for their next step they made it so you can pick your categories and they auto populate the chart.
 

Physiocrat

Has No Mates
Joined
Jun 29, 2010
Messages
8,978
Has anyone a link to all the definitions used for the site?

In particular the expected goals etc
 

Borys

Statistics Wizard
Joined
May 10, 2013
Messages
9,135
Location
Bielsko Biala, Poland
bosnian_red said:
Take the PSxG-GA stat instead, as it'll be the figure of post shot expected goals minus the actual goals conceded. So the higher the number, the better the shot stopper. Just post shot xG would just show what goalkeeper should've conceded the most goals as they faced the best shots, nothing to do with their actual shot stopping.
Click to expand...
Yeah that's what I used, post corrected.
 

giorno

boob novice
Joined
Jul 20, 2016
Messages
26,797
Supports
Real Madrid
Stack said:
Do you know if there are any scouting platforms that look at body shape when receiving the ball, receiving on front foot, receiving on back foot, inside turns, outside turns, receiving on the half turn, receiving on the full turn, making space under pressure for passes, making space to receive, movement creating inside passing channels, movement creating outside passing channels, underlapping runs, overlapping runs, body shape when closing down opposition players, angled closing down runs, curved closing down runs, holding out of channels, holding out of spaces, creating space for team mates with movement, weight of passes, driven passes, delayed passes, chipped passes etc etc etc
Those are the things coaches are looking at when assessing players and the things coaches want to hear from scouts.
Click to expand...
Platforms? I don't think so. Clubs own internal departments likely do, at least the more advanced and well funded ones(like liverpool's). Statsbomb might very well have done it for clubs they're partnered with, too. But obviously that stuff will be protected under a non-disclosure agreement, so won't be made public
 

Raees

Pythagoras in Boots
Joined
May 16, 2009
Messages
29,469
bosnian_red said:
Yeah! Statsbomb loves doing radar charts to compare players so to get the Grealish/Bruno comparison above I just copy/pasted the data in excel and pulled the specific stats into a radar chart. Doesnt take long. But would be nice if for their next step they made it so you can pick your categories and they auto populate the chart.
Click to expand...
Ah sweet good shout that.
 

VorZakone

What would Kenny G do?
Joined
May 9, 2013
Messages
33,077
kundalini said:
Statsbomb recently bought an Egyptian company called Arqam that had been doing the raw data for them. People watch each match and count a billion different events. Where was the shot taken from ? How many defenders between the shot taker and the goal ? Where is the keeper stood ? What happened just before the shot ? What height was the ball off the ground when the shot or header happened ? Each year Statsbomb add more detail. Pressures is a big thing for them. It must take hours to do each match.

This is what their website says:

StatsBomb Data
Our unique event data collection spec has over 3,400 events per match of on and off the ball data including pressures, ball carries, possession chains and more. Data generated from a blend of Computer Vision and human driven collection with automated validation checks and a highly experienced quality assurance team, makes it the most accurate event data in the industry
Click to expand...
The sheer fact that there's so much human viewing involved does make me a bit sceptical. Are these people who watch the games 100% focused on each event? Is the way of entrying data foolproof?

I know from experience how people mismanage simple Excel sheets.
 

Grande

Full Member
Joined
Jun 22, 2007
Messages
6,377
Location
The Land of Do-What-You-Will
bosnian_red said:
You're 100% right, easy to take data out of context but there's a lot of great info there too. All about translating it and displaying what's useful. This source is huge as it's basically what professional clubs could have had access to (though many are in pretty novice stages with it still as it's newer in football), but now it's open to everyone.

For goalkeepers in particular, I feel like post shot xG - gA is the most accurate measure of their shot stopping. Basically totals the amount of post-shot xG (expected goals based on how likely the goalkeeper is to actually save the shot, normal xG doesn't take into account the actual finish/sooting), and then subtracts the actual goals conceded from that. That way every goalkeepers shot stopping is compared on a per shot basis.

Looking at that stat in particular, you can compare De Gea 17/18, 18/19, 19/20, 20/21 (all premier league stats) with the overall last 365 days comparing top 5 leagues player and european competitions.

De Gea 17/18 - 87th percentile (+0.21)
De Gea 18/19 - 60th percentile (+0.02)
De Gea 19/20 - 63rd percentile (+ 0.06)
De Gea 20/21 - 31st percentile (-0.06)
De Gea last 365 days - 76th percentile (+0.07)

Pope for example is at +0.31 over the past year, in the 98th percentile of shot stopping. Donnarumma 87th, Neuer 99th, Ter Stegen 90th, Ederson 78th, Allison 58th ( ! ), and so on.
Click to expand...
I think PSxG-GA/90 is even more precise. Can you fin how Dean Henderson has done the past seasons compared to those other keepers in that respect in the stats?
 

bosnian_red

Worst scout to ever exist
Joined
Aug 13, 2011
Messages
58,132
Location
Canada
Grande said:
I think PSxG-GA/90 is even more precise. Can you fin how Dean Henderson has done the past seasons compared to those other keepers in that respect in the stats?
Click to expand...
The stats are per 90 already!
Dean Henderson shot stopping wise:
  • 19/20 prem - +0.14 (77th percentile)
  • Last 365 days big 5 leagues and European comps - -0.04 (41st) ... sample size for this one would be really small and hurt by a few mistakes he made in the small sample
His only 2 underlying stats that are 90+ are the "launch %" (percentage of kicks longer than 40 yards) and the average pass length... so for hoofing.
Comparing De Gea and him for crosses stopped %, De Gea last season was at 5.3% (27th percentile), Henderson at 7.9% (79th percentile). De Gea this season down in the 7th percentile though, which makes sense as it's been awful this year.

Now there isn't a stat for what counts as a stoppable cross, so could be bad luck/low sample size... but I doubt it. Every player has the same bias here so it wouldn't affect it that much.

Also a thing to note - comparing individual seasons is league specific, comparing past 365 days makes it compare all top 5 leagues. So play styles across competitions would influence it a lot more, significantly harder to stop crosses in the premier league probably with the more physical style (and as we've seen so many goalkeepers struggle in that aspect when they come here).
 
You must log in or register to reply here.