Conclusions from Bolton

[HodgysBrokenThumb]

Another naive question, if you have time to continue answering, Blades Analytics . How are goalkeeping stats interpreted? The much-vaunted clean sheets are down to most/all of the team; the number of saves covers the routine to the brilliant; one 25-yard shot can dip and bend, another go in a straight-line; is it harder to keep goal for the bottom team in league 2, against poorer attackers but behind weaker defenders, than for Man City, against better attackers but who perhaps hardly get into the Man City half? As I always like to claim, are goalkeepers a different breed in a unique role, where stats are less useful?
Thanks for your posts - always good to have a thread which makes us look at things in a different way.

 

[Blades Analytics]

Hi BA, I thought that the xG was a percentage. i.e. 1 = 100% how do we then get 1.5? I know that footballers talk about 110% but this is 150%. I realise it's not a typo and just wondered how this worked?

xG is not a percentage as such. Read this mate and if still doesn't make sense let me know?

https://www.uptheblades.com/latest-articles/2018/8/22/explaining-expected-goals

 

[Bergen Blade]

Extremely interesting read. Bergen Blade is going to love it!

Yes, really enjoy the Conclusions articles and I'm very grateful for the work of Blades Analytics and Beans. For years I've felt it's been unfair that fans of bigger clubs have had more access to live games, highlights, more tactical focus and in depth analyses than we have.

Not just enjoying the stats, but also the highlighting of Stevens benefiting from better playmaking from central and possibly some tactical tweaks, i.e. when a bit more interpretation goes along with the numbers, and looking at complementary skills and the style and general balance of the team.

I have a question regarding analysing from recorded/filmed games. It's been a bugbear of mine that BladesPlayer and also iFollow often tend to zoom in on the man on the ball far too much. Meaning we often see this:



Rather than this:





(Sorry they're old screenshots - I just took some from a former moan) In the first screen it's impossible to see the passing options, the runs, the non-runs, the marking, the defensive line, etc. Do professional football league analysts have access to different videos, or do they have to make do with the same we (often) get on BladesPlayer/iFollow/EFL highlights in general. Premier League or Sky matches aren't filmed like this, so why do EFL camera men insist on doing it?

 

[Blades Analytics]

Another naive question, if you have time to continue answering, Blades Analytics . How are goalkeeping stats interpreted? The much-vaunted clean sheets are down to most/all of the team; the number of saves covers the routine to the brilliant; one 25-yard shot can dip and bend, another go in a straight-line; is it harder to keep goal for the bottom team in league 2, against poorer attackers but behind weaker defenders, than for Man City, against better attackers but who perhaps hardly get into the Man City half? As I always like to claim, are goalkeepers a different breed in a unique role, where stats are less useful?
Thanks for your posts - always good to have a thread which makes us look at things in a different way.

Great question, and GKs are far from less useful stats, i'd argue they are actually becoming one of the best covered positions!

you're completley correct in what you said. However we actually look at other GK stats, expected saves (xS), Save efficiency % which is a mathmatical equation involving shots on target, saves and goals and there is new research looking into positioning, distribution (as it becomes a big part at top level). The best one is expected saves or at least expected goals saved, as that takes into account shot placement, shot quality and distance and angle from goal.

 

[Blades Analytics]

Yes, really enjoy the Conclusions articles and I'm very grateful for the work of Blades Analytics and Beans. For years I've felt it's been unfair that fans of bigger clubs have had more access to live games, highlights, more tactical focus and in depth analyses than we have.

Not just enjoying the stats, but also the highlighting of Stevens benefiting from better playmaking from central and possibly some tactical tweaks, i.e. when a bit more interpretation goes along with the numbers, and looking at complementary skills and the style and general balance of the team.

I have a question regarding analysing from recorded/filmed games. It's been a bugbear of mine that BladesPlayer and also iFollow often tend to zoom in on the man on the ball far too much. Meaning we often see this:

View attachment 43603

Rather than this:

View attachment 43604



(Sorry they're old screenshots - I just took some from a former moan) In the first screen it's impossible to see the passing options, the runs, the non-runs, the marking, the defensive line, etc. Do professional football league analysts have access to different videos, or do they have to make do with the same we (often) get on BladesPlayer/iFollow/EFL highlights in general. Premier League or Sky matches aren't filmed like this, so why do EFL camera men insist on doing it?

So professional analysts even down to national league level have cameras setup with a full pitch view. In my work with some agencies and scouts i have access to some of the professional software tools and i can select 2/3 different pitch views or angles. Obviously the higher the club the more options they will have. Most clubs now will be doing analysis real time in the stands to give managers better information at half time and for tactical tweaks.

Must say from years ago i always appreciated your analysis and views Bergen Blade you do a fantastic job and will still do so! hopefully we can add to that too

 

[Rodley]

Yes, really enjoy the Conclusions articles and I'm very grateful for the work of Blades Analytics and Beans. For years I've felt it's been unfair that fans of bigger clubs have had more access to live games, highlights, more tactical focus and in depth analyses than we have.

Not just enjoying the stats, but also the highlighting of Stevens benefiting from better playmaking from central and possibly some tactical tweaks, i.e. when a bit more interpretation goes along with the numbers, and looking at complementary skills and the style and general balance of the team.

I have a question regarding analysing from recorded/filmed games. It's been a bugbear of mine that BladesPlayer and also iFollow often tend to zoom in on the man on the ball far too much. Meaning we often see this:

View attachment 43603

Rather than this:

View attachment 43604



(Sorry they're old screenshots - I just took some from a former moan) In the first screen it's impossible to see the passing options, the runs, the non-runs, the marking, the defensive line, etc. Do professional football league analysts have access to different videos, or do they have to make do with the same we (often) get on BladesPlayer/iFollow/EFL highlights in general. Premier League or Sky matches aren't filmed like this, so why do EFL camera men insist on doing it?

Bloody hell, if Blades Analytics and Beans get Bergen Blade involved in some form of analysis of tactical systems / roles, that would be some fantastic stuff! Surely a collaboration is in order?

 

[ucandomagic]

This stuff has really interested me for a while. It isn’t just the stats, but the reason that the stats are important that is fascinating. The problem is that however analytical we think a scout or a coach is, he is human and through just our own eyes what we humans do is quickly form an opinion and subconsciously look for reasons to justify that opinion.

Michael Lewis’s book Moneyball highlighted this in one sport - baseball and how gathering cold stats is a huge addition to our ability to assess. Interestingly, it is the second tier clubs that benefit most - because the top clubs buy the best and just throw them away if it doesn’t work. The lower clubs, however, are piecing together teams from players who generally each have a weakness that has kept them out of the top clubs but combined carefully these weaknesses can be cancelled.

Lewis’s most recent book The Undoing Project goes back to identify that all of this was proven by 2 Israeli psychologists and was first used with huge success to select Tank Commanders in the Israeli army. It then moved into economics - earning them a Nobel prize.

Keep it coming BA - this stuff is why it is not impossible that we could do a Leicester.

Magic Takes Time but Needs Good Data!

 

[SwissBlade]

Basic question I know, but I couldn't immediately find it on Google: Do players wear GPS during games?

Yes, as Danny04 hall said they are like a bra and consist of GPS, Heart Rate monitors and i think accelerometers.

We've had the vests in training since Clough was in charge and we used to use Catapult vests going back to Clough and Adkins.

https://www.catapultsports.com/

However, they weren't always allowed during games, just training. They were allowed in the EFL from 2015/ 16 season as FIFA only approved them in Feb 2015

https://www.bbc.com/sport/football/34267968

https://www.fifa.com/about-fifa/new...r-electronic-performance-and-tra-2564815.html

Important to remember that ALL clubs use it, therefore Cloughie et al would also have used it if available, and frankly I dont care how 'all consuming' it becomes to Wilder as long as results on the pitch keep going on the same upward curve. Having said that he's far too well grounded to become 'all consumed' by anything.

Blades Analytics may correct me on this, however, whilst the data available and performance analysis was improving in Cloughs reign at United, I don't think that we had a large budget for the Sports Science side and particularly the Performance analysis as the data was probably not recorded as accurately as it is now.

Adkins talked openly about using the data from the vests (distance covered, sprints, Heart Rate Zones etc) after the excellent display in the FA Cup against Man United, i think in relation to the next game which was midweek. His comments related to exertions in that game and how this would affect their recovery and preparation for the next game. We drew the next game 3-3 with Wigan having being 3-0 down.

Wilder has pretty much said from the start and reiterated this at the end of the promotion season that he is "old school". He certainly has some old school principles in how he manages players and i think with regards to the performance analysis, he will undoubtedly just get an overview of the results. But the way that he has embraced the Sports Science and player conditioning aspects shows that he is kidding us a little bit.

Blades Analytics Conclusions from Bolton talk a lot about Stevens and his positioning in the formation. This won't have been by accident. Yes Wilder and Knill will have seen for themselves that the form of Stevens hasn't been great recently, however, what the data does is allows them to see very quickly where the weaknesses in Stevens game have been.

Interestingly Stevens pushed inside and played a bigger part in the central midfield, this isn't immediately obvious during the game (note the FIFA/ EFL didn't approve use of data during games), but in the types of analysis done on player movement, its strikingly clear.

Most of the game Stevens played inside, however for the fleck goal, his movement and the ball from McGoldrick was played into space where you'd expect a winger/ wingback to be. The ball in from Stevens to fleck was not hit with too much power, height, just played in to fleck to finish with ease.

I'd love to see the average positions for Bolton during the game, just to see how much space there was in that left corner where Stevens got into.

You can have all the data you want, but its really how you use it which can be the difference

 

[Blades Analytics]

Yes, as Danny04 hall said they are like a bra and consist of GPS, Heart Rate monitors and i think accelerometers.

We've had the vests in training since Clough was in charge and we used to use Catapult vests going back to Clough and Adkins.

https://www.catapultsports.com/

However, they weren't always allowed during games, just training. They were allowed in the EFL from 2015/ 16 season as FIFA only approved them in Feb 2015

https://www.bbc.com/sport/football/34267968

https://www.fifa.com/about-fifa/new...r-electronic-performance-and-tra-2564815.html



Blades Analytics may correct me on this, however, whilst the data available and performance analysis was improving in Cloughs reign at United, I don't think that we had a large budget for the Sports Science side and particularly the Performance analysis as the data was probably not recorded as accurately as it is now.

Adkins talked openly about using the data from the vests (distance covered, sprints, Heart Rate Zones etc) after the excellent display in the FA Cup against Man United, i think in relation to the next game which was midweek. His comments related to exertions in that game and how this would affect their recovery and preparation for the next game. We drew the next game 3-3 with Wigan having being 3-0 down.

Wilder has pretty much said from the start and reiterated this at the end of the promotion season that he is "old school". He certainly has some old school principles in how he manages players and i think with regards to the performance analysis, he will undoubtedly just get an overview of the results. But the way that he has embraced the Sports Science and player conditioning aspects shows that he is kidding us a little bit.

Blades Analytics Conclusions from Bolton talk a lot about Stevens and his positioning in the formation. This won't have been by accident. Yes Wilder and Knill will have seen for themselves that the form of Stevens hasn't been great recently, however, what the data does is allows them to see very quickly where the weaknesses in Stevens game have been.

Interestingly Stevens pushed inside and played a bigger part in the central midfield, this isn't immediately obvious during the game (note the FIFA/ EFL didn't approve use of data during games), but in the types of analysis done on player movement, its strikingly clear.

Most of the game Stevens played inside, however for the fleck goal, his movement and the ball from McGoldrick was played into space where you'd expect a winger/ wingback to be. The ball in from Stevens to fleck was not hit with too much power, height, just played in to fleck to finish with ease.

I'd love to see the average positions for Bolton during the game, just to see how much space there was in that left corner where Stevens got into.

You can have all the data you want, but its really how you use it which can be the difference

Some excellent salient points in there! just to kind of chip in and address a few of those.

Wilder "old school" - you're exactly right. Wilder wants to come across as old school to us fans for that blood and thunder approach we love about him. However, no fan should be under any impression that he is only old school. Listen to any CW pre season interview and you'll notice he says the words looked at the stats in about 90% of the interviews he did. CW and AK look at performance on the ball data and the more fitness type data CONSTANTLY.

The Stevens point is excellent too. That is minor tactical tweak but the way it was carried out was akin to a real element of why we won in the game (interestingly i have included their average positioning in the game in the write up so re-check that out if you missed it). It's these kind of small tactical tweaks that evolve the different dimensions of our game, evolution not revolution. CW alluded to it very well in an interview last week where he was discussing a small tactical "thing" we do in games, and that both him and Knilly had seen opposition repeat this in recent games, so it had been picked up on. That's the job of a performance analyst, stats are only part of the picture but the real job is to analyse opposition or our own team to the tightest of margins to ensure we understand what they do well and how they do it, in my experience statistics, heat maps, touch maps etc can be a very good indicator of just that. Loving the discussion by the way anymore questions please feel free to fire away!

 

[SwissBlade]

The Stevens point is excellent too. That is minor tactical tweak but the way it was carried out was akin to a real element of why we won in the game (interestingly i have included their average positioning in the game in the write up so re-check that out if you missed it). It's these kind of small tactical tweaks that evolve the different dimensions of our game, evolution not revolution. CW alluded to it very well in an interview last week where he was discussing a small tactical "thing" we do in games, and that both him and Knilly had seen opposition repeat this in recent games, so it had been picked up on. That's the job of a performance analyst, stats are only part of the picture but the real job is to analyse opposition or our own team to the tightest of margins to ensure we understand what they do well and how they do it, in my experience statistics, heat maps, touch maps etc can be a very good indicator of just that. Loving the discussion by the way anymore questions please feel free to fire away!

Cheers for that, I meant it would've been good to see the Bolton positioning side by side with the Blades positioning to see the space in behind Bolton. From the Blades one there looks like almost an entire vertical third (the left wing) of the Bolton pitch which was wide open for exploiting. Obviously this is average positioning so it wasn't left wide open all game. But i'd expect McGoldrick was given the instruction before he came on - get the ball into space on the left for Enda"



And something else which interests me, I bet the positioning map for the Blades v Pigs game at the lane last season didn't look anything like this!

 

[ucandomagic]

BA - as stats enable us to compare players’ performances do you have the data as to why teams like Brighton and Fulham have not kept a player like Ollie Norwood on their books? What have the players that have replaced him got that is better than him? To us he seems perfect, but is that because the players around need his strengths to overcome their weaknesses and in a better team his strengths are less important. Is he too slow, is he too 1-footed - always uses outside of right instead of left?

I’d be very interested to know your thoughts.

 

[I should coco]

It's been a bugbear of mine that BladesPlayer and also iFollow often tend to zoom in on the man on the ball far too much.

It has ever been thus with all televised games. The camera follows the man with the ball, so the midfielder who makes a lung bursting 60 yard run but doesn't get used gets no credit. This is why the MotM on live games is nearly always a scorer or a keeper who makes some good saves. You just can't appreciate what the the other players are doing off the ball.

 

[Blades Analytics]

BA - as stats enable us to compare players’ performances do you have the data as to why teams like Brighton and Fulham have not kept a player like Ollie Norwood on their books? What have the players that have replaced him got that is better than him? To us he seems perfect, but is that because the players around need his strengths to overcome their weaknesses and in a better team his strengths are less important. Is he too slow, is he too 1-footed - always uses outside of right instead of left?

I’d be very interested to know your thoughts.

I think in Norwood's case it was a simple case of Fulham signing numerous midfielders in the summer and he just got further and further back in their plans, same with his parent club Brighton. Fulham had the option of signing Seri from Nice who 12 months ago was minutes away from signing for Barcelona. Now i love ON and his style but let's be clear here he is a top level championship player, perhaps bottom end PL player. Fulham's recruitment has been based on building a top 12 side in the PL already! Although if you ask the more knowledgeable Fulham fans you find that really liked and appreciated what Norwood bought. I think same story can be applied to Brighton, Hughton has his favourites and some of their new signings last year in that area of the team really hit the ground running. I don't think their is a specific technical reason why Norwood hasn't got that chance with Brighton or Fulham, just that their managers preferred others or thought they needed a different type. What does surprise me, with his statistical profile, his set piece delivery, his age and recent champonship sucess why only really us and Hull seemed to be gunning for him. Factor in the price too and we have a complete bargain.

 

[The Tea Lady]

However, they weren't always allowed during games, just training. They were allowed in the EFL from 2015/ 16 season as FIFA only approved them in Feb 2015 https://www.bbc.com/sport/football/34267968

Thanks. This is what I was thinking off.

 

[The Tea Lady]

Conclusions from Bolton talk a lot about Stevens and his positioning in the formation. This won't have been by accident. Yes Wilder and Knill will have seen for themselves that the form of Stevens hasn't been great recently, however, what the data does is allows them to see very quickly where the weaknesses in Stevens game have been.

As upthread I'm not yet convinced his form was ever that bad.

I'm trying to think of a cycling analogy...maybe: You're chain is skipping, so it seems like it's old and worn out and needs replacing, but what's happened in fact is the gears weren't indexed - and maybe even the back wheel wasn't sitting properly. You sort those out and in a sense the performance of the chain improves, but really there was nothing wrong with it in the first place.

This is a general point, of course: It's a team game, sometimes individual parts need attention (I think there's far too much emphasis on this, that's really what I'm getting at) but sometimes it's the way they work with each other, and sometimes it's they way they are put together.

It could be that Stevens did have a dip in form (and I know Blades Analytics thinks this) but with his "expected assists" for instance being pretty high then I think there are other ways of looking at his contributions.

 

[The Tea Lady]

Thanks. This is what I was thinking off.

*of

 

[I should coco]

What does surprise me, with his statistical profile, his set piece delivery, his age and recent champonship sucess why only really us and Hull seemed to be gunning for him.

Steve Bruce is still looking for Alex Bruce#2 as the final piece of the jigsaw to turn Villa into title winners.

 

[The Tea Lady]

I think in Norwood's case it was a simple case of Fulham signing numerous midfielders in the summer and he just got further and further back in their plans, same with his parent club Brighton. Fulham had the option of signing Seri from Nice who 12 months ago was minutes away from signing for Barcelona. Now i love ON and his style but let's be clear here he is a top level championship player, perhaps bottom end PL player. Fulham's recruitment has been based on building a top 12 side in the PL already! Although if you ask the more knowledgeable Fulham fans you find that really liked and appreciated what Norwood bought. I think same story can be applied to Brighton, Hughton has his favourites and some of their new signings last year in that area of the team really hit the ground running. I don't think their is a specific technical reason why Norwood hasn't got that chance with Brighton or Fulham, just that their managers preferred others or thought they needed a different type. What does surprise me, with his statistical profile, his set piece delivery, his age and recent champonship sucess why only really us and Hull seemed to be gunning for him. Factor in the price too and we have a complete bargain.

One factor ON mentioned in his interview was having a young family, and his missus in particular wanting to move back Up North and settle. That may also be a limiting factor.

Regardless, whoever negotiated this, all parties should take enormous credit.

(Quality of life was also something McGoldrick mentioned.)

 

[SwissBlade]

As upthread I'm not yet convinced his form was ever that bad.

I'm trying to think of a cycling analogy...maybe: You're chain is skipping, so it seems like it's old and worn out and needs replacing, but what's happened in fact is the gears weren't indexed - and maybe even the back wheel wasn't sitting properly. You sort those out and in a sense the performance of the chain improves, but really there was nothing wrong with it in the first place.

This is a general point, of course: It's a team game, sometimes individual parts need attention (I think there's far too much emphasis on this, that's really what I'm getting at) but sometimes it's the way they work with each other, and sometimes it's they way they are put together.

It could be that Stevens did have a dip in form (and I know Blades Analytics thinks this) but with his "expected assists" for instance being pretty high then I think there are other ways of looking at his contributions.

I like your analogy, so we can say that Stevens wasn't indexed properly.

I think the problem is that Stevens was probably just not contributing to the attacking play well enough before, again i will happily be proven wrong, but I am guessing that when he was in those positions he wasn't getting much joy.

So perhaps the answer was to subtly move him into a position where he could support the midfield better inside but by doing this it created more space for him to attack, instead of being there for most of the game.

 

[Houston_Blade]

I've just had chance to read both articles on the match and xG, good stuff. For the Leon bashers:

Leon Clarke 17/18

• xG – 14.63
  
  
• Goals Scored - 19
  

Very brief one here: Leon exceeded his Expected Goals. That suggests he’s a player with high finishing skill or quite simply he had a purple patch where everything he hit ended up in the net.

 

[Sothall_Blade]

CW alluded to it very well in an interview last week where he was discussing a small tactical "thing" we do in games, and that both him and Knilly had seen opposition repeat this in recent games, so it had been picked up on

The actual quote in The Star from CW was:-

"There’s a little routine we do on the pitch, and we didn’t see it before. We seemed to be the only team doing it."But we came back in last week and I said to Alan that one team had started doing it, and he said another team did it the night before!"I'm not going to reveal what it was, but that's what happens. It's a really small thing, but people have picked up on it.

So what "little routine" do you think he's referring to?
I would guess at the "throw and immediate angled-pass to set the wing-back free" as used to great effect on many occasions.
I've not seen too many other teams do it although it seems relatively easy to copy.
Also would have thought it's fairly easy to stop it working by simply putting a man on the thrower but teams very rarely do.

See 02:07 on this video for an example:-

 

[Revolution]

I've just had chance to read both articles on the match and xG, good stuff. For the Leon bashers:

Leon Clarke 17/18

• xG – 14.63
  
  
• Goals Scored - 19

Very brief one here: Leon exceeded his Expected Goals. That suggests he’s a player with high finishing skill or quite simply he had a purple patch where everything he hit ended up in the net.

I said at the start of the season that if I had to Bet on one player to regress in goals scored it would be Leon based on this stat.

I also would bet O'Connell would do better given he didn't score and his xg was 4.

Similarly, our goals conceded was miles higher than expected goals for much of last season, which can suggest bad luck and regression to the mean.

 

[Damski]

This is great, thanks for posting here and bringing your good work to my attention. Have given you both a follow on Twitter, looking forward to your analysis of the rest of our season.