Author Topic: Statistics and Analytics - insight into our performance  (Read 191509 times)

Offline ElstonGunn

  • is Billy Beane
  • Legacy Fan
  • ******
  • Posts: 3,271
  • We all Live in a Red and White Kop
Re: The Lucas-Gerrard Partnership and Using Stats to Talk about Liverpool
« Reply #80 on: January 10, 2014, 08:30:29 pm »
Fascinating analysis. But just one quibble; there's a couple of typos in your lists (hint ... too many " SoT against"s).
Yep! Sorry, will go back and try to fix tonight.

Offline ElstonGunn

  • is Billy Beane
  • Legacy Fan
  • ******
  • Posts: 3,271
  • We all Live in a Red and White Kop
Re: The Lucas-Gerrard Partnership and Using Stats to Talk about Liverpool
« Reply #81 on: January 10, 2014, 08:32:46 pm »
To be honest, apart from the last couple of posts, I dont see what this thread has to do with Gerrard/Lucas specifically. Isn't it just an LFC stats thread?
I'm obviously only just one opinion, but I'm happy for us to discuss the issues the stats point at as well. It's worth asking why the stats are what they are. Though I definitely don't want this to just turn into another thread where people exchange their general views about Gerrard, so I think it's best if we try to keep the discussion as grounded in the stats as possible. 

Offline ElstonGunn

  • is Billy Beane
  • Legacy Fan
  • ******
  • Posts: 3,271
  • We all Live in a Red and White Kop
Re: The Lucas-Gerrard Partnership and Using Stats to Talk about Liverpool
« Reply #82 on: January 10, 2014, 08:37:10 pm »
Why would this matter assuming 'shots' are long range, or from outside the area?  How many has Mignolet conceded from shots outside the area anyway?  There was one recently v City which was goalkeeping error.  And one v Arsenal.  Don't recall any more off the top of my head.  He's a good shot stopper generally.  Less confident at coming off the line and dealing with high balls.  I'd be more concerned with defending set pieces as opposed to looking at stats re shots at goal.
Really just look at any decent study of the issue, and you'll find that shots stats have much more predictive value than any other basic counting stat you'd look at. It's definitely worth looking at shot quality too (or else you end up thinking any club Villas Boas manages is elite), but it's actually a bit surprising how good pure shots data is in most cases at telling you how good a team is.

Offline ElstonGunn

  • is Billy Beane
  • Legacy Fan
  • ******
  • Posts: 3,271
  • We all Live in a Red and White Kop
Re: The Lucas-Gerrard Partnership and Using Stats to Talk about Liverpool
« Reply #83 on: January 10, 2014, 08:40:23 pm »
I also read somewhere recently that shots from any position are twice as likely to succeed if the preceeding pass was a through ball rather than a cross. Found that very interesting as I haven´t seen any statistical analysis account for that. The same article showed that Arsenal & Liverpool lead the league in through balls, followed by Man City.
You're talking about this, I think: http://cartilagefreecaptain.sbnation.com/2013/11/16/5111212/shot-matrix-iii-the-incredible-through-ball

The question with all this shot quality analysis is how repeatable it is. We know TSR is pretty damn repeatable (that is, it's a fairly stable quality of a team), whereas I don't think anyone has done a serious look on the persistence of chance quality.

Offline ElstonGunn

  • is Billy Beane
  • Legacy Fan
  • ******
  • Posts: 3,271
  • We all Live in a Red and White Kop
Re: The Lucas-Gerrard Partnership and Using Stats to Talk about Liverpool
« Reply #84 on: January 10, 2014, 08:48:23 pm »
I posted this over the weekend if anyone's interested, I attempt, and I think succeed in improving TSR by adjusting for shot quality.

http://woolyjumpersforgoalposts.blogspot.co.uk/2014/01/creating-some-new-metrics-using-optas.html?m=1
And just as I say the above, there you go. This is a really good idea and, honestly, I don't think you could've expected a much better r^2 than .42 year-on-year.
You are doing some very good work here. You are good at explaining your thought process and making it simple for the reader to understand in a statistical world where that is a rarity.

Just wish I could put more stock into the OPTA CCCs as they can be a bit weird and very prone to bizarre human interpretation by their loggers!
Probably true, but they do at least have a pretty repeatable conversion rate league wide IIRC.

Offline barbudo

  • No new LFC topics
  • Kopite
  • *****
  • Posts: 772
  • We all Live in a Red and White Kop
Re: Using Stats to Talk about Liverpool
« Reply #85 on: January 10, 2014, 09:20:05 pm »
Citing R^2 values or ANOVAs with and without Gerrard is dandy for anyone with a stats course or two under their belts, but if you want to convert the less numerate RAWKites, a table comparing TSR with league position at the end of last season might do the trick.

ElstonGunn's done more than enough work on this; is there anybody else with access to these data willing to enlighten us? And, if that's not enough of a challenge, how about a weekly TSR graph for this season for the Reds?
Bethink ye, bethink ye ... that ye might be mistaken! (Oliver Cromwell)

Offline ElstonGunn

  • is Billy Beane
  • Legacy Fan
  • ******
  • Posts: 3,271
  • We all Live in a Red and White Kop
Re: Using Stats to Talk about Liverpool
« Reply #86 on: January 10, 2014, 09:32:13 pm »
Citing R^2 values or ANOVAs with and without Gerrard is dandy for anyone with a stats course or two under their belts, but if you want to convert the less numerate RAWKites, a table comparing TSR with league position at the end of last season might do the trick.

ElstonGunn's done more than enough work on this; is there anybody else with access to these data willing to enlighten us? And, if that's not enough of a challenge, how about a weekly TSR graph for this season for the Reds?
There are a couple in Woolster's post that are relevant to this (hopefully that's all right with him):

TSR's very close correlation with Goal Difference:



TSR's strong consistency year-on-year:



Offline barbudo

  • No new LFC topics
  • Kopite
  • *****
  • Posts: 772
  • We all Live in a Red and White Kop
Re: Using Stats to Talk about Liverpool
« Reply #87 on: January 10, 2014, 09:38:50 pm »
That's great -- thanks for the quick response. As always, it's not the points that lie on the regression line that catch the eye, but the outliers. "Who?" and "Why?" are the obvious questions to ask.
Bethink ye, bethink ye ... that ye might be mistaken! (Oliver Cromwell)

Offline ElstonGunn

  • is Billy Beane
  • Legacy Fan
  • ******
  • Posts: 3,271
  • We all Live in a Red and White Kop
Re: Using Stats to Talk about Liverpool
« Reply #88 on: January 10, 2014, 10:06:36 pm »
Another model that thinks LFC are big favorites for the CL, from @11tegen11:





I'm curious to see his model in more detail (he said he's writing a post later this week). It doesn't seem very TSR-based (as it loves Arsenal), but differs significantly from MCofA's shot-quality-heavy model, in loving Chelsea and not loving us. (Exciting that a model "not loving us" still puts us at 79% top 4 though!)
« Last Edit: January 10, 2014, 10:09:53 pm by ElstonGunn »

Offline GregCharrua

  • Legacy Fan
  • ******
  • Posts: 2,038
  • We all Live in a Red and White Kop
Re: Using Stats to Talk about Liverpool
« Reply #89 on: January 10, 2014, 10:10:50 pm »
I love this thread :D

Thanks for the posts!

Offline robgomm

  • He just can't get enough of Luis Suarez.
  • Legacy Fan
  • ******
  • Posts: 10,087
Re: Using Stats to Talk about Liverpool
« Reply #90 on: January 10, 2014, 10:13:48 pm »
We're surely going beyond 70 points this season. It'd take quite a drop off in form not to reach that and that'd put us in with a good shot. But of course the models - and I know this is jarringly obvious - can go only on what's happened and can't account for variables that might, say, increase the TSR of one of the teams below us (or our own!). It may be possible to do a hypothetical model that does: say if we sign Salah than estimate how many games he'd play and the impact he'd have on our TSR.

Offline GregCharrua

  • Legacy Fan
  • ******
  • Posts: 2,038
  • We all Live in a Red and White Kop
Re: Using Stats to Talk about Liverpool
« Reply #91 on: January 10, 2014, 10:16:19 pm »
We're surely going beyond 70 points this season. It'd take quite a drop off in form not to reach that and that'd put us in with a good shot. But of course the models - and I know this is jarringly obvious - can go only on what's happened and can't account for variables that might, say, increase the TSR of one of the teams below us (or our own!). It may be possible to do a hypothetical model that does: say if we sign Salah than estimate how many games he'd play and the impact he'd have on our TSR.

That sort of estimation you might as well just go with your gut feeling and make a number up. Salah being signed to begin with is totally unknown (and not terribly likely) factor, and even if he was, whether he'd slot right in or struggle for goals or even gametime is totally unknown, and more likely to be negative than positive. Really the only way to do it is to work with concretes and project forward. And even that is, of course, just a model. :P

Offline robgomm

  • He just can't get enough of Luis Suarez.
  • Legacy Fan
  • ******
  • Posts: 10,087
Re: Using Stats to Talk about Liverpool
« Reply #92 on: January 10, 2014, 10:23:15 pm »
That sort of estimation you might as well just go with your gut feeling and make a number up. Salah being signed to begin with is totally unknown (and not terribly likely) factor, and even if he was, whether he'd slot right in or struggle for goals or even gametime is totally unknown, and more likely to be negative than positive. Really the only way to do it is to work with concretes and project forward. And even that is, of course, just a model. :P

Sorry, yeah, I meant creating another model that takes those things into account. It would be necessarily more tenuous I think aye :D

But say Salah is signed (I'm not saying the model should predict who we sign, was speaking hypothetically about him being signed), then there was something Elston talked about the other day where you can fairly accurate predict the number of goals a player going from the Championship to the Premiership will score. We may be able to do that with the Swiss league too.

There may be numbers on the average amount of minutes January signings get :D And then you turn it into a projection. It's using concretes but widening the net (risking less accurate projections).

Offline GregCharrua

  • Legacy Fan
  • ******
  • Posts: 2,038
  • We all Live in a Red and White Kop
Re: Using Stats to Talk about Liverpool
« Reply #93 on: January 10, 2014, 10:47:09 pm »
Is this the article about Championship form vs Premier League form for strikers you are talking about? (http://www.prozonesports.com/news-article-analysis-does-form-translate-from-the-championship-to-the-premier-leagueij.html)

I remember reading something somewhere else about this same topic but can't remember where it was... anyone have a link?


Offline robgomm

  • He just can't get enough of Luis Suarez.
  • Legacy Fan
  • ******
  • Posts: 10,087
Re: Using Stats to Talk about Liverpool
« Reply #94 on: January 10, 2014, 11:09:27 pm »
Is this the article about Championship form vs Premier League form for strikers you are talking about? (http://www.prozonesports.com/news-article-analysis-does-form-translate-from-the-championship-to-the-premier-leagueij.html)

I remember reading something somewhere else about this same topic but can't remember where it was... anyone have a link?

http://www.redandwhitekop.com/forum/index.php?topic=304864.msg12318934#msg12318934

Offline The Woolster

  • Main Stander
  • ***
  • Posts: 198
  • We all Live in a Red and White Kop
Re: The Lucas-Gerrard Partnership and Using Stats to Talk about Liverpool
« Reply #95 on: January 10, 2014, 11:13:14 pm »
You are doing some very good work here. You are good at explaining your thought process and making it simple for the reader to understand in a statistical world where that is a rarity.

Just wish I could put more stock into the OPTA CCCs as they can be a bit weird and very prone to bizarre human interpretation by their loggers!

Thanks, very kind of you to say so!

I may be biased, as I use the stat a lot, but I don't think there is an issue with it. Its very consistent year on year, the conversion is about 37% each year plus or minus a couple, the proportion of shots that are CCCs is consistently about 13%, and they get double checked as I've seen them change after I've tracked them, so its not down to a single person's view.

Offline The Woolster

  • Main Stander
  • ***
  • Posts: 198
  • We all Live in a Red and White Kop
Re: Using Stats to Talk about Liverpool
« Reply #96 on: January 10, 2014, 11:25:06 pm »
Citing R^2 values or ANOVAs with and without Gerrard is dandy for anyone with a stats course or two under their belts, but if you want to convert the less numerate RAWKites, a table comparing TSR with league position at the end of last season might do the trick.

ElstonGunn's done more than enough work on this; is there anybody else with access to these data willing to enlighten us? And, if that's not enough of a challenge, how about a weekly TSR graph for this season for the Reds?

This is probably the best graph for it from James Grayson, from here http://jameswgrayson.files.wordpress.com/2012/07/ptstsr2.png

Black dots were Champs, blue dots 2nd to 4th, red dots relegated


Offline The Woolster

  • Main Stander
  • ***
  • Posts: 198
  • We all Live in a Red and White Kop
Re: Using Stats to Talk about Liverpool
« Reply #97 on: January 10, 2014, 11:35:00 pm »
That's great -- thanks for the quick response. As always, it's not the points that lie on the regression line that catch the eye, but the outliers. "Who?" and "Why?" are the obvious questions to ask.

The big outliers above the line, at the higher end at least, are Utd or City, because of better than average finishing (or chance quality), the outliers below the line are us and Spurs and for the opposite reasons.

Offline The Woolster

  • Main Stander
  • ***
  • Posts: 198
  • We all Live in a Red and White Kop
Re: Using Stats to Talk about Liverpool
« Reply #98 on: January 10, 2014, 11:43:03 pm »
Sorry, yeah, I meant creating another model that takes those things into account. It would be necessarily more tenuous I think aye :D

But say Salah is signed (I'm not saying the model should predict who we sign, was speaking hypothetically about him being signed), then there was something Elston talked about the other day where you can fairly accurate predict the number of goals a player going from the Championship to the Premiership will score. We may be able to do that with the Swiss league too.

There may be numbers on the average amount of minutes January signings get :D And then you turn it into a projection. It's using concretes but widening the net (risking less accurate projections).

You could try an agent based model, like this... http://eplindex.com/43543/results-prediction-model-premier-league-forecasts-arsenal-liverpool-chelsea.html

Offline houkura

  • Legacy Fan
  • ******
  • Posts: 1,058
  • We all Live in a Red and White Kop
Re: Using Stats to Talk about Liverpool
« Reply #99 on: January 10, 2014, 11:50:13 pm »
We're surely going beyond 70 points this season. It'd take quite a drop off in form not to reach that and that'd put us in with a good shot. But of course the models - and I know this is jarringly obvious - can go only on what's happened and can't account for variables that might, say, increase the TSR of one of the teams below us (or our own!). It may be possible to do a hypothetical model that does: say if we sign Salah than estimate how many games he'd play and the impact he'd have on our TSR.

As far as signings go-it's an interesting point. Several of the league projection guys usually note at the end of their explanations something along the lines of "...now if LIverpool sign a top midfielder things could change..." or "..if Liverpool sign a top right back things could change..." It certainly changes things for better or worse it's hard to tell. All signings are a gamble. Another thing to remember is these numbers change every week and while we're right now around 65-80% likely to qualify for the Champions League in most projections-if we lose a game or two those numbers change drastically. These numbers are for right now only. Every game is so very important. The best thing to do is just keep winning games!

PS-I like the thread name change-was needed.
"Good to be here, good to be anywhere."  --Keith Richards

Offline TSC

  • RAWK Supporter
  • Legacy Fan
  • ******
  • Posts: 25,342
  • We all Live in a Red and White Kop
Re: Using Stats to Talk about Liverpool
« Reply #100 on: January 10, 2014, 11:54:49 pm »
Really just look at any decent study of the issue, and you'll find that shots stats have much more predictive value than any other basic counting stat you'd look at. It's definitely worth looking at shot quality too (or else you end up thinking any club Villas Boas manages is elite), but it's actually a bit surprising how good pure shots data is in most cases at telling you how good a team is.

My post which you quoted was responding to a post which was focusing on the amount of 'shots' conceded when Lucas and Gerrard were in midfield.  So my response was what does it matter if they're all long range & don't result in goals.

Offline The Woolster

  • Main Stander
  • ***
  • Posts: 198
  • We all Live in a Red and White Kop
Re: Using Stats to Talk about Liverpool
« Reply #101 on: January 11, 2014, 12:19:00 am »
If you don't know about Nate, who does the Oh You Beauty blog, then you should. His match visualisations after every league game are the best out there

http://ohyoubeauty.blogspot.co.uk/2014/01/visualized-liverpool-2-0-hull.html

Offline ElstonGunn

  • is Billy Beane
  • Legacy Fan
  • ******
  • Posts: 3,271
  • We all Live in a Red and White Kop
Re: Using Stats to Talk about Liverpool
« Reply #102 on: January 11, 2014, 12:46:38 am »
My post which you quoted was responding to a post which was focusing on the amount of 'shots' conceded when Lucas and Gerrard were in midfield.  So my response was what does it matter if they're all long range & don't result in goals.
Er, I know? My response was completely relevant. It matters because there's proof that it does. Not sure why "shots" is in quotes either.

Offline ElstonGunn

  • is Billy Beane
  • Legacy Fan
  • ******
  • Posts: 3,271
  • We all Live in a Red and White Kop
Re: Using Stats to Talk about Liverpool
« Reply #103 on: January 11, 2014, 12:47:38 am »
If you don't know about Nate, who does the Oh You Beauty blog, then you should. His match visualisations after every league game are the best out there

http://ohyoubeauty.blogspot.co.uk/2014/01/visualized-liverpool-2-0-hull.html
My favorite. I've been trying to get Rawk reading him for a while.  ;D

Offline TSC

  • RAWK Supporter
  • Legacy Fan
  • ******
  • Posts: 25,342
  • We all Live in a Red and White Kop
Re: Using Stats to Talk about Liverpool
« Reply #104 on: January 11, 2014, 12:53:26 am »
Er, I know? My response was completely relevant. It matters because there's proof that it does. Not sure why "shots" is in quotes either.

Wel ok then how many goals outside the box have we conceded when Gerrard and Lucas have played in midfield?  And how many have we conceded outside the box without both in midfield?

Offline ElstonGunn

  • is Billy Beane
  • Legacy Fan
  • ******
  • Posts: 3,271
  • We all Live in a Red and White Kop
Re: Using Stats to Talk about Liverpool
« Reply #105 on: January 11, 2014, 01:13:54 am »
Wel ok then how many goals outside the box have we conceded when Gerrard and Lucas have played in midfield?  And how many have we conceded outside the box without both in midfield?
I don't know. Feel free to do the data collection and let us know. I'm also not sure what you think the answer would prove. I expect we have forced a higher proportion of shots from outside the box when L&G played together (a good thing), but not nearly enough to make up for the basic TSR difference.

Online JackWard33

  • RAWK Supporter
  • Legacy Fan
  • ******
  • Posts: 25,725
  • President of the Harry Wilson fanclub
Re: Using Stats to Talk about Liverpool
« Reply #106 on: January 11, 2014, 01:16:57 am »
Wel ok then how many goals outside the box have we conceded when Gerrard and Lucas have played in midfield?  And how many have we conceded outside the box without both in midfield?

Sample size and variance are going to mean those numbers are pretty meaningless

Offline TSC

  • RAWK Supporter
  • Legacy Fan
  • ******
  • Posts: 25,342
  • We all Live in a Red and White Kop
Re: Using Stats to Talk about Liverpool
« Reply #107 on: January 11, 2014, 01:22:39 am »
I don't know. Feel free to do the data collection and let us know. I'm also not sure what you think the answer would prove. I expect we have forced a higher proportion of shots from outside the box when L&G played together (a good thing), but not nearly enough to make up for the basic TSR difference.

Well that's no good.  My original post was in response to someone claiming Gerrard and Lucas combo gave so many 'shots' away.  (It was in brackets because even crap shots still get classed as shots).  In response I asked a reasonable question.  How many of these shots from outside the box which Lucas and Gerrard have gave away resulted in goals.  And how many have resulted in goals when they've not been in the centre.

it would appear it's either rather difficult to get these stats, or else they're so insignificant they render the argument irrelevant.

I assume the latter in the absence of 'proof'.  And you know why I've put that in 'quotes' eh?

Offline TSC

  • RAWK Supporter
  • Legacy Fan
  • ******
  • Posts: 25,342
  • We all Live in a Red and White Kop
Re: Using Stats to Talk about Liverpool
« Reply #108 on: January 11, 2014, 01:25:55 am »
Sample size and variance are going to mean those numbers are pretty meaningless

Exactly.  Yet someone earlier  came on and claimed Gerrard should not be played in the middle because with him and Lucas we concede so many 'shots'.  My response was 'so what'.  As long as it doesn't lead to goals.  And anyway, there are no stats to support this view.

Online JackWard33

  • RAWK Supporter
  • Legacy Fan
  • ******
  • Posts: 25,725
  • President of the Harry Wilson fanclub
Re: Using Stats to Talk about Liverpool
« Reply #109 on: January 11, 2014, 01:27:31 am »

I assume the latter in the absence of 'proof'.  And you know why I've put that in 'quotes' eh?

Goals from shots outside the box are scored at a low rate so that you'd have to play a huge amount of games for the numbers to be truly meaningful

The point about giving away a lot of shots - as we do unfortunately - (and more so with the poorer CM combination) is that you're effectively playing roulette; the more shots you're allowing, even from outside the box the greater chances of a goal being scored (a goal keeper error, deflection, 'worldy' or whatever)

Offline ElstonGunn

  • is Billy Beane
  • Legacy Fan
  • ******
  • Posts: 3,271
  • We all Live in a Red and White Kop
Re: Using Stats to Talk about Liverpool
« Reply #110 on: January 11, 2014, 01:38:00 am »
Goals from shots outside the box are scored at a low rate so that you'd have to play a huge amount of games for the numbers to be truly meaningful

The point about giving away a lot of shots - as we do unfortunately - (and more so with the poorer CM combination) is that you're effectively playing roulette; the more shots you're allowing, even from outside the box the greater chances of a goal being scored (a goal keeper error, deflection, 'worldy' or whatever)
Yeah, this pretty much. Shot numbers are more meaningful--in any sized sample, really, but pariticularly in one this small--than the pure outcome-focused stat of goals, which is subject to a lot of randomness. One of the big insights of stats analysis is that focusing on process-based information (like how many shots you take and allow) tells you more about underlying quality than outcome-based ingormation (like goals).

Offline TheBestSideDrew

  • Main Stander
  • ***
  • Posts: 116
  • We all Live in a Red and White Kop
Re: The Lucas-Gerrard Partnership and Using Stats to Talk about Liverpool
« Reply #111 on: January 11, 2014, 09:34:37 am »
Very cool! Hadn't seen this. The big difference I see here is that he incorporates possession into the calculation. I'm always very skeptical of how important possession is, so I wonder if there's any evidence that incorporating possession into the calculation improves accuracy over and above just the shot data.

Hi Elston, I was just having a quick look at the link to an LFC performance dashboard (http://public.tableausoftware.com/views/LFC_2013_14_Perf_Dashboard/LFC#1)  that someone posted earlier in the thread; worth a look over a cuppa - quite interesting. This does seem to show a correlation between possession and win/loss which would suggest it could be a useful indicator..?

Online Eeyore

  • "I have no problem whatsoever stating that FSG have done a good job.".Mo Money, Mo Problems to invent. Number 1 is Carragher. Number 2 is Carragher. Number 3 is Carragher. Number 4 is Carragher. Likes to play God in his spare time.
  • Campaigns
  • Legacy Fan
  • ******
  • Posts: 31,558
  • JFT 97
Re: Using Stats to Talk about Liverpool
« Reply #112 on: January 11, 2014, 11:11:02 am »
Yeah, this pretty much. Shot numbers are more meaningful--in any sized sample, really, but pariticularly in one this small--than the pure outcome-focused stat of goals, which is subject to a lot of randomness. One of the big insights of stats analysis is that focusing on process-based information (like how many shots you take and allow) tells you more about underlying quality than outcome-based ingormation (like goals).

The big problem for me with looking at the numbers in isolation is that they don't really tell the whole story. Away to Villa is a pretty good example, last season they tore us apart on the counter and at the start of this season they did the same to Arsenal. They have forwards who are a real threat in behind but are extremely poor at breaking down a low block.

In that game we made a conscious decision to defend deep, allow no space behind the back four and allow Villa to have the ball in areas where they could take pot shots but couldn't hurt us. As TSC was saying about Gerrard and Lucas sometimes you actually want the opposition to try their luck from distance because you are negating their strength whilst at the same time allowing yourself space to counter attack.

At the start of the season we had no Suarez and had to rely almost entirely on Sturridge for goals, that necessitated us to be tight at the back and to keep clean sheets if we wanted to win games. At the moment we are playing in a completely different way Suarez is scoring a ridiculous number of goals and we are basically playing we will score more than you. We are prepared to give teams chances to attack in numbers safe in the knowledge that we have the hottest striker on the planet.

The better sides are prepared to take up that offer of playing in effect basketball where you have an attack and we have an attack whereas the vast majority of sides simply don't have the quality to do that especially at Anfield. So you end up with a pretty small initial sample size that is then made up of different ways we have approached games and different ways the opposition have approached games.

In short you have to look a bit deeper than shots at goal or any single metric and look at what each team set out to achieve because conceding 5 shots on target in a game you have played end to end is very different to conceding 5 shots on target in a game you have looked to keep tight and nick 1-0.
"Ohhh-kayyy"

Offline steveeastend

  • Learnt to play them drums
  • RAWK Supporter
  • Legacy Fan
  • ******
  • Posts: 15,853
Re: Using Stats to Talk about Liverpool
« Reply #113 on: January 11, 2014, 02:18:30 pm »
I dont need to compare anything.

His finishing ability, whether its from open or set play hasnt diminished (long range missiles aside) and he still offers more of a goal threat than any other player outside the strikers.



I don't know why but you clearly closing your eyes on what's going on if you say he hasn't lost anything in his finishing abilities.

Coutinho, Sterling and Henderson offer more goal threat from midfield compared to Gerrard because his age denies him to be the player he once was and that won't change anymore.
« Last Edit: January 11, 2014, 02:20:11 pm by steveeastend »
One thing does need to be said: in the post-Benitez era, there was media-led clamour (but also some politicking going on at the club) to make the club more English; the idea being that the club had lost the very essence of what it means to be ‘Liverpool’. Guillem Ballague 18/11/10

Offline BabuYagu

  • It's Portuguese for 'BabyYoghurt'. The John Motson of RAWK. Or Barry Davies. Or Charley Boorman, even. Expertly silent fist-pumper. Needs to pay more attention. Repeatly analing goalkeepers.
  • Legacy Fan
  • ******
  • Posts: 17,350
  • wakelet.com/@BabuYagu
    • Wakelet of the Articles I have written
Re: Using Stats to Talk about Liverpool
« Reply #114 on: January 11, 2014, 03:41:08 pm »
I don't know why but you clearly closing your eyes on what's going on if you say he hasn't lost anything in his finishing abilities.

Coutinho, Sterling and Henderson offer more goal threat from midfield compared to Gerrard because his age denies him to be the player he once was and that won't change anymore.

I think what he means is the actual ability is still there... if you give all those players the same chances you would back Gerrard.

The problem is his legs don´t allow him to open up from long distance anymore and his age means he cannot burst into the box regularly anymore and needs to be more disciplined. So he has very few chances to use his finishing. Whereas Coutinhos dribbling, Sterlings pace and Henderson´s running mean they will all get several good chances per game to get shots off.

I would agree they all carry a far greater goal threat than Gerrard in CM who pretty much is only a threat at set pieces.
My first article on Anfield Index on Shaqiri. Enjoy. bit.ly/2mAq3Qd

Offline Deo

  • No new LFC topics
  • Kopite
  • *****
  • Posts: 749
Re: Using Stats to Talk about Liverpool
« Reply #115 on: January 12, 2014, 07:46:13 pm »
So, I just checked:

Until now, Stoke had an average of 3.1 shots on target per game at the Britannia Stadium. Today, they had 12.

This is really worrying, but a win is a win I guess! :D
"Our fans cannot give red cards or award penalties, but they can score goals." - Rafa Benitez

Offline rscanderlech

  • Legacy Fan
  • ******
  • Posts: 4,023
  • We all Live in a Red and White Kop
Re: Using Stats to Talk about Liverpool
« Reply #116 on: January 12, 2014, 09:01:52 pm »
So, I just checked:

Until now, Stoke had an average of 3.1 shots on target per game at the Britannia Stadium. Today, they had 12.

This is really worrying, but a win is a win I guess! :D
According to WhoScored.com, they had 4 on target, 13 overall.

Offline Trendisnotdestiny

  • Finally, the custom title that cannot be beat
  • RAWK Supporter
  • Legacy Fan
  • ******
  • Posts: 15,630
  • Go for Goal Sunshine! - N Saunders
Re: Using Stats to Talk about Liverpool
« Reply #117 on: January 12, 2014, 09:23:02 pm »
Any word on the Gerrard-Lucas partnership we saw today?  Fascinating thread and I look forward to more analysis.
THIS IS ANFIELD SIGN:
It’s there to remind our lads who they’re playing for and to remind the opposition who they’re playing against! - Bill Shankly

We have everything we need - Jurgen Klopp

You need to get more wives mate, it fixes everything. Apart from then you have loads of wives, which is a nightmare.  -  Djozer

Offline Deo

  • No new LFC topics
  • Kopite
  • *****
  • Posts: 749
Re: Using Stats to Talk about Liverpool
« Reply #118 on: January 12, 2014, 10:08:10 pm »
According to WhoScored.com, they had 4 on target, 13 overall.

I guess the app I've got it from counted all the very long range shots, blocked shots and the Gerrard header too.
"Our fans cannot give red cards or award penalties, but they can score goals." - Rafa Benitez

Offline ElstonGunn

  • is Billy Beane
  • Legacy Fan
  • ******
  • Posts: 3,271
  • We all Live in a Red and White Kop
Re: Using Stats to Talk about Liverpool
« Reply #119 on: January 15, 2014, 05:55:59 pm »
New MCofA projections: http://cartilagefreecaptain.sbnation.com/2014/1/15/5309920/premier-league-projections-week-21-zero-sum

No real change, we're still at 83% for top 4. This from him in the comment section was interesting, however:

Quote
On Liverpool

The defense has been up and down, and they’re not amazing. But the attack is more than “impressive” — they’ve been the best in the league, at a level with Manchester City and much better than Arsenal or Chelsea. That’s the key for the model, that Liverpool have been just ridiculous at creating high-quality chances.

And the defense has had its share of top performances. They allowed 0 danger zone shots and 0 SiBoT in the win over Hull City, they shut us down, they didn’t get overrun by Manchester City, and there are a bunch of others. Their defensive numbers are below Arsenal’s, City’s, and Chelsea’s, but not by as much as their attacking numbers are better.

I’m not really nervous about Liverpool. If they collapse, the primary beneficiaries are probably either Spurs (yay!) or Everton (cool). And if there’s something new to learn and to incorporate into the model, that’s great. (And, of course, I could be right and Liverpool could miss top 4. Things with a 15% chance of happening, they happen all the time.)