Topic: Statistical Comparison of LFC players to Rest of EPL - starting w/LB

[BabuYagu]

Results are in. Here are the best left backs in the Premier League for the 2016/17 season (I have used 2015/16 data for Moreno as his numbers were too small otherwise and we needed to include him for comparison purposes. All other players are 2016/17 though).



Green
This grouping would be what I would consider the best grouping of left backs in 1-v-1 situations in the Premier League at the moment. I would say for that grouping Davies & van Aanholt look the best, although van Aanholt will be 27 when the season starts. We seem to target players a few years younger than that so we have a core squad coming into their peak 2 or 3 years from now. With this in mind, and Davies likely unavailable (or over priced for what we want to spend) as he is at Spurs, Robertson then looks the best option available to us.

Light Green
Bertrand, Cresswell & Olsson may also be very good options although the very low number of tackles they are making is a concern. Are they simply not being tested much in 1-v-1 situations? Are they world class at staying on their feet and jockeying people away from danger without making a challenge? Or are they too passive and simply back off ensuring they are rarely beaten, foul or make a challenge but cause a lot of problems for their sides?

Yellow
This would be the league average group. Alonso was the surprising name in there for me. Although maybe this explains why Chelsea want a 60m left back this summer?

Orange
This would be the group that seem weak in 1-v-1 situations. They are all fouling & being dribbled past more than twice as often as Nathaniel Clyne. Interesting names in group are Chilwell (bullet dodged or still learning?), Milner (sorry James, but I figured this would be his weakest category given he is not a defender and at 31 will be losing some of that explosiveness) Monreal (explains Arsenals need for a new leftback too) and Danny Rose (always thought he was a liability in defence, his improved attacking game and better center backs mopping up behind him maybe is seeing his weaknesses exposed less?)

Outliers

Gibbs & Moreno are both way out on their own, and not in a good way for either. Moreno looks like he could be a good fullback. Hell, maybe even a world class one. However, again, his decision making comes into play. He dives into every challenge. He attempts 4,3 tackles per 90 compared to the second highest in the league 2,99. The league average is 2,1. If he stayed on his feet half the time you expect he would simultaneously reduce his fouling and dribble past numbers too. You 1/2 all his numbers and he would be the best full back in the league in this category by some distance. But that has always been the problem with Alberto, if. He has everything a full back could need to be the Naby Keita of full backs except the brain of one.

Why are we not looking at Kieran Gibbs?

We just looked at him and the view is one of pure shite. He rarely makes a tackle but when he does, he fouls 3 people at once and simultaneously gets dribbled past too, somehow. His tackle success is the worst in the league too.This probably explains why at 27 years old he is leaving his physical peak, has never been first choice and averages about 10 full games per season in minutes played.  The only positive is he hasnīt killed any kid with one of his aimless tackles. Yet.

I also decided to look back comparitively at all Liverpool fullbacks as far back as I can (2012/13) for comparison. Hereīs what that looks like. I include the grouping circles in the same place as a point of reference. Here you go.



18 July 2017
Update:- I have added player names and numbers rather than rankings to the table. & a table comparing Robertson, Milner & Moreno only.

I have noticed a recurring theme in how people look at footballers.

We see every minute of our own players, warts and all. Therefore we know them best. We know exactly how good they are. We also are better placed than anybody to say when they stink. We also are attached to our own players and thus we develop a bias. I love Alberto Moreno. Not the full back so much, but the person. I want him to succeed. Imagine a full back who is solid with his pace and attacking numbers. I canīt help but dream. But itīs just not happening. He makes too many little mistakes (and big ones) which when you add them up I cannot escape from the fact that he costs us goals. He isnīt shite. He doesnīt do everything wrong. But he does enough wrong for him to be a problem for sides that will place a lot of defensive responsibilities on him and whose goalkeeper(s) seem to have issues with shots from wide angles and crosses.

What we end up with then is people who cannot see past the rose tinted lens of those they like and will defend them to the hilt. The will look out for any and all information that supports their pre-conceived opinion of him. Likewise, there are those who have written the player off and will look for any and all evidence to support that opinion. When comparisons to non-Liverpool players come into the question the waters get muddier still. Simply because we donīt watch every minute of the players we are comparing them too. We probably havenīt see the "warts and all" yet unless they are highlighted to us by pundits, blogs we read or we happen to pick up on it in the very small sample size of the total number of games they play.

This happened to me with Cavani early in his career. I saw him be very very shite over the course of about 5 games for Palermo. He went to Napoli for a fee and hype I couldnīt wrap my head around and he proceeded to be equally shite there for 5 or so more games. When I started getting into stats, analytics, metrics, whatever you want to call it, I stepped back and looked at him statistically. What I noticed is that stats are an unbiased picture of everything a player has ever done that is currently meassurable on a football pitch. Those stats included all those shit games of Cavani I watched. They also included to 70 or so non-shit games he had over the same period. You see - my opinion was off simply because I didnīt have enough evidence. Watching a player 10 or so times always felt like enough to get a good impression of a player. I was wrong. The conclusion I now have on Cavani based on stats married with clips (my opinion was formed pre-every touch videos and highlights clips) and watching the player is that Cavani can be brilliant. He can also be shite. On average, he is brilliant a lot more often than shite. He can win you games on his own. There will be games he just doesnīt show up for too. Also he looks like an ugly Dothraki extra from Game of Thrones.

^{This would be a fuck of a lot easier with a real bow Khaleesi!}

There has to be a better way to do this though, right? To remove that bias? Thus I present you with "Babuyaguīs attempt to look at shit in an unbiased way to prevent people just dismissing stuff that doesnīt agree with their own pre-conceived ideas about things and instead have a rational conversation about the players identifying their supposed strengths and weaknesses suggested by stats and then use scouting to verify or dismiss what the stats suggest which is how I imagine it works at clubs more or less but just on a bigger scale with more data and professional scouts, right".

Thatīs just a working title at the moment. Iīm concerned itīs a bit wordy but that is the least important aspect of what I am doing. For now I just want to engage people with the idea that stats have some use yet can be misleading. Videos have a use but can be misleading. Watching players has a use but can be misleading. Marrying that all together seems to be the best approach of all though.

So first was to look at our fullbacks seeing as Robertson may soon be joining the club. Or someone else if it falls through. Or no one. Or we might keep Moreno and give him another chance as he is quite good actually, the stats suggest this. Or heīs shite and aside from running fast in straight lines resembles more an olympic sprinter than a footballer. Either way - itīs a topic of some discussion right now that it would be good to have an unbiased look at if possible.

Therefore I asked myself, in Liverpoolīs system what are the vital strengths a full back should have. I will try to look at them one at a time to give as much focus to each one as possible. The first topic therefore is:-

1. Good in 1-v-1 situations.
Our defenders get isolated a lot more in our system that in others. To try and show this I will look at Tackles, Fouls & Dribbled past. If someone is fouling & being dribbled past more than they are winning tackles that would indicate they are weak in these situations. I will compare every left back in the league who played more than 1,000 minutes. I have also added in Gibbs & Chilwell (as they have been linked to us) despite them falling a little short of 1,000 minutes. I also added Walker & Clyne as they are seen as the two best right backs in the league to give us some standard bearers to compare the rest to. Please note, Blind & Shaw & Trent Alexander-Arnold have not been included due to Blind playing as much as a CB as a LB (Which screws the numbers a lot) and Shaw & TAA barely played in the league last season. I have also added the 2015/16 data for Moreno to include him in the comparisons seeing as he is the player we are trying to upgrade upon.

I will post actual numbers for things later, for now I will just show you where each player ranks in terms of tackle success %, fouls committed per 90 and dribbled past per 90. I will then sort the players by their average rank and add a scouting comment where I perceive there may be a weakness (e.g. a player who has low numbers across the board may be too passive, is a player winning lots of tackles diving in too much meaning he is also losing a lot of tackles also, etc).

Player Name	Tackling Success Rate	Tackles per 90	Fouls per 90	Dribbled Past per 90	1-v-1 Score	Comments for Scouts
Clyne	76.25%	2,16	0,38	0,62	1,00	No perceived Weakness
van Aanholt	74,42%	1,55	0,54	0,61	1,15	Passive?
Cresswell	71,43%	0,89	0,47	0,64	1,11	Very Passive?
Davies	71,79%	2,02	0,67	0,67	1,35	No perceived Weakness
M.Olsson	70,00%	0,72	0,86	0,22	1,08	Very Passive?
C.Daniels	71,19%	1,74	0,35	0,94	1,29	Dribbled Past too easily?
Clichy	72,73%	1,86	0,52	1,02	1,52	Dribbled Past too easily?
Robertson	67,92%	1,74	0,59	0,76	1,34	Tackling?
Fuchs	82,29%	2,78	0,84	1,51	2,35	Dribbled Past? Dives In?
K.Walker	70,83%	2,40	1,20	0,60	1,80	Fouls too much?
Bertrand	70,83%	1,33	0,86	0,47	1,33	Tackling? Very Passive?
Chilwell	78,26%	2,97	0,91	1,68	2,59	Dribbled Past? Dives In?
M.Alonso	67,86%	1,87	0,84	0,87	1,70	Tackling?
Friend	75,81%	2,85	1,29	1,01	2,30	Fouls? Dives In?
Nyom	71,62%	2,53	1,71	0,65	2,35	Fouls?
Moreno	74,80%	4,34	1,02	1,09	2,12	Dribbles Past? Dives in Most?
Pieters	77,14%	2,99	1,71	0,91	2,63	Fouls? Dives in lots?
Baines	69,84%	2,01	1,12	0,83	1,95	Fouls?
Ward	67,14%	1,95	0,97	0,86	1,84	Tackling?
Holebas	67,74%	1,92	1,33	0,68	2,01	Tackling? Fouls?
Monreal	67,78%	2,57	0,94	1,4	2,34	Tackling? Dribble Past? Dive In?
Rose	68,00%	2,94	1,53	1,23	2,76	Fouls? Dribble Past? Dive In?
Milner	67,02%	2,68	1,08	1,4	2,48	Tackling? Dribbles Past? Dives In?
Gibbs	58,33%	1,54	1,41	1,03	2,44	Literally everything. Also appears as passive. How?

Notes:-
Passive? = In the bottom 5 in terms of tackles made. Failure to make errors could be the player just backing off which could be worsening the situation. Or he could be good at jockeying the player away from danger thus not needing to make tackles at all. Scouting required.
Dives In? = Player in the top 5 in terms of tackles made. Is he diving in too much rather than jockeying the player when the situation needs it? Scouting required.
Tackling? Player in the bottom 7 in terms of Tackling success rate. May need scouting to ascertain why and if itīs a legitimate concern or not.
Fouls? Player in the bottom 7 in terms of Fouls per 90. May need scouting to ascertain why and if itīs a legitimate concern or not. He could be simply making a lot of tactical fouls to kill counter attacks before they can blossom into something more dangerous.
Dribbles Past? Player in the bottom 7 in terms of players successfully dribbling past per 90. May need scouting to ascertain why and if itīs a legitimate concern or not. He could be allowing himself to be beaten while also pushing the player towards less dangerous areas of the pitch (e.g. giving up his outside to prevent someone cutting in on a stronger foot).

Iīve left out the names of all the players under 30 who are not right backs and are not Keiran Gibbs. The reason being that over 30īs would be just adding another Milner aged player in the same position which I suspect we would avoid. So you can use these guys as comparisons for those current unknowns the number are throwing up. Also, how fucking good is Clyne in 1-v-1īs based on this? Jaysus! I said elsewhere but wingers should just pass him the ball and save themselves the trouble.

Now we cannot draw anything conclusive from the above as at the moment, they are just numbers. Also it is only ONE aspect of being a fullback we are looking at right now. An important one for Liverpool but still just one none the less. It could all mean nothing if the player is making 5 mistakes a game costing us goals. But we will come to that part later. For now, it should highlight some good players to scout further. It could highlight specific weaknesses to pay attention to.

We probably donīt want to put all our eggs in one basket so out of those 14 unknown players we should probably each pick at least 5 to scout and pay attention to as our later checks could throw up some glaring weaknesses in our current favourites. Thoughts on who you like?

Comparison of Robertson with Liverpoolīs 2 existing full backs.

Player Name	Tackling Success Rate	Tackles per 90	Fouls per 90	Dribbled Past per 90	1-v-1 Score	League Rank per Category & Overall
Robertson	67,92%	1,74	0,59	0,76	1,34	Rank 17th, 6th, 10th, 8th overall
Moreno	74,80%	4,34	1,02	1,09	2,12	Rank 6th, 15th, 19th, 15th overall
Milner	67,02%	2,68	1,08	1,4	2,48	Rank 22nd, 16th, 21st, 21st overall

[BabuYagu]

Oh and for people who like visuals, this will help give you an idea where this thread is ultimately going and what inspired me to start on this.



These numbers are useful but donīt mean much in terms of game situations. Although itīs nice to see that Robertson has no real weaknesses in this graphic I made and isnīt in the bottom 2 in any category. But perhaps we can create our own visual like this but that has more meaning?

[northern Monkey]

Looking forward to more on this.

(Posting to remind myself to check in later).

[Doc Red]

Appreciate the effort, Babu!
Love this idea of comparison, even if it is restricted to the "defensive" aspect of a full back.
Would be interesting to know the rate of attempted tackles per 90. There might be some fullbacks that have a low rate of tackles (like Hyppia probably had) as they read the game better but also had a low success rate of tackling. Similarly, a high rate of tackles and a high rate of success would pretty much confirm a gung ho approach that though has a higher rate of success, might not suit a system like ours with full backs that play high up and mistakes can lead to massively exposing ourselves for a counter.

Are you thinking of doing the same for other positions?  (sorry, greedy I know!)

[BabuYagu]

Appreciate the effort, Babu!
Love this idea of comparison, even if it is restricted to the "defensive" aspect of a full back.
Would be interesting to know the rate of attempted tackles per 90. There might be some fullbacks that have a low rate of tackles (like Hyppia probably had) as they read the game better but also had a low success rate of tackling. Similarly, a high rate of tackles and a high rate of success would pretty much confirm a gung ho approach that though has a higher rate of success, might not suit a system like ours with full backs that play high up and mistakes can lead to massively exposing ourselves for a counter.

Are you thinking of doing the same for other positions?  (sorry, greedy I know!)

Perhaps a spoiler but I think you might have already described the difference between Robertson and Moreno. Moreno makes by far, and a huge margin, the most tackles per 90 on this list. It almost looks like a compulsive disorder at this point. The average is 2.1 tackles per 90. Moreno is 4.3. Second highest has 2.9. Robertsonīs tackles per 90 is about 1.75 which is slightly lower than average but not dangerously low with a few players clocking in with less than 1.

This also passes the eye test too. He doesnīt strike me as someone who has a problem in knowing when you tackle or not. He's a scratch under  6 foot and has very good pace and backs himself in 1-v-1īs when the need arises. Whereas the eye test says Moreno dives into tackles way to much and his position probably means he needs to dive in even more on top as he is usually in the wrong place and needs to make a tackle just to solve that little problem he makes for himself.

Anyway Iīll be posting a similar table tomorrow complete with numbers instead of just rankings and player names. Iīll include tackles made per 90 in that also for you. Iīll also try to do as many categories as we can. Problem with attacking play is I think that will come down more to scouting that data. I just donīt have those metrics for things like packing that would show how effective they are from an attacking sense. Plus itīs really hard to compare full backs attacking numbers in a low block counter attacking side to those in a possession oriented side. Iīll add what I can though that I think is fairly measurable.

[PoetryInMotion]

Great work there mate.

And Cavani is a great great player. Lewandowski level. Not a player who often passes the eye test of a casual viewer though. You really have to follow him to know how good he is. I made an argument for him in a different thread, will post it here if I find it.

If we replaced Suarez with him (even though he's a marginally lesser player), I'm sure we would not have had a drop off in our side like how it was in 2014-15. He hounds and presses like Suarez and his movement is insanely intelligent.

As for Left Backs, I really like the metrics that you're using. I don't understand how a lot of people undervalue Clyne and have even called for replacements. Some have mentioned about Alexander-Arnold taking over this or next season. The way I see is, he is one of those players whose value won't be apparent but when he is not there, you know it and there is a big impact and a drop off. Alexander-Arnold has a good cross in him and is one of our best prospects, but he has a long way to go to get close to Clyne defensively. And our constant weakness in our squad for the last half a decade has been defense - not midfield, not attack. We have to upgrade our defense (LB and CB cover/competition). Not remove the already present better elements of our defense in Clyne. He might be one of the Best Full Backs we've had in a long long time and one of the best in the league, but if we read the complaints and posts on him last season, you would think he was Konchesky.

[BabuYagu]

You can also see when you look at the above how much of a situation Arsenal had with full backs. I never considered Monreal that bad but this indicates he isnīt particularly good in 1-v-1īs. I got the impression he is more someone who will break up attacks with interceptions and tactical fouls, never thought his tackling was particularly bad though. This suggests if you can isolate him 1-v-1 you might have some success.

Before I done any work on this I expected Clyne to be #1 and Gibbs to be #shit. Glad that bore out in the results. Gives me some confidence at least that the numbers are showing what I want them to show. Walker & Fuchs are exactly what I expected them to be also having watched them a lot. Few get past Walker, he will dive in and either win the ball or foul you. His pace bails him out of problems he creates for himself. Fuchs is the same but without the fouls. He dives in a lot allowing players to skip by him too easily. He wins a lot of the ball though so when people were using tackles to find full backs before I saw a lot of people saying "we could do worse than sign Fuchs". Yes... we could sign Gibbs. But we could also do a lot better too.

[BabuYagu]

Oh and speaking of Kolasinac, he doesnīt look particular solid either in this category.  His numbers would rank him very slightly better than Monreal. He dives in wayyyy too much (2nd in terms of total tackles), wins a fair share (6th for tackle success), fouls a lot (joint bottom) and gets dribbled past an above average amount too (15th).

Overall he looks similar to player #15. Interestingly Kolasinac and player #15 both profile as center backs in terms of size and weight with only 1cm and 1kg of difference between them which supports my theory that smaller players are better equipped to deal with 1-v-1 situations at full back isolated against wingers.

[rickardinho1]

Awesome work Babu. I appreciate how much time you've put into that as I've done some similar compilations myself, so thanks for that.

FYI below is a table I put together recently in the Robertson transfer thread that might be useful to include in this thread too.

-----

Here's a statistical comparison between Robertson and some of the best fullbacks in the PL (plus Mendy).

Stats are per-90 (from whoscored)

                Milner - Robertson - Bertrand - Clyne   -   Davies  -  Rose  -  Walker -  Monreal   -  Mendy
Tackles           2.6   -  1.6    -    1.3     -     2.2      -     1.7     -   2.8    -    2.2    -     2.5       -    1.9
Interceptions   1.2   -  1.7   -     1.6     -    1.2      -      0.5    -   1.2    -     1.2    -     2.5       -    1.6
Fouls               1.1   -  0.5   -     0.9     -     0.4     -      0.5    -   1.4    -     1.1    -     0.9       -    1.2
Clearances       2.8  -   2.3   -    1.6     -     1.5     -      3.4    -    2.9    -     2.8    -    2.8       -     2.4
Dribbled past  1.4   -   0.7   -     0.5    -     0.6     -      0.6    -    1.2    -    0.5     -     1.4      -     0.9     
Blocks             0.1   -   0.5    -    0.4    -     0.1     -      0.3   -     0.2    -    0.2     -     0.3      -     0.1
Dispossessed   0.6   -  0.6    -    0.5    -     0.5     -      0.4    -    1.1    -    0.8     -     0.5      -     1.1
Dribbles           0.9  -   0.8    -    0.5    -     0.6     -      0.3    -    1.3    -    1.1      -    0.5      -     1.5
Bad Control     0.9   -   0.5    -   1.4     -     0.7     -      0.4    -    1.5    -    0.8      -    0.4     -      1.0

[10 years on]

Clyne-Good defender, rarely beaten by his man, caught out of position occasionally but what full back isn't. Very frustrating going forward quite often gets to the byline and just hits the first man with the cross.

Milner-Surprisingly very good defensively, quicker then you think he is aswell. Certainly much better then clyne going forward but has a tendency to just cut back onto his right foot and just loft crosses into our small forwards.

Moreno- Very likeable guy, has showing promise but lacks concentration/ a chromosome.

Robertson- Similar to clyne probably better going forward, but he's not as strong or quick and is worst in one v one situations.

Flano- Not sure how he's still a Liverpool player, was decent in fairness under rodgers for that half a season but so was the whole team. He can't run, he gets beaten by league 2 players in one v one's, he seems overweight, technically he has to be one of the worst players to play for the club. Surely will leave this summer, probably to a championship side.

Trent- Definitely looks a promising player, not sure what he's like positionally though, seems a bit of a headless chicken at times.

[Suareznumber7]

Perhaps a spoiler but I think you might have already described the difference between Robertson and Moreno. Moreno makes by far, and a huge margin, the most tackles per 90 on this list. It almost looks like a compulsive disorder at this point. The average is 2.1 tackles per 90. Moreno is 4.3. Second highest has 2.9. Robertsonīs tackles per 90 is about 1.75 which is slightly lower than average but not dangerously low with a few players clocking in with less than 1.

This also passes the eye test too. He doesnīt strike me as someone who has a problem in knowing when you tackle or not. He's a scratch under  6 foot and has very good pace and backs himself in 1-v-1īs when the need arises. Whereas the eye test says Moreno dives into tackles way to much and his position probably means he needs to dive in even more on top as he is usually in the wrong place and needs to make a tackle just to solve that little problem he makes for himself.

Anyway Iīll be posting a similar table tomorrow complete with numbers instead of just rankings and player names. Iīll include tackles made per 90 in that also for you. Iīll also try to do as many categories as we can. Problem with attacking play is I think that will come down more to scouting that data. I just donīt have those metrics for things like packing that would show how effective they are from an attacking sense. Plus itīs really hard to compare full backs attacking numbers in a low block counter attacking side to those in a possession oriented side. Iīll add what I can though that I think is fairly measurable.

That's really interesting that Moreno has such a high number of tackles when his interception rate is fairly high as well.  His positioning, or lack of it, would certainly explain those stats. 

[rickardinho1]

Clyne-Good defender, rarely beaten by his man, caught out of position occasionally but what full back isn't. Very frustrating going forward quite often gets to the byline and just hits the first man with the cross.

Milner-Surprisingly very good defensively, quicker then you think he is aswell. Certainly much better then clyne going forward but has a tendency to just cut back onto his right foot and just loft crosses into our small forwards.

Moreno- Very likeable guy, has showing promise but lacks concentration/ a chromosome.

Robertson- Similar to clyne probably better going forward, but he's not as strong or quick and is worst in one v one situations.

Flano- Not sure how he's still a Liverpool player, was decent in fairness under rodgers for that half a season but so was the whole team. He can't run, he gets beaten by league 2 players in one v one's, he seems overweight, technically he has to be one of the worst players to play for the club. Surely will leave this summer, probably to a championship side.

Trent- Definitely looks a promising player, not sure what he's like positionally though, seems a bit of a headless chicken at times.

I don't think Flanagan will be sticking around too long, but that bit in bold is pretty harsh on him. I remember him in 2014 actually being quite a smooth passer on the ball, and his tackling is really solid... The tackle on Soldado is a perfect example of that. He obviously lacks that bit of extra skill and pace needed to contribute in attack though, and does get dribbled past quite a lot (statistically) which is why I think he'll be leaving this summer.

[BabuYagu]

That's really interesting that Moreno has such a high number of tackles when his interception rate is fairly high as well.  His positioning, or lack of it, would certainly explain those stats.

If Moreno had a footballing brain he would be a dream of a player. He would be the Keita of full backs. You look at someone like Woodburn who is just 5 years old and already plays the game on a level Moreno never could even comprehend.

Morenoīs numbers remind me of a quote from one of my footballing idols

If I have to make a tackle, Iīve already made a mistake
- Paolo Maldini

Tackling is certainly a skill and there are times you need to make a tackle for sure. Winning those tackles you need to make is therefore imperative so I value the tackle success rate metric. But out of everything I ranked, that is certainly the one I put least value in personally. I would rather a defender who made few fouls and was rarely beaten than one who was good at tackling.

I see tackles per 90 being a metric people place a lot of value in these days but I personally donīt value it myself as highly. I do value highly the tackles lost metric though as that indicates the amount of times you took yourself out of the game.

"Stay on your fucking feet you quilt"
- Rafa Benitez

That last quote might be tricky to source

[10 years on]

I don't think Flanagan will be sticking around too long, but that bit in bold is pretty harsh on him. I remember him in 2014 actually being quite a smooth passer on the ball, and his tackling is really solid... The tackle on Soldado is a perfect example of that. He obviously lacks that bit of extra skill and pace needed to contribute in attack though, and does get dribbled past quite a lot (statistically) which is why I think he'll be leaving this summer.

Is passing is average, if you play football every day of the week you have to be able to pass a ball. His tackling is definitely probably the strongest part of his game, but if your second touch is a tackle what's the point.

[BabuYagu]

I don't think Flanagan will be sticking around too long, but that bit in bold is pretty harsh on him. I remember him in 2014 actually being quite a smooth passer on the ball, and his tackling is really solid... The tackle on Soldado is a perfect example of that. He obviously lacks that bit of extra skill and pace needed to contribute in attack though, and does get dribbled past quite a lot (statistically) which is why I think he'll be leaving this summer.

The biggest problem I think Flanagan has defending 1-v-1 situations is that he gets himself to square. Hard to explain what I mean by that but basically you always want to be side-on facing the player and showing them away from what you consider to be the danger area. Flanagan gets into positions where he is facing the player which if you have ever played full back you will know it causes you problems. You can get easily nutmegged or a quick change of direction can leave you badly off balance or cross-legged. He also gets too tight one the player has turned.

In the below video there are two moments where his 1-v-1 skills would be tested. The first at 22 seconds in this video you can see what I mean. He is standing square when the ball is controlled, that can get knocked through his legs or Arnautovic can get him off balance with a change in direction easily. Luckily Arnautovic is off balance himself when controlling the ball which gives Flanno the chance to nip in and take the ball. 2:22 in the video he gets too tight once the player has turned and fouls. Two 1-v-1 scenarios and in both his body shape or positioning is wrong.

<a href="https://www.youtube.com/v/0-ZxFVk5-VA" target="_blank" rel="noopener noreferrer" class="bbc_link bbc_flash_disabled new_win">https://www.youtube.com/v/0-ZxFVk5-VA</a>

[BabuYagu]

Here are the actual names and numbers now for the first table.

Player Name	Tackling Success Rate	Tackles per 90	Fouls per 90	Dribbled Past per 90	1-v-1 Score	Comments for Scouts
Clyne	76.25%	2,16	0,38	0,62	1,00	No perceived Weakness
van Aanholt	74,42%	1,55	0,54	0,61	1,15	Passive?
Cresswell	71,43%	0,89	0,47	0,64	1,11	Very Passive?
Davies	71,79%	2,02	0,67	0,67	1,35	No perceived Weakness
M.Olsson	70,00%	0,72	0,86	0,22	1,08	Very Passive?
C.Daniels	71,19%	1,74	0,35	0,94	1,29	Dribbled Past too easily?
Clichy	72,73%	1,86	0,52	1,02	1,52	Dribbled Past too easily?
Robertson	67,92%	1,74	0,59	0,76	1,34	Tackling?
Fuchs	82,29%	2,78	0,84	1,51	2,35	Dribbled Past? Dives In?
K.Walker	70,83%	2,40	1,20	0,60	1,80	Fouls too much?
Bertrand	70,83%	1,33	0,86	0,47	1,33	Tackling? Very Passive?
Chilwell	78,26%	2,97	0,91	1,68	2,59	Dribbled Past? Dives In?
M.Alonso	67,86%	1,87	0,84	0,87	1,70	Tackling?
Friend	75,81%	2,85	1,29	1,01	2,30	Fouls? Dives In?
Nyom	71,62%	2,53	1,71	0,65	2,35	Fouls?
Moreno	74,80%	4,34	1,02	1,09	2,12	Dribbles Past? Dives in Most?
Pieters	77,14%	2,99	1,71	0,91	2,63	Fouls? Dives in lots?
Baines	69,84%	2,01	1,12	0,83	1,95	Fouls?
Ward	67,14%	1,95	0,97	0,86	1,84	Tackling?
Holebas	67,74%	1,92	1,33	0,68	2,01	Tackling? Fouls?
Monreal	67,78%	2,57	0,94	1,4	2,34	Tackling? Dribble Past? Dive In?
Rose	68,00%	2,94	1,53	1,23	2,76	Fouls? Dribble Past? Dive In?
Milner	67,02%	2,68	1,08	1,4	2,48	Tackling? Dribbles Past? Dives In?
Gibbs	58,33%	1,54	1,41	1,03	2,44	Literally everything. Also appears as passive. How?

I have also replaced average ranking with overall score which is basically fouls+dribbled past / total tackles made. Things to note though is that players who make almost no tackles at all like Olsson, Cresswell, Bertrand will all come out very highly here because itīs a failure rate table. And to quote my favourite TV show

"You cannot lose if you donīt play"
- Marla Daniels, The Wire

This probably takes 1-v-1 defending as far as I think we can with the data we have. Now you would need to look at the players individually and pay attention to those scouting comments to see how they pan out. Iīll let you know the next category and how we will work it out as soon as another work gap allows me.

[BabuYagu]

Arsenal, like Liverpool, had a problem with their full backs this summer. In Kolasinac (someone many championed us to sign) have they really improved? In terms of 1-v-1 defending, the numbers suggest not really.

Player Name	Tackling Success Rate	Tackles per 90	Fouls per 90	Dribbled Past per 90	1-v-1 Score	Comments for Scouts
Monreal	67,78%	2,57	0,94	1,40	2,34	Tackling? Dribble Past? Dive In?
Gibbs	58,33%	1,54	1,41	1,03	2,44	Literally everything.
Kolasinac	75,36%	3,00	1,70	0,91	2,61	Fouls? Dives In?

Weīll see how he checks out in other categories too. My initial thought is he looks a lot more like a center half playing left back and struggles in 1-v-1īs against more agile players as any center back tends to when they need to go out there. His stats are scarily similar to Erik Pieters who is pretty much the exact same height and build.

Oh and again note, its only because of Gibbīs very low number of actions per 90 that he ranks above anyone.

Whereas the best according to the table - Clyne - has a similar height and build to most wingers.

[BabuYagu]

Player Name	Tackling Success Rate	Tackles per 90	Fouls per 90	Dribbled Past per 90	1-v-1 Score	League Rank per Category & Overall
Robertson	67,92%	1,74	0,59	0,76	1,34	Rank 17th, 6th, 10th, 8th overall
Moreno	74,80%	4,34	1,02	1,09	2,12	Rank 6th, 15th, 19th, 15th overall
Milner	67,02%	2,68	1,08	1,4	2,48	Rank 22nd, 16th, 21st, 21st overall

Now this isnīt to say Milner is a bad defender. That he isnīt an upgrade on Moreno. etc. It merely indicates that maybe his weakness is 1-v-1 defending. I plan to go into a lot of detail on other aspects of full back defending and thatīs before we think about them going in the other direction with the ball at their feet which, as a former striker, winger and central midfield, Milner has the upper hand on the vast majority of full backs for sure.

[DanA]

Some great stuff and it mirrors up with what I think we see on the pitch.

I wonder though how much data the Moreno stuff is based on. Is it just his 290 odd minutes of league football?  I feel like the 4+ tackle figure is a bit of an outlier. My gut feeling though is that Moreno's biggest problem is positional, you see him get caught up the pitch constantly where as I don't remember seeing it happen to Milner at all.

I think Clyne is a really good fullback too, he just lacks a little going forward and isn't too good in the air. With a short team that's something to consider, I think a bigger LB would balance the squad a lot, even at a touch under 6 foot I think Robertson improves things a little.

[BabuYagu]

Some great stuff and it mirrors up with what I think we see on the pitch.

I wonder though how much data the Moreno stuff is based on. Is it just his 290 odd minutes of league football?  I feel like the 4+ tackle figure is a bit of an outlier. My gut feeling though is that Moreno's biggest problem is positional, you see him get caught up the pitch constantly where as I don't remember seeing it happen to Milner at all.

I think Clyne is a really good fullback too, he just lacks a little going forward and isn't too good in the air. With a short team that's something to consider, I think a bigger LB would balance the squad a lot, even at a touch under 6 foot I think Robertson improves things a little.

I used the 2015/16 season for Moreno. It was either that or leave him out. Everyone else is 2016/17.
Moreno's exact numbers used are 123 total tackles, 92 tackles won, 29 fouls conceded, 31 times dribbled past, 2548 minutes played. In roughly the same number of minutes last season, Ryan Bertrand attempted 37 tackles. Robertson 53. Marcos Alonso 56.

Morenoīs numbers are accurate. He loves tackling more than Rooney loves wearing kids pyjamas.

[DanA]

I used the 2015/16 season for Moreno. It was either that or leave him out. Everyone else is 2016/17.
Moreno's exact numbers used are 123 total tackles, 92 tackles won, 29 fouls conceded, 31 times dribbled past, 2548 minutes played. In roughly the same number of minutes last season, Ryan Bertrand attempted 37 tackles. Robertson 53. Marcos Alonso 56.

Morenoīs numbers are accurate. He loves tackling more than Rooney loves wearing kids pyjamas.

That makes sense, was thinking it otherwise wouldn't be enough info to include Moreno.

[Redcap]

I plan to go into a lot of detail on other aspects of full back defending

This is the bit I'll be most interested in, because I don't think it's borne out in public stats at the moment, which is what makes Moreno look like the Messi of fullbacks on sites like Squawka and Whoscored.

[Doc Red]

This is the bit I'll be most interested in, because I don't think it's borne out in public stats at the moment, which is what makes Moreno look like the Messi of fullbacks on sites like Squawka and Whoscored.

Some fantastic work there Babu, and again, appreciate the massive effort it must have taken to compile and list all that data!
I was always surprised to keep seeing Moreno's stats get trotted out showing him to be right at the top end in defensive stats such as tackling, even though with the ol' eye test it was plain to see his positioning was weak and he always dove in (though the odd brilliant saving tackle might have also made him look great).

So nice to see data compiled together to offer context that actual challenges our perceptions or confirmation bias (those that might have had).
We can kind of understand why Klopp was exceedingly hesitant to put Moreno on the field to the degree he'd put a 30yr+ old midfielder there instead.
 

[BabuYagu]

This is the bit I'll be most interested in, because I don't think it's borne out in public stats at the moment, which is what makes Moreno look like the Messi of fullbacks on sites like Squawka and Whoscored.

Well he already looks like a bag of shite when it comes to 1-v-1 defending here. Heīs 15th in 1-v-1 defending and I think this is probably his best category as this is where his pace most comes to use which might be all Moreno has that sets him apart as a footballer, explosiveness.

The problem I think we have is we look at players still in terms of what they are doing right. Lots of tackles, interceptions, etc. The problem is that itīs the opposite side of that which is the problem. When players make tackles but donīt get the ball. You are then out of the game. When you foul. When youīre dribbled past. When you step out to make an interception and miss (missed interceptions is a stat Iīve wantd for a long time for example).

Iīm not sure the data exists to properly show this. Even things like defensive errors seems to only show the very blatent errors that even an idiot like Savage could tell you it was an error. It would be nice to have a high level defenderīs opinion on what are defensive errors. Most defenders will make several errors per game. Not always resulting in goals or even being a huge problem, but just positional, or body shape. Itīs probably a pipe dream wanting that and itīs probably where scouting has to take over from analytics.

Flanagan, for example, always worried me in 1-v-1īs. Was never sure why. Then I saw a former Brazilian international full back on TV here showing correct body shape to deal with situations. He said things like this seems to be getting lost in coaching these days and the art of defending gets lost as more and more players become full backs as a fall back option later in their youth careers. I then went back and looked at Flanagan again with what he said ringing in my ears and he was spot on, he plants his feet square early and also gets too close once a they have turned. This allows him to be easily put off balance or needing to make a foul to stay with someone who makes a sharp turn. Instead he should be favouring a direction and showing the player towards it. If they go there, they are going away from danger and you can jockey him. It also makes it harder to turn if you have a 1/2 a yard headstart in that direction.

Thing I donīt get though is fullback isnīt really the fallback option for Flanagan. It just seems like something he does that was never coached out of him. I would expect other players who were moved to full back like Moreno, Milner, Rose etc all display similar body position & shape issues which results in them getting dribbled past or fouling more than average. They have the agility and acceleration to match people stride for stride, so I am assuming they are letting the winger lead them, instead of leading the winger where they want and setting up little traps for them to fall into.

I will also make the bold prediction that this doesnīt happen with natural fullbacks in Germany. They are so fucking detail oriented over there I just cannot imagine something like that gets past the quality checks in youth football. I have literally nothing to back that up with and would probably need to run my numbers on Germany, find full backs whose numbers are similar to our small, agile, fouling machines at full back and then watch them to see if they are making these little mistakes.

[Gnurglan]

Good stuff in here!

I wish I could learn to interpret stats, so this is helpful. Not surprised to see Clyne get good numbers. We never see him caught out the way we do with Moreno. He's just more solid.

I want to add something about body shape. It's so important. I remember when I watched a training session with a top division side over here. They had this simple training. Attacker tries to dribble defender in a 1-v-1. The defender's job was quite easy. Show the player a path. Once the attacker made his move, the defender would be expected to take a step forward and thus force the attacker away from danger even more. Basic defending, but my God what a difference between them. The most efficient defender was one who put his feet right and could time his move the best. Needless to say, he wasn't the fastest. But literally no-one got past him. His timing was excellent. I always thought that defender was shite until I saw that training. He was the kind that couldn't trap a bag of cement, but he could cancel out the best playmaker in the league.

[McrRed]

Effi'in brilliant, Babu. Love your work.

Bit unfair on Moreno as 15-16 was a stinker of a season for the whole defense. He never got a chance last season but the overall improvement across the back was encouraging.
I'm not saying he's Messi, just that the data for him is skewed.

His recovery stats are brilliant. His positional ones not so good - but only he, Klopp and the dressing room would know if that was by instruction.

Clynes good one on one but his back post figures are atrocious. TAA has a very small sample size but has excellent awareness and I was delighted to see him cut out back post threats a couple of times in his short tenure.

Basically, I think Klopp now has alternatives at full back, depending on the opposition attacking threat. Good, solid options and a couple of exciting ingenues to blood in.

[JCB]

Just spotted this thread and quickly skimmed through this; will have a proper read this weekend when I get some more time, however I quickly grabbed Babuyagu's data and threw it in Tableau to visualise it better for clarity and insight.

What I might attempt is to colour code defenders that play in high/low block teams and then display the stats to see if a pattern emerges. If you have any ideas let me know and I'll try to graph it this weekend as well.

NB. The size of the circle is the tackling success rate (normalised).





Edit: Updated table as per BabuYagu and Poetry

[BabuYagu]

Hey JCB, would love your help with charts some day. Thanks also for all the help sourcing data too.

One thing though, I would change your colour coding so that green is always the best outcome. At a glance, for example, Gibbs could appear better than Clyne in 1-v-1 situations which is like saying Fellaini appears as good as Messi at dribbling.

[JCB]

I quickly threw then together with no deep thought in output.. but will take your comments on board for the next time. More than happy to help with chart visualisation, I'll pm you either tonight or tomorow with some xG/xA data outputs as well

[PoetryInMotion]

Fouls per 90, Dribbled past 90 and 1 v 1 score are the lesser the better, so the color codes flipped (green for lowest) for those columns would make the visual perfect.

[Ravishing Rick Dude]

I know he's a c*nt,and plays for the biggest c*nt on the planet,but Antonio Valencia was very good as rb for them,and he's one of the most important players for them. Really really tough to beat...I haven't looked his stats though.

[BabuYagu]

I know he's a c*nt,and plays for the biggest c*nt on the planet,but Antonio Valencia was very good as rb for them,and he's one of the most important players for them. Really really tough to beat...I haven't looked his stats though.

Valencia has a weird tendancy to get sent of in friendlies. Remember that time he took out Sterling v England. That tackle against Salt Lake was pretty shit too. He wasnīt even trying to play the ball, just kicked the guy in both calf muscles for a laugh.

Agreed though, always thought he was decent for them. He makes great recovery runs.

[JackWard33]

Nice work Babu

Some of these indicators are slightly flawed (e.g. is a full back better or worse when they're making more tackles?) and ideally they'd be possession adjusted (and one day game state adjusted but we don't have that yet)

What is pretty clear from this (and from detailed passing stats) is that as a fan base we under rate Clyne and over rate Milner
I'm not totally sure why that is but it's definitely happening a lot

[BabuYagu]

Tableau is some excellent shit to play with. Thanks again there Seņor JCB!!!



Will make another table excluding Clyne, Walker, Ghoulam & Kolasinac seeing as they are either not LB (Clyne played 1 game there in 2 years) or  they are not in the Premier League last season.

[Ravishing Rick Dude]

Valencia has a weird tendancy to get sent of in friendlies. Remember that time he took out Sterling v England. That tackle against Salt Lake was pretty shit too. He wasnīt even trying to play the ball, just kicked the guy in both calf muscles for a laugh.

Agreed though, always thought he was decent for them. He makes great recovery runs.

Yeah, he's a red card waiting to happen. Absolutely prick of a player

[BabuYagu]

Green
This grouping would be what I would consider the best grouping of left backs in 1-v-1 situations in the Premier League at the moment. I would say for that grouping Davies & van Aanholt look the best, although van Aanholt will be 27 when the season starts. We seem to target players a few years younger than that so we have a core squad coming into their peak 2 or 3 years from now. With this in mind, and Davies likely unavailable (or over priced for what we want to spend) as he is at Spurs, Robertson then looks the best option available to us.

Light Green
Bertrand, Cresswell & Olsson may also be very good options although the very low number of tackles they are making is a concern. Are they simply not being tested much in 1-v-1 situations? Are they world class at staying on their feet and jockeying people away from danger without making a challenge? Or are they too passive and simply back off ensuring they are rarely beaten, foul or make a challenge but cause a lot of problems for their sides?

Yellow
This would be the league average group. Alonso was the surprising name in there for me. Although maybe this explains why Chelsea want a 60m left back this summer?

Orange
This would be the group that seem weak in 1-v-1 situations. They are all fouling & being dribbled past more than twice as often as Nathaniel Clyne. Interesting names in group are Chilwell (bullet dodged or still learning?), Milner (sorry James, but I figured this would be his weakest category given he is not a defender and at 31 will be losing some of that explosiveness) Monreal (explains Arsenals need for a new leftback too) and Danny Rose (always thought he was a liability in defence, his improved attacking game and better center backs mopping up behind him maybe is seeing his weaknesses exposed less?)

Outliers

Gibbs & Moreno are both way out on their own, and not in a good way for either. Moreno looks like he could be a good fullback. Hell, maybe even a world class one. However, again, his decision making comes into play. He dives into every challenge. He attempts 4,3 tackles per 90 compared to the second highest in the league 2,99. The league average is 2,1. If he stayed on his feet half the time you expect he would simultaneously reduce his fouling and dribble past numbers too. You 1/2 all his numbers and he would be the best full back in the league in this category by some distance. But that has always been the problem with Alberto, if. He has everything a full back could need to be the Naby Keita of full backs except the brain of one.

Why are we not looking at Kieran Gibbs?

We just looked at him and the view is one of pure shite. He rarely makes a tackle but when he does, he fouls 3 people at once and simultaneously gets dribbled past too, somehow. His tackle success is the worst in the league too.This probably explains why at 27 years old he is leaving his physical peak, has never been first choice and averages about 10 full games per season in minutes played.  The only positive is he hasnīt killed any kid with one of his aimless tackles. Yet.

[BabuYagu]

Okay I think that concludes 1-v-1 defending for full backs. I think next weīll try to look at ability to prevent passes and crosses into the box. Have a feeling Milner & Robertson will both do well here. Letīs see.

[BabuYagu]

Nice work Babu

Some of these indicators are slightly flawed (e.g. is a full back better or worse when they're making more tackles?) and ideally they'd be possession adjusted (and one day game state adjusted but we don't have that yet)

What is pretty clear from this (and from detailed passing stats) is that as a fan base we under rate Clyne and over rate Milner
I'm not totally sure why that is but it's definitely happening a lot

Cheers mate

I mentioned this about tackles. Ideally, you want him to never tackle and instead just shepherd the player back towards his own goal away from danger. So players who tackle more than average I highlighted as being a problem they might "Dive in" whereas the opposite might also be true. If you never make a tackle, you will rarely commit fouls or be beaten. Is that because you did your job well or just backed off inviting the player to advance into more dangerous areas? Stats donīt really show this though so weīd need to look at the players involved.

Possession adjusted tackles is something I see Ted Knutsen do. But then it occurred to me that teams that have the most possession (Spurs & Liverpool & Arsenal) also have the most tackle happy left backs (Rose, Milner, Moreno, Monreal) whereas teams like Southampton, West Ham and Swansea have tackle shy full backs in Bertrand, Cresswell & Olsson.

Then you have Gibbs at Arsenal on the flip side. Which makes me wondering whether number of tackles is more down to the individual concerned than anything else. I reckon Moreno could join Barcelona and play in one of their 80% possession matches and still manage to rack up 5+ tackles just be aiming a few kicks at Busquets to get it out of his system.

The thing I worked out when looking at this is that there is only 1 world class full back in the premier league in dealing with 1v1 situations and itīs Nathanial Clyne. His numbers are sick. Robertson looks like a huge upgrade at LB for dealing with 1v1īs based on this. Hope it comes out like that too. We could use someone who locks down wingers on that side of the pitch the way Clyne does on the other side.

[BabuYagu]

Effi'in brilliant, Babu. Love your work.

Bit unfair on Moreno as 15-16 was a stinker of a season for the whole defense. He never got a chance last season but the overall improvement across the back was encouraging.
I'm not saying he's Messi, just that the data for him is skewed.

His recovery stats are brilliant. His positional ones not so good - but only he, Klopp and the dressing room would know if that was by instruction.

Clynes good one on one but his back post figures are atrocious. TAA has a very small sample size but has excellent awareness and I was delighted to see him cut out back post threats a couple of times in his short tenure.

Basically, I think Klopp now has alternatives at full back, depending on the opposition attacking threat. Good, solid options and a couple of exciting ingenues to blood in.

I am starting to believe that these numbers show "traits" in players, if you will. And that you can pick someone up and put him in a different team, or different league, or different system and he will always perform in a similar way. To test this I decided to look back for a few seasons at Moreno & Clyne to see what the numbers show. Just for kicks Iīll add in all full backs since Rodgers. My guess will be that Cissokho, Enrique will appear as good. Johnson average. Flanagan similar to Moreno.

[Redcap]

Possession adjusted tackles is something I see Ted Knutsen do. But then it occurred to me that teams that have the most possession (Spurs & Liverpool & Arsenal) also have the most tackle happy left backs (Rose, Milner, Moreno, Monreal) whereas teams like Southampton, West Ham and Swansea have tackle shy full backs in Bertrand, Cresswell & Olsson.

Is tackling is something you do in the middle third of the pitch? I mean, to take a random, arbitrary example of a back four player - Lucas made 1.5 tackles and 1.5 interceptions per game last season, playing mostly as a CB. Compare that to when he played more as a DM in 15/16, when he made 4 tackles and 1.9 interceptions per game. Similarly, you wouldn't see as many tackles made up in the final third of the pitch, and not just because it's mostly occupied by attackers.

My thesis is the number of tackles a player makes is mainly a function of:

1. The primary zone they operate in

a) Middle third players will make more tackles than defensive and attacking thirds
b) The amount of opponent's attacks that go through a player's part of the pitch
c) The amount of responsibility a player has for defending that part of the pitch

2. Their individual characteristics, comprising:

a) Positioning
b) Recovery pace
c) Intelligence (knowing when to stay and when to go)
d) Actual ability to make a tackle

I suspect a reason Moreno has been relatively less successful in his defensive responsibilities, and more successful in his offensive responsibilities, is Coutinho. Coutinho's been our main left-sided attacker for the last few seasons. He's certainly no slouch defensively, but you also won't see him sprinting back to make a tackle near our goal line to help his fullback. I suspect that's because he's neither quick enough, nor has sufficient stamina to do that very much. On the other hand, he's also been our best attacking outlet, which I think is why in 2015/16, there was a table that showed that one our biggest key pass generation combinations was Moreno to Coutinho. Coutinho cutting in with Moreno overlapping is also a natural attacking combination that takes advantage of both players' strengths. This goes to 1c) and also to a lesser extent 1b). Moreno probably had more responsibility covering his part of the pitch than say.. Clyne (Mané) or Clyne (Milner) in the last couple of seasons. Not ideal, when his sense of positioning wasn't good to begin with, and he tended to rely on recovery pace.

I don't have data to support any of this. Someone will probably produce a heat map that shows all the tackles he makes are central and outside our area