- Part 1: The manager and the mentality
- Part 2: The real problem area – midfield
- Part 3: The Sarcevic factor
- Part 4: Defensive disaster?
- Part 5: Defensive disaster II
- Part 6: Taylor or Ladapo?
- Part 7: The creative void
Argyle’s centre backs: Edwards, Wootton, Canavan, Grant. These four have probably been collectively scrutinised more than any group in Argyle’s squad. It’s easy to see why: this season, Argyle have conceded at least three goals more frequently than they have kept a clean sheet, while only three sides in League One have conceded more times.
So far, parts 1, 2 and 3 of this series have all focused on why the manager and midfield have contributed to Argyle’s shocking defensive start, and the lack of control exerted, and protection offered, by Argyle’s midfield has certainly been a majorly significant factor. However, the defence itself must take its portion of the blame, and lessons need to be learned to turn the situation around.
Before we progress, though, a quick disclaimer. The defensive positions (goalkeeper, full-back, wing-back and centre-back) are just about the hardest positions to analyse, especially using statistical models. Consider either midfielders or attackers. Both are quite easy to understand because they are trying to make something happen – they’re trying to dribble past someone, create a chance or score a goal etc. – and this can be easily recorded and analysed.
On the other hand, defenders are trying to stop something from happening. When an attacker makes his move, a defender must respond: if they try to dribble, they must tackle; if they try to pass, they must intercept; if they try to shoot, they must block. In this way, a defender’s actions are a response to the attacker. For example, the number of tackles one amasses is just as much of a reflection of the number of times a player tried to dribble past them as proof of how good they are at tackling.
Let me demonstrate this using the data I have gathered. Collectively, Argyle’s centre-backs have completed 651 defensive actions this season, of which 540 – 82.9% – have been interceptions. Strip that back even further and you’ll see that just over half of those interceptions, and roughly 40% of all defensive actions, are defensive headers, the majority of which are against goal-kicks and long passes from defence. Therefore, on average, roughly a third of defensive actions a centre-back completes amounts to little more than successfully contesting aerial duels from long, aimless passes.
In fact, on average each central defender contests an aerial duel every 6.32 minutes, but only make a tackle every 39.0 minutes. This doesn’t mean that they are collectively better at winning headers than making tackles, but that the opposition challenge them to headers more often than they dribble at them. The same goes for full-backs and goalkeepers, whose save percentage is as much a reflection of the quality of the shots they face as their skill.
So, what does this mean? Well, firstly, there are a series of firm conclusions that can be drawn using data, but most require even more contextualisation and analysis than usual to generate any accurate conclusions. Second, because of this extra analysis that is required to understand the chronic defensive issues that are plaguing Argyle, this article being split into two. This part will focus on aerial duels and the midfield’s role in the defensive weaknesses; the second will focus on defensive errors and Matt Macey. With that in mind, let’s start with that which we’ve already touched upon: aerial duels.
For centre-backs, being able to successfully compete in aerial duels is absolutely critical. As has already been shown, it is the test they will face more often than any other. Should they be unable to win headers, then failure will inevitably stalk their career. Which leads us to Scott Wootton. He has faced a lot of criticism following his performances at the start of the season, and much of it has been justified. However, his biggest issue is that he is, by far, Argyle’s worse defender when it comes to aerial duels; only Gary Sawyer, a full-back by trade, struggled more during his short spell at centre-back.
2018/19 | Aerial won/90 | Aerial lost/90 | Aerial success (%) |
Edwards | 11.20 | 4.64 | 70.71% |
Canavan | 7.92 | 4.34 | 64.62% |
Grant | 7.50 | 4.34 | 63.33% |
Wootton* | 5.35 | 7.96 | 40.20% |
Sawyer* | 5.31 | 9.92 | 34.88% |
*excludes matches when Wootton or Sawyer played at full-back
I’ve spent a lot of time in this series of articles writing about how a negative mindset allows for the opposition to play further up the field, put Argyle’s defence under more pressure and create more goal-scoring chances. Yet, a player who consistently loses defensive headers poses an even greater threat. Let me give you some examples of how this impacts a team defensively. First, a quick example of how losing a header from a simple goal-kick helps a team gain ground:
Wootton lost the header, Edwards was forced into a half-clearance to guarantee that McLaughlin (#11) could not run through on goal, and from there Southend’s midfield are were possession beyond Argyle’s midfield. That was the moment that the danger emerged, when Argyle’s defence was exposed to the opposition. From that moment, Southend had the opportunity to exploit the space that was generated by the lost header. They worked it to the edge of the area and almost grabbed a late winner.
Let’s look at another one, this time one that does lead to a goal, a late equaliser for Wycombe:
This time, Wootton positioned himself perfectly to deal with the long ball, but was nudged out of position far too easily by Akinfenwa. Wycombe’s striker took the ball down and set up the equaliser three touches later.
And, of course, there’s the worst example:
I sincerely hope that there is no need to explain what he did wrong there. Sadly, that’s not a one off; here’s another example you might all remember from last season:
Ciftci – for ten man Argyle – was sent clean through by, again, a simple long ball from defence that went straight over Wootton’s head. He may not have scored, but as a defender you can’t allow a simple pass to beat the entire team and an opponent a one-on-one opportunity.
Next, the corner from which Charlton netted their equaliser last month:
Macey came off his line but could not reach the ball – possibly due to an infringement by Lyle Taylor – but it was Wootton who lost the header at the back post to set up the equaliser. It’s not the first time he has lost a header from a cross leading to a goal.
Below (last example, I promise) you will find back-to-back highlights of Wootton losing headers leading to goals – the second of which was ruled out for off-side.
In that second highlight, Wootton lost a header in the box to Jack Marriott, a player 15cm (6in) smaller than him. That is – quite frankly – not good enough for a player who should be expected to challenge for an aerial duel every six minutes.
Wootton has the height, but lacks the technique, timing and strength required to win aerial duels with the regularity that is required of a central defender. There are, for certain, many other factors that determine the success and failure of such a player – and Wootton has other defensive weaknesses that need improving – but aerial ability is such a core requirement of being a central-defender that if you don’t have it, then you are immediately a liability.
Against Southend, Wycombe and Peterborough, he won just 34.6%, 27.2% and 27.2% of his aerial duels respectively, losing 33 in total. In each match, the opposition was able to rapidly gain ground simply by hitting long passes to Tom Hopper, Craig Mackail-Smith or Matt Godden, and each was able to bring the ball into play beyond Argyle’s midfield with alarming consistency.
Even without considering his further strengths and weaknesses, starting Wootton at centre-back runs the risk of allowing the opposition to rapidly transition from defence to attack and gain possession in dangerous areas by just playing long, direct passes. This is his greatest weakness and so long as it remains unaddressed he must not start.
For certain, other players have made errors in this way, the most notable of which was Edwards’ lost header against Tom Elliott as Millwall won a penalty in the League Cup tie earlier this season. Yet, Wootton poses this threat more than anybody else, and for that reason simply cannot start at centre-back unless it is necessary. It’s not that he’s guaranteed to give away a goal, but, unfortunately, with him in the side the risk is substantially higher.
Wootton aside, the remaining trio have all demonstrated enough to suggest that any pairing – most likely between Edwards and one of Canavan or Grant – could form the rear-guard of a successful Argyle side in the same mould as during last season’s run. Though a lot of discussion has centred around the absence of Zak Vyner from this side as a quicker foil for one of these stronger, slower centre-backs, pointing to that as the key defensive deficiency is misguided. While a player akin to Vyner would probably improve the defence slightly, one is not necessary to turn things around.
Don’t forget, last season, Argyle’s charge up the table started with an eight-match unbeaten run between Gillingham and Doncaster which saw Edwards and Bradley (whom I once saw described as having the turning circle of a milk float) successfully paired together. Only six were conceded in that time, four of which came from set pieces and none as a result of an opponent directly dribbling at – and past – the duo. The absence of smaller, quicker centre-back certainly wasn’t felt then.
Indeed, just consider league leaders Portsmouth. Their centre-backs (Clarke, Burgess, and Whatmough) have an average height of 188cm and none are anywhere near as quick as Vyner. Yet, only two sides have conceded fewer goals. In fact, there are barely any sides in this division that could claim to have a player similar in style to Vyner – Curtis Nelson at Oxford and Rhys Bennett at Peterborough are the only obvious examples that spring to my mind – and how many other sides are seeing their defenders exposed quite as much as Argyle are?
A point that has been made repeatedly, especially in the discussions on Fox and Sarcevic, is that the performance of the midfield heavily determines the success of the defence. The best defences begin from the front: first by denying the opposition control of the ball through possession retention; second by thoroughly shielding the defence from the opposition midfield. From the moment the opposition break the midfield’s defensive line, they are in the position to create goal-scoring chances.
The more that line fails, the more chances the opposition will create, the more shots they will take and the more goals they will score. Whereas the current defence has been more frequently failed by their midfield this season, there were only a handful of matches in which Vyner was equally as exposed last season, the main three being Charlton, Rotherham and Gillingham. In each, he struggled, just as any defender would. Take Gillingham’s first goal last season:
The goal may appear to come from the final, defence-splitting pass to release Parker, but it really originated the pass before that, when Reilly broke the midfield line by slipping the ball into Martin. From that moment, Vyner had to reposition himself. Instead of blocking the passing lane to Parker, he had to close the gap between himself and Bradley because – with Martin running at the heart of the defence – that became the biggest danger. In repositioning himself, Vyner opened up the passing lane for Martin to diagonally slip the ball between himself and Miller. One pass through the midfield, and Gillingham were able to open Argyle’s defence up.
That goal was not Vyner’s fault. He did exactly as he should have done and prioritised the biggest danger to the goal ahead of the secondary threat. That all came from a pass that broke the midfield line.
Here’s the thing: no matter what your instinct tells you, few (if any) central-defenders are designed to be dribbled at by an opponent. Instead, centre-backs are designed to win headers and be dominant in close-quarters in order to deal with long-balls and crosses while blocking, tackling or challenging for the ball with individuals in a crowded penalty area, because they have to deal with these scenarios far – far – more often than they are challenged to a foot-race. In fact, should a centre-back find himself repeatedly in a situation like this, then it is a warning sign that there is a wider problem with the team.
In Vyner’s case, he was not successful at Argyle because of his speed and agility, that was just the cherry on the cake. Instead, he did what normal centre-backs do. He averaged 17.21 defensive actions – more than Wootton but less than Canavan, Edwards and Grant, since he had less to do behind a better midfield – of which more than 90% were interceptions. For all the talk of Vyner being a more desirable player to improve this defence, he actually averaged 37.5% fewer dispossessions than Bradley during last season’s run, the slower, stronger alternative.
Vyner did indeed perform very well for Argyle, but he played in a side that protected him far better than this current one protects their defence. As a result, though Vyner could deal with one-versus-one situations better than this current crop, he was only infrequently required to do so. It wasn’t his speed that made him successful, it was his ability to deal with regular defensive scenarios behind a midfield that prevented him from being exposed on a regular basis.
This can be seen in the shot data from this season. I group all shots into the following sections: penalties, direct free-kicks, set pieces, counter attacks, rebounds, dribbles and open play. Compared to last season’s run (from Oldham to Peterborough) there are a series of obvious trends that demonstrate how exposed Argyle’s defence has been this season.
Firstly, there has been a colossal rise in counter attacks. Compared to last season, opposition sides are both taking these shots and scoring from them more than four-times more frequently, which has resulted in the percentage of counter attacking shots, out of total shots, climbing from less than 4% to nearly 15%.
Counter attacking shots | Oldham – Peterborough 2017-18 | 2018/19 |
Shots/90 | 0.31 | 1.70 |
Goals/90 | 0.10 | 0.43 |
Of total shots (%) | 3.59% | 14.6% |
On target (%) | 83.3% | 57.1% |
Conversion rate (%) | 33.3% | 25.0% |
Across this season and last, counter attacks have the highest conversion rate – last season’s run once every third shot, this season once every fourth shot. This is because counter attacks provide one of the best opportunities to expose defenders without midfield protection. With the frequency of counter attacks more than quadrupling, it’s not surprising that Argyle are conceding a counter attacking goal – on average – just shy of every other game.
But it’s not only counter attacks that are the problem. The ease with which teams are breaking Argyle’s midfield line can also be seen from open play too, as opposition shots are more frequently coming from less defensive pressure which is a result of the lack adequate midfield protection:
Open play shots | Oldham – Peterborough 2017/18 | 2018/19 |
No marking (%) | 13.1% | 18.7 |
No marking/90 | 0.71 | 1.22 |
Stand-off defending (%) | 64.5% | 68.2% |
Stand-off defending/90 | 3.51 | 4.44 |
Man marking (%) | 16.8% | 12.2% |
Man marking/90 | 0.92 | 0.79 |
Double marking (%) | 4.67% | 0.93% |
Double marking/90 | 0.25 | 0.06 |
Allow me to elaborate more on what this means, for the sake of clarity. From open play, sides are increasingly more likely to be shooting at goal with either no marking (that is to say that there is no defender between the player shooting and the goalkeeper). Additionally, they are more likely to be shooting at goal with a defender between them and goal, but not pressuring the attacker. Meanwhile, on both accounts they are less likely to be directly man-marked or double-marked.
These shots are only from open play. This does not include counter attacks, set pieces, rebounds etc. When in possession, opposition sides are currently able to create more chances to shoot with less defensive pressure compared to last season.
Let’s take this a step further. Remember the shot matrix diagram from part 3? For those who don’t zones 1, 2 and 3 are collectively the “danger zone”, the areas from which the opposition are most likely to score.
This time, we’re only looking at open play shots from the “danger zone”.
Open play shots | Oldham – Peterborough | 2018/19 |
“Danger zone” shots/90 | 1.93 | 2.31 |
No marking (%) | 28.2% | 36.8% |
No marking/90 | 0.56 | 0.85 |
Stand-off defending (%) | 33.3% | 36.8% |
Stand-off defending/90 | 0.66 | 0.85 |
Man marking (%) | 28.2% | 23.8% |
Man marking/90 | 0.56 | 0.55 |
Double marking (%) | 10.3% | 2.63% |
Double marking/90 | 0.15 | 0.06 |
Not only does this side average 20% more shots inside the “danger zone” per-90 minutes than last season’s side, they are substantially more likely to be unmarked or unpressured when they shoot. It goes without saying that an attacker who shoots in these conditions is far more likely to score than one who is marked. It’s no wonder that last season’s side averaged an open play goal every 197 minutes compared to one every 135 this season.
At this point, you might be thinking that surely this is just evidence for how poor the defenders have been since they are allowing opponents to take shots under less pressure. This is where the midfield comes into it.
Think about the example of Vyner that I gave earlier. In it, once the midfield line was broken, he had to reposition himself to deal with the attacking threat, which open up the space for Gillingham’s striker to be played in for an unmarked shot at goal. While there are other causes of this – for example Wootton losing the header against Southend that resulted in a shot at goal or a defender giving the ball away, a la Canavan against Peterborough (more about that in the next part) – the flimsy midfield line being broken by the opposition, through a pass or a dribble, has been the most common cause of the defence being exposed.
in 2018/19, there have been many instances of this; Doncaster’s second is one such example:
Here, if you were to pause the clip after the midfield line was broken and the ball was played into Wilks, you’d see the following: Grant is blocking Wilks from shooting or crossing across the six-yard box; Edwards is marking Marquis; Wylde is attempting to block a cut-back to Kane and Smith-Brown is guarding the far post. All are in the positions they need to be in.
Unfortunately, neither Fox nor Songo’o drop into the position they are supposed to be in – the one which would prevent the simple pass across goal – and so Blair was found with remarkable ease and finished under no pressure ahead of both midfielders. The goal started and ended with Argyle’s midfield line being broken.
Here’s another example – one that didn’t end in a goal – for Coventry.
Instead of beating the midfield line, they beat the defence on the flank. However, with Songo’o out of position, Wootton was isolated in two-versus-one situation which allowed Andreu the time to get off an uncontested shot.
Peterborough’s fifth:
Same story: beat the midfield line; pass the ball through the defence; attackers engineer a shot on goal.
Against Portsmouth, the opening goal came after Portsmouth beat Argyle’s midfield line and played a precision pass to dissect Argyle’s defence, but an even better example came minutes after:
Beat the midfield line; pass the ball through the defence; attacker engineers an unchallenged shot at goal.
Swindon’s second?
Beat the midfield line; lay the ball to midfield runner who sprints through Argyle’s defence unmarked and gets off a relatively uncontested shot.
Oxford’s opener yesterday? You guessed it. They beat the midfield line. They passed it through Argyle’s defence (though, yes, Edwards should have done better at his attempted interception, instead of allowing it to squirm through), and Mackie scored with no defenders between him and goal.
It’s not the case that all of the shots against Argyle – or all of the goal conceded – have come as a result of the opposition beating the midfield line. However, it is a majorly significant factor in explaining how the rate of uncontested shots has increased, and with it the rate of goals. More frequently, Argyle’s midfield line is being broken with ease from open play to pose problems for Argyle’s defence and generate more clear goal-scoring opportunities.
You might have also noticed that, in those examples, the centre-backs themselves were not dribbled past once; instead, the ball was passed around them.
Despite being slower and less agile than most of the strikers they have faced, players like Edwards and Canavan have only been dribbled past 0.24 and 0.19 times per-90 respectively, meaning that – collectively – only one of them is beaten every other game. Meanwhile, the duo’s ratio of being beaten to making tackles is roughly 1:3. In other words, they are rarely required to be quick and agile in the presence of a striker, and when they are they win the ball back more often than they are beaten.
2018/19 | Edwards | Canavan |
Beaten/90 | 0.24 | 0.19 |
Tackles/90 | 0.80 | 0.57 |
Beaten : Tackle | 0.30:1 | 0.33:1 |
Indeed, because of the way most teams – or the ones that aren’t total disasters – are set up, centre-backs shouldn’t have to deal with a one-versus-one situation against a striker who has the space to run at, and past, them. However, even in most of these cases the centre-backs aren’t dribbled past because there is no need to. Once you – as an attacker – are put into a position like this, it is far easier to pass the ball around the defender and into a player who is running into the vacated space.
Watch those highlights back again and notice how – once the midfield line is broken – the defence is forced to converge around the player who is attacking the goal with the ball to deny them the opportunity to shoot. Again, just as in that first example I showed you regarding Vyner, this repositioning to face the immediate threat results in nothing other than space for another attacker, and it is that attacker who will be the one who will actually take the shot. All of which (the attacker running at goal, the defenders converging around him and the subsequent space generated) is a direct result of the midfield line being broken.
Overall, though Argyle’s centre-backs are taking the brunt of the criticism, a majorly significant reason for the increase in chances, big chances and goals has been the inadequate midfield protection, either through poor ball retention, poor defensive awareness or poor positioning. While the next part of this series will focus in greater detail on the individual errors made by defenders, the role of the full-backs in all of this and Macey’s own issues, the most significant individual factor in Argyle’s horrific defending of late has been the midfield: it wasn’t for lack of reason that I dubbed it ‘the real problem area’.