Three Team Possession: What Is It Good For?

In this article I will explore three team possession – when and where it has value, some lower complexity activities that can be used to build up to it, and how adding constraints can diminish (or enhance) its value. I have chosen to explore this activity because I have seen used quite frequently within the North American soccer sphere and I think it is very much a double-edged sword.

I want to begin by analyzing what this activity is good for using the 5Ws framework (who, what, when, where and why). Because the activity is non-directional and by default there are no positions, a lot of context of the normal game does not transfer. Notably the attacking principle of Penetration is almost completely lost and the Who/Where dynamic suffers greatly for it. The one thing that will always transfer is the What, the repetitions required for skill development (first touch, dribbling, striking, scanning/multi-object tracking). Most of the other attacking and defending principles transfer to this activity even if how to utilize them in a directional context is lost. For example, I will encourage (young) players to think about the concepts of Support and Width by using coaching points like “two small passes, one big pass.” This phrasing is intended to give players roles to fill and implicitly ask them to look for their teammates and make decisions about where to move based on what options the player in possession already has. In keeping with the emphasis on multi-object tracking, because defenders are numbers down, third defender screening and denying multiple passes by occupying key spaces also has heightened value if the teams are big enough.

The relatively relaxed structure of the activity means that there is plenty of room for general tactical development (decision-making) as well, at least in the context of keeping possession (e.g. When pressure is great, come closer and offer a wider angle of support). I break the attacking mission down into three parts for young players: 1) keep the ball, 2) play forward and 3) create chances to score. Three team possession gives players practice at accomplishing the first. Once players have ability to consistently maintain possession using the overload, they need to be pushed again. In my situation, I was most recently working with U9/U10 boys who had never played larger than 4 v 4. Most of the group were not good enough to consistently execute the core fundamentals of the game (scanning before receiving, intelligent first touch into space, pass to another player who will also be able to retain possession despite whatever pressure they face). This activity therefore had value, as it gave them the opportunity to practice these skills in an overload situation (4 v 2, 6 v 3) where they could find success (as defined by retention of possession) at least 50% of the time.

For my boys, my preferred lead-in activity to three team possession was 3 v 1 rondos. The 3 v 1 rondo is most useful in building in the Support concept while giving a player technical repetition. At its core, the activity relies on the notion that one defender should not be able take away both passing options at the same time, so two players who simultaneously interpret a situation (pressuring defender) the same way should be able to work together to find a pass that moves the ball further away from pressure and makes retaining possession easier. Where the overload is greater (e.g. 5 v 2) there is less emphasis on the Mobility of the players, but in 3 v 1 the need to readjust is constant. Each player is always the next pass, rather than potentially being two or three passes down the line. This ensures that no time is wasted – every action by every player matters in every single moment. I found it necessary to rotate the defenders on timers rather than on winning possession to ensure that players were not continually stuck in the middle for whatever reason. This can be made competitive by keeping score of the possession chains scored against each defender over a given duration. If your players need this activity, they might not be able to count and keep the ball simultaneously, so bear that in mind.

Two specific variations of 3 v 1 rondos to get players used to movement are the open corner and open side rondos. These can also be done without defenders to impart the basic concept if necessary. These rondos require the attacking players to always support the possession with passes to either connected side or corner. The basic idea is that every time the ball moves, the players must move as well to ensure the player in possession has two passes. I prefer the open side rondo to the corner rondo primarily because there is a spectrum of supportive positions the player can choose from rather than a binary “you are in the right place” choice presented by the corner model, but the binary aspect can be useful in creating the expectation of movement. Once players understand the basic concept of support you should be able to do away with these stipulated movements and let the players have freedom in a 3 v 1, which includes the freedom to stay in place if they can complete the mission (retaining possession) without moving. Once players have success finding ‘the’ open player with just one defender the complexity can be scaled up. This brings us first two 4 v 2 and then 6 v 3 (which is more a matter of convenience than preserving the concept of overloading by two).      

For lead-ins to the 3 v 1 rondos, I would usually opt for something that emphasizes technical repetition with two players passing the ball and taking directional first touches (as opposed to stopping the ball). Sometimes this is two players standing still and doing ball mastery/change of direction techniques between their first touch and the return pass, but this is also a good opportunity to introduce players to different movement forms that enable scanning, such as side-shuffling and backing away, as well as asking for the ball to specific areas (a space they are running into or to left/right foot), which requires the passer to look at their target and interpret their actions. You can also ask them to experiment with the type of pass they are giving here, both in the foot it comes from and the surface (inside, lace, outside) chosen.

Once players are finding some levels of success within the basic three team possession activity it can be modified with specific restrictions or incentives to focus in more detail on certain concepts or skills. The USSF espouses a constraints-based approach to practice design without ever explicitly using that terminology (my experience). I would say this is unfortunate as codification and shared terminology can be an important part of allowing two people to communicate clearly, though perhaps this terminology was not prevalent when the coursework was designed or when I took it. I am linking a basic overview of the concept from someone I came across when I was exploring this topic at the time.

From here it is my intent to explore different types of constraints that can be imposed to place emphasis on specific actions and behaviors. Before getting too deep into the weeds,  I want to take a brief detour into the Hippocratic Oath and the concept of non-maleficence, or in layman’s terms “first, do not harm.” If your practice is the sum of all the good design and well-delivered coaching points less all the negative knee-jerk reactions and negative design elements that go unaddressed, then both sides of the equations can min-maxed to create progress. What I see all too often is that in the intent to alter one side of the equation, there are knock-on effects to the other side (economics uses the term externalities), often running counter to the intended effect. A non-soccer example for this is MRSA (methicillin-resistant staphylococcus aureus), wherein our efforts to control or eliminate bacteria populations have inadvertently created strains of bacteria that resist some forms of pharmaceutical treatment. This occurrence makes the broader scope mission (maintaining a healthy populace) more difficult than it had been previously.

Probably one of the most commonplace soccer examples of this is a “two touches or less” mandate. This requirement is well-meaning in its intention: if a player has fewer touches available to them then they should shift their information-gathering processes to before they receive the ball, rather than only engaging these skills after receiving possession and using up half their allotted touches. It also increases repetition volume be ensuring the continuous transfer of the ball. The issue with this mandate it that it restricts a player’s opportunities to make meaningful choices (and therefore learn) while selecting for a quality that is not even universally desirable.

To explore that in a bit more detail, seeing a player’s unrestricted actions ensures we are working to get them from where they really are to where they need to be instead of viewing them through the forced perspective of two-touch play. The first step toward enacting real change is to understand reality as it is, rather than for what you hope it might be. Selecting for specific kinds of decision-making means limiting the types of mistakes a player can make. This in turn limits the type of feedback the player receives, both in experiencing the outcome of their decision and whatever feedback the coach tacks on. Giving a player freedom should result in an accurate assessment of the player which enables the coach to create an alignment between the specific needs of each player and the feedback they receive.  

Moving on to the merits of tempo, we want players who voluntarily choose to play quickly when it makes sense to do so. Two almost ubiquitous examples of this are 1) the player is faced with pressure and possession is risked or 2) increased tempo will create even more space for the next attacker. Because these two situations account for most situations where a player receives the ball, enforcing high tempo through constraints is not harmful in those instances. However, there are instances where there are no good options available within two touches (at least, within a player’s skill ceiling) or taking three/four will create more success (space) for the attacking team.

Finally, to return to externalities, two-touch creates situations where a player is punished for touching the ball during the run of play, which is unthinkable in the normal game. This necessitates a huge uptick in referee interventions, which means less practice counter-pressing and developing good habits in transition (as well as redirection the coach’s focus from helping the players improve to managing the activity). A common sight under this constraint is a player shielding the ball from his immediate marker so a teammate can swoop in and take over possession. This occasionally happens in normal play, but nowhere near as much as a two-touch mandate might suggest.

Quantifying Constraints – Imperfect Rules for an Imperfect World

Here is a graph quantifying decision-making. The general notion here is that intelligence is good, tempo is good, slap constraints on both (task rewards/punishments) to get the best of both worlds.

Art versus Science: Part of what makes limited touches a poor constraint is that it is being used to measure speed of play, ignoring the fact that you can take four or five touches very quickly or two touches very slowly. Furthermore, we are judging the quality of a player’s decision based on whether possession was lost within the next two passes (an attempt to give the player responsibility for not just the technical quality of the pass, but also the tactical quality of the pass). These units of measurement are chosen for their accessibility and clarity rather than for their perfect alignment with what we trying to judge. It is much easier to count to two in units of one, than to try to judge if a player took 1.9 or 2.1 seconds to release the ball.

We use these forced approximations and other simple rules to try to introduce some degree of objectivity into a complex game with 23 simultaneously moving parts where objective truth can be hard to come by. The presence of a mutually agreed upon record of what happened allows players to be held accountable for their decisions and to use what happened in the past to reinterpret similar situations in the future and make better decisions (i.e. learn). This pursuit is valuable enough that we will accept the odd false positive (or false negative) knowing that the trend line will still have value. The coach’s responsibility is then twofold: to design an activity that minimizes the incidence of false positives and false negatives as well as to recognize and counteract the positive or negative influence of these instances with contrasting feedback.

Based on what the two-touch mandate values, you could be forgiven for valuing the quadrants accordingly:

Yet, the reality should look like this:

These hierarchies agree on the value of production in quadrants 2 and 4, but not in 1 and 3, so let us consider what the bottom-left and top-right (those) quadrants might look like. In the top-right, we have extra touches coupled with high intelligence. Fortunately, this concept has already been codified in the Spanish-speaking world, where it is called la pausa. I really love this Busquets example because it shows him passing up a ‘good’ two-touch play (the short lateral pass) for a better three-touch play (the penetrative seam pass).

As I mentioned before, the lack of directionality to three-team possession makes it harder to properly reward these specific types of passes, whereas a directional game can implicitly reward this allowing the team to pursue the subsequent part of the attacking mission (get further up the field and create shooting opportunities).

In the bottom-left you have players that play quickly but without making intelligent decisions. This forces us to ask, what does intelligence look like on a soccer field? I am not going to pretend to be some genius who can objectively define something this complex, but I do not believe it takes a genius to say that intelligent decision-making requires the consideration of all possibilities. In execution this means awareness of all possibilities AND the courage to pursue all possibilities. While it is possible to choose poorly with total awareness, for the young players who can benefit the most from three team possession the primary cause of low intelligence play is an ignorance of alternatives, which brings us back to scanning/multi-object tracking.

So, what does it look like when a player has underdeveloped information-gathering techniques? In my experience, this player will often return the ball from where it came (directionally speaking in general but often directly to the origin of the pass) or put their back to the edge of the playing field and refuse the engage with the center of the pitch. This allows them to fulfill the requirement of two-touch play with the lowest degree of difficulty available to them (also called the path of least resistance/principle of least effort). This is often an ego thing (fear or making a mistake/looking bad in a group setting), so while we could try to coax players to attempt the more difficult choice, it is often simpler and less time-consuming to rig the game to force the issue. If courage is a matter of choice, then remove choice from the table. In the absence of alternatives, the extraordinary becomes ordinary and what is necessary becomes normal.

Examples of Constraints

The two-touch mandate is a blanket punishment on any use of more than two touches while the conditional scoring comes from a possession chain of X passes.

What we need to understand in exploring constraints is what behaviors need to be punished immediately, what can simply be ignored (neither punished nor rewarded) and what should be rewarded and ensuring that there is an alignment between a constraint’s intended effect and its actual outcome while accounting for externalities. There are those that would argue that nothing should ever disallowed (within the laws of the game) and we should only use incentives. This gives the players complete freedom with a carrot to strive for, but I have found a corrective pressure from the other end of the spectrum is often useful alongside it.  

To return to two-touch, an example of using incentives would be only counting individual possessions toward the advancement of a possession chain when the player uses two or less touches, while potentially adding a punishment in the form of rolling back progress for excessive touches: starting with the fifth touch every touch that is not a pass reduces the possession chain by one. This creates a gray space where more touches are acceptable when necessary, but also makes it clear that moving the ball quickly has value and too many touches is very much something that exists.

Here are some further examples of things I have either seen or thought of using the 5W framework:

  • Who
    • A possession chain only completes where every single member of the attacking two teams has completed a pass.
      • This can be combined with other restraints (Where or What) to create positive peer pressure and incentivize teams to incorporate everyone, though depending on the emotional maturity of the team you need to be wary of creating and “us against them” situation within a given team.
    • Players are only rewarded (or are rewarded further) for passing to their team (or the other allied team) within the two attacking teams (e.g. blue passing to blue or blue passing to yellow when red is defending).
      • The intent here is to emphasize scanning and reward players for discerning between good passes and better passes. The specific instances (what is good and what is better) do not necessarily translate to the game because of a lack of Where, but the general concept does.
      • Put another way, this constraint evolves the activity from one where advancement of the chain is the only goal to one where advancing the chain is the bare minimum, with advancement in specific ways acting as the new North Star (analogous to creating opportunities to play forward).
  • What
    • Players are given a point for completing X passes in a row.
      • This constraint is the fundamental building block of this activity because it is what forces the defending team to pressure. The whole point of possession is to continually transfer it forward to get closer to the goal. Without having penetrative potential to force defensive action, we need to introduce a constraint that forces the defending team to pressure and allows the attacking team to create and take space.
    • Players earn a point for completing a give and go.
    • Players only earn points for first-time passes.
      • When played like this, you can do away with possession chains just play to a number as two players passing the ball back and forth has the same effect of forcing pressure.
    • In contrast to the two-touches or less mandate, there is “three-touch must” – a player must take at least three touches before releasing the ball. This has its own set of issues, but I like to use it occasionally to force players to be very deliberate with their first touch, dribble and practice ball manipulation skills.
  • When
    • Given one of the most common criticisms of this is a lack of directionality, we can introduce that here. Stick goals at either end of the playing space and then after completing X passes the possessing team(s) attack the goal further away from them. The defending player closest to the goal becomes the goalkeeper (or not, the choice is yours). This can be done with pop-up goals as wells, emphasizing width via a six-goal game, whatever benefits your players.
  • Where
    • The other common “problem” is players making a ring around the edge of the space and not interacting with the center, so we can create incentives for them to do that. Something I have used is a central square with points for playing into and out of the square, with more points depending on the relationship of the sides the ball enters and exits the square from (same = 0, contiguous = 1, opposite = 2). This creates an incentive not just to use the central space, but to use it access the opposite side of the field and keep the ball moving toward what would nominally be considered forward.
      • I have considered doing this with other shapes as a well. A triangle seems the simplest (play out any face other than the one the ball entered for a point), with the opportunity to add additional incentives (+1 for one-touch play) to get players to look for opportunities to play first time around the corner.
      • I have also seen coaches use mini-spaces like the pips on the 5 face of a die (corners plus center) that teams must use to score (only passes into pips are worth points).
        • This was often combined with a When element – no hanging out in the pips (akin to basketball’s paint), thus forcing players to time their runs into space. This was mandated by the size of the pips, but it is probably an important consideration for any incentivization of central space, lest one player hog it or lazily occupy it and only it.
      • At its most basic, incentivizing the middle is a shallow attempt to introduce some level of directionality to the game, which is worth pursuing if you want to use this kind of activity for more advanced players.
    • Players are rewarded for completing particular types of passes based on origin and destination (say from one edge to the other – big switches of play).
    • One thing I have taken to when I have the equipment is just putting random stuff in the way – mannequins and pop-up goals being the most common – which forces both passer and receiver to change how they play in response to these conditions. A mannequin creates opportunities for the receiver to move out of its cover-shadow or for the passer to lift the ball over it or bend the ball around it, both of which are skills that would be used less without the obstacle being present.
      • I have combined this with the previous constraints to encourage large, lofted balls over goals placed in the middle of the field.
  • How
    • This is a bit different, but I have seen three team possession used for handball with two different aims. The first is to eliminate the technical burden of controlling the ball, that freeing up more cognitive capacity to think about positioning. I am not a huge fan of this, but I think it can have value if used sparingly.
    • The second is to use handball to incorporate more heading and aerial control elements. Because it usually does not make sense the lift the ball unless it is the only way through, these types of repetitions can be difficult to get in bulk, but by playing handball you guarantee the ball is never on the floor and create opportunities for players to practice these touches. As it is handball, players can choose to catch the ball, but incentives are created for legal soccer touches before catching the ball (chest control, thigh control, header to self/other). This can extend to incorporating volleys as well rather than thrown passes (control and volley), but the shared height of heads and feet is something to be cautious of.

Closing Thoughts

This is my first opinion-oriented article and I would love to get feedback from you if you’ve stuck with me to this point. Some specific things I would be interested in hearing about are different constraints and variations you have used or seen in the past, thoughts on some of my constraints and thoughts on three team possession in general.

I’d also appreciate any direction in terms of whether or not this is too long and what things you’d be interested in seeing me write about in the future.

Sheffield United vs Tottenham Hotspur – 17.1.2021

Some quick thoughts on Sheffield United vs Tottenham

  • Tottenham
    • Rodon looked allergic to passing forward. This will be compounded by Aurier’s inconsistency and Mourinho’s obviously distaste for using him in build up. I also wonder about the impact of Bergwijn in this. Coming from Swansea and having played under Graham Potter and Steve Cooper I would have expected a bit more from Rodon. Did will defensively though.
    • Interesting to see Ndombele as part of the double pivot in contrast to his earlier involvements under the striker. He likes to invite pressure so you can see why Mourinho was reluctant to trust him in possession near Tottenham’s back four, but he was given responsibility today in a deeper role and had a good game capped off by a fantastic goal.
    • I missed the first 15 or so, but while Tottenham looked dangerous in transition they did not create much from sustained attacks. Sheffield United are clearly not in a great place, but they generally did enough defensively to keep Tottenham at bay.
  • Sheffield United
    • Is Henderson to Ramsdale the worst positional downgrade in the league?
      • This is a thought I have had for some time and I am mostly interested in seeing what other takes there are. I am not a goalkeeper guy but I can’t help but look at a lot of the goals Sheffield United concede and feel he consistently could have done different or better. I would expect that to be the case for goalkeepers with most goals scored everywhere, but I think he consistently bears a larger portion of the blame than a good keeper should.
        • I think this is potentially compounded by the habits formed playing for a side like Bournemouth versus the habits required to play for a team like Sheffield United.
      • Today Ramsale begins to come for Aurier’s goal before retreating, whereas had he chosen to stay his positioning and balance would have been better prepared for the shot. He is also at the six for Kane’s goal whereas had he been on the line/two yards out I think he makes or should make this save. He is recovering after playing short and trusting his team so you can see why he was where he was, but on the balance of trust in teammates versus considering “what if they mess up?” he needs to err on the side that consistently reduces the team’s goals against. Given the jumbled mess in front of him that means more wisely worried and less trust, regardless of how egregious Norwood’s mistake is. 
      • I also wondered about his percentage of balls played long vs Henderson but looking at the data there does not seem to be too much difference (T-tests might say otherwise but my curiosity is sated). The numbers for PSxG/SoT suggest Sheffield United are not worse defensively than last year at 0.30 holding fast (other metrics are worse but this is consistent), but the difference in goalkeeper performance for PSxG +/- is quite large and likely to get worse if Ramsdale continues to perform at his current level.
    • Sheffield’s attacking limitations in wide areas
      • There was a piece in the Athletic a little while ago about the importance of left-footed center backs. I cannot be bothered to pay for it but it’s here if you care to or are already subscribed. Beyond pass trajectories and whatever else that article delves into, the ability to use the outside foot is important because it allows a player to protect the ball and play forward into wide areas where there are generally less opponents and losses of possession are more forgiving (further distance to goal, more players between point of concession and goal. Many center backs will pass their troubles onto their fullbacks rather than attempting to play forward themselves, but this issue is magnified when a player is on the “wrong” side. This further reinforces what a loss Jack O’Connell is for this team. He is a capable defender, left-footed, contributes to the attack in wide areas and was their primary set piece target when fit. All of this preamble is to say that Ethan Ampadu’s relative discomfort/limitations affected Sheffield United’s ability to play down the left-hand side, especially given the outside center back’s initial responsibility as the base of their flank play. On the other side, I have not seen much of Jayden Bogle before today, but he made a lot of interior runs and seems more focused on offering a presence centrally rather than in wide areas, which meant that both of Sheffield United’s flanks were limited going forward by the tendencies or the qualities of the players within them. While Fleck delivered a great cross for McGoldrick’s goal, neither Lundstrom or Basham contributed significantly down the right.
    • Norwood’s nightmare
      • Last season Norwood often filled the role of the floater who linked Sheffield’s left and right flanks. He often receives the ball with little support but lots of space and is responsible for choosing whether to continue to probe the current side or switch the point of attack. He was caught out multiple times today due to lack of awareness and lack of quality and was a consistent source of Tottenham transition opportunities. It was hardly surprising Ampadu was pushed forward into this role as it theoretically solved two problems for Sheffield United (Ampadu’s suitability as the base of left-sided attacks, Norwood’s struggles).
    • Fluidity as a double-edged sword
      • Attacking fluidity is great in theory because you should get into better positions as no time is lost trying to get the right people in the right areas. You can get people into dangerous areas from multiple adjacent areas and use this movement to pull teams apart and create space. In practice, at least for Sheffield United today, you wind up with players in areas of the pitch who do not have attacking identities or qualities tied to those areas.
      • If the result of this fluidity is that McGoldrick is wide left, Enda Stevens is in the half-space and Fleck is attacking the box, that arrangement is suboptimal given the different qualities of these players. What you are hoping for is that the increased space afford to the players from the fluidity makes up for the difference in technical/tactical quality, which generally has not been and was not the case for Sheffield United today.
      • Contrast this with the goal, where Enda Stevens is attacking the flank, Fleck is supporting him and facing the block and both forwards are in the box.

Match Analysis

Manchester United vs Southampton (13.7.2020)

Fair Use Disclaimer: This post uses copyrighted material from NBC Universal Media LLC which has been used without specific authorization. Under Section 107 of the Copyright Act 1976, allowance is made for “fair use” for purposes such as criticism, comment, news reporting, teaching, scholarship, and research. Fair use is a use permitted by copyright statute that might otherwise be infringing.

If you wish to use copyrighted material from this site for purposes that go beyond fair use , you must obtain permission to use the work from the original copyright holder.


Hosts Manchester United have shot out of the blocks in Project Restart, winning five of their last six matches (one of those in the FA Cup) and finally settling into a consistent starting line up with the interregnum allowing star midfielder Paul Pogba to recover from his injury woes. They have started the same eleven in each of their last four Premiership matches. Visitors Southampton are also in good form with a record of 3-1-1 since the resumption of the league, including a hard fought 1 – 0 win against incumbent champions Manchester City in which they rode their luck and the fine form of keep Alex McCarthy to an uneasy victory. Tonight’s game could be similarly touch and go for the Saints with the red side of Manchester scoring three or more in each of the games this settled first team has started. Southampton will look to use their aggressive pressing to unsettle a Manchester United team that has yet to be really contested in their own half since the restart.

Starting XIs

First Half Analysis

From the outset it was clear Southampton were not overawed by the occasion, and several clear patterns emerged within the opening 10 minutes of the game. First, United had a clear preference for playing out from the back to their left-hand side, with Nemanja Matic dropping left of Harry Maguire to create a back three with Luke Shaw and Aaron Wan-Bissaka pushing on. When Southampton successfully prevented the ball from being playing forward in wide areas (with Armstrong and Redmond) and inside (with Ward-Prowse or Romeu pushing on), United played backward and laterally away from the continued pressure. From there Lindelöf was forced to play long in the absence of other options, with Greenwood, Fernandes and Wan-Bissaka marked tightly. Second, Southampton had a clear intent to keep their front four relatively tight while in possession to allow for support and combination play, with belated runs from the fullbacks to create avenues to progress the ball in wide areas. This overload caused Matic and Pogba to be more worried about the passes they were screening than getting out to contest switches of play from in front of them.

Redmond and Armstrong are highlighted in white here, tucking inside an creating opportunities for quick combinations if they can be found. Pogba, highlighted in black, is preoccupied with what is behind him and does not pressure Romeu (red), allowing him an uncontested switch of play to Walker-Peters on the right flank.

Southampton were generally sharp in possession and moved the ball quickly on to the next available pass, but the game might have been different had Martial capitalized on a slack first touch from Ward-Prowse. The Southampton set piece specialist had dropped in between the Saints center backs and his errant touch allowed Martial to rob him and charge toward a retreating Alex McCarthy, who stayed big and spared his captain’s blushes.

Here we see Luke Shaw (far touchline) about to receive a pass from Matic. The two black arrows indicate the ideal support positions for Pogba (left) and Rashford (right), with the white arrows indicating how United might combine were Shaw able to play first time around the corner to Rashford. Also highlighted here at far right is Fernandes, who nearly always looked to stay near the striker and behind both Romeu and Ward-Prowse. By preventing Pogba from playing forward, Southampton simultaneously denied Fernandes the immense influence he has enjoyed since coming to Old Trafford.

United were finding relatively little joy in terms of playing forward thanks to continuous Southampton pressure. The first goal came from this pressure, with Lindelöf being drawn to the left side to cover for Harry Maguire. With Lindelöf passing the ball back to De Gea from United’s left hand side, De Gea’s normal short circulation pass was not available, and so he played it up the gut to Pogba. Pogba had Lindelöf underneath at this point as an easy set but chose to turn at the top of his own penalty area and lost out to Danny Ings. Nathan Redmond recovered the loose ball from Pogba’s resulting foul and broke into United’s penalty area before crossing for Stuart Armstrong to finish at the far post.

With this goal United appeared to make a bit of an adjustment, with Luke Shaw and Wan-Bissaka staying deeper and available for uncontested rotational (rather than penetrative) passes from the United center backs and Martial switching positions with Rashford. This immediately yielded success in beating the press, with Shaw able to turn, play accurately forward to Martial and receiving his striker’s return pass beyond the initial Southampton press. From there Shaw played an early cross to Rashford in behind, but his smart finish was disallowed on account of mistiming his run and straying offside.

Southampton continued to pressure United, with Pogba being dispossessed just inside the Southampton half but the counterattack petering out after an overhit through ball. United’s circulation continued to look improved and created opportunities for them to play forward in wide areas on both sides of the field. This advanced possession allowed Pogba to receive facing forward in position to put the ball into Southampton’s penalty area. Martial’s initial success in taking the ball down and away from Jack Stephens drew in Kyle Walker-Peters, creating enough space for Manchester United’s number 9 to slip in Rashford for his 16th Premier League goal of the season.

The safer passes to the fullbacks again counted for United’s second, with Wan-Bissaka able to find Pogba in a central pocket against Southampton’s press. With Pogba free to play forward he found Fernandes behind Southampton’s midfield, who passed the ball on for Martial to have a run at Kyle Walker-Peters. The fullback lost contain and allowed Martial to come inside. Jack Stephens was slow getting out and Martial finished powerfully toward the near post, putting Manchester United in front going into the water break.

United seemed settled down from seeing the results of their adjustments, and Southampton seemed a bit hesitant to commit their central midfielders forward in the attacking half, resulting in more possession for United in the attacking half. Southampton clogged the middle and attempted to squeeze United on the flanks as the looked to make inroads there, generally limiting United to possession around the penalty area without ever really allowing them inside it. United generally counterpressed well in these situations when they did lose the ball, forcing Southampton to play early long balls that were uncontested recoveries or favorable duels for the United defenders. The switch from Martial and Rashford also appeared to be more of a spontaneous thing than a strict switch, with Rashford appearing more prominently down United’s left.

The most noteworthy moment after the game had settled was a late challenge from Oriol Romeu on Mason Greenwood which was acknowledged by the referee in the form of advantage played but went unsanctioned in the referee’s notebook, when it was at very least a yellow and could have been worse given the ball was very much gone was Romeu raked his studs down Greenwood’s Achilles.

Second Half

Southampton started the second half spritely, again getting at United in wide areas. They had the first chance of the period with Shaw twice denying Walker-Peters but unable to move the ball on before Stuart Armstrong was able to double-down on him and abscond with the ball. Adams and Ings both threatened runs in behind, driving United’s defense deeper and creating a pocket for Nathan Redmond to receive in the left half-space with a chance to run at defenders. Initially there was no further wide threat so Wan-Bissaka squared up to Redmond, but as Che Adams overlapped from his striker position Lindelöf sent his fullback to deal with that threat and stepped out. Redmond was quick to pull the trigger as United attempted to transfer the responsibility of pressure, ultimately pulling his shot wide of the near post.

Southampton again exploited Pogba’s hesitance to pressure for the second shot of the second half, with Armstrong running off Matic onto a chipped ball into the penalty area from Romeu. The Scot laid it back to Golden Boot hopeful Danny Ings who sold Victor Lindelof on an attempted shot before having the real effort blocked by Nemanja Matic. There was a hopeful penalty shout from Southampton when Armstrong backed into Maguire as United looked to scramble the ball clear, but the referee gave nothing and the Video Assistant Referee was similarly unmoved.

The game continued in stop-start fashion with lots of fouls, throw-ins and pressure forcing both teams into errors and playing for territory rather than risk turnovers in their own half. There was also a lack of sharpness to both attacks in the final third, perhaps speaking to the congested fixture list since the restart beginning to take its’ toll. There were a few half moments before both managers looked to affect the game with their substitutions, namely a long distance Rashford snapshot after a handball from Kyle Walker-Peters prevented a more penetrative pass and a Southampton counter after Rashford saved a mishit pass from Bruno Fernandes. Rashford’s momentum carried him out of the pitch and Southampton eventually worked the ball into the area he would have been screening and found an easy forward pass to Kyle Walker-Peters. The fullback’s low driven cross nutmegged both Luke Shaw and Harry Maguire but evaded Ćhe Adams at the far post.

The aforementioned substitutions, simultaneously made on 64 minutes, spoke to the competing objectives of the two managers. For United it was Fred on for Paul Pogba, who had provided a moment of quality for United’s first goal but was also responsible for Southampton’s opener, while Southampton brought on Shane Long for Ćhe Adams. The priorities of Manchester United were also clearly demonstrated by De Gea waving his defense upward on ensuing goal kick rather than risking playing out.

United had two more almost moments as the game continued with chances derived from the individual qualities of their forward players. First, Rashford ghosted Walker-Peters with a reverse elastic before finding Martial making a run in behind from in between the two center backs. Rashford continued his run into the area and was found by Martial’s cutback inside the six, only for his first-time effort to be stuffed point-blank by the covering Ryan Bertrand. Second, after the drinks break and substitutions by both sides (Will Smallbone on for Stuart Armstrong, Brandon Williams on for Luke Shaw, who appeared to twist an ankle), Martial blocked Stephens’ forward past and was first to the rebound inside the United half. Martial ran away from Stephens, around Bednarek who had slipped in trying to offer cover, and all the way to the Southampton penalty area, where Stephens had recovered just enough to disrupt the Frenchman’s shot.

Solskjaer’s intent to protect what United had achieved up to the point was cemented by his substitutions in the 84th minute, with Daniel James coming on for Mason Greenwood and Scott McTominay coming on for Bruno Fernandes. Neither Fernandes nor Greenwood had been particularly influential up to that point. Greenwood was a victim of United’s greater quality in building down the left and Southampton’s good low block defending not allowing United to switch the point of the attack in the attacking half. Fernandes was also starved by Southampton’s pressure, with the driven, line-breaking passes toward him from Matic and Pogba that were crucial parts of United’s attack in prior games being denied by Southampton’s central superiority and active defending. With this pair of changes Fred moved into the #10 role with McTominay alongside Matic in the double pivot.

While the Red Devils did a good job of positioning and being first to crosses and long balls, their unconvincing clearances from those wins allowed Southampton to maintain possession in the attacking half and to continue to probe United for weakness. They found it down their left, with Bertrand drawing Wan-Bissaka out before laying the ball back to Romeu to find Redmond on the left side of United’s penalty area. Redmond duly took on the covering Matic and drew an excellent save from De Gea after driving into the box.

With United having shown their hand by using all three substitution intervals, Hassenhütl withdrew central midfielder Oriol Romeu for striker Michael Obafemi, seemingly to play as a right forward with Smallbone moving centrally alongside Ward-Prowse. The game was hardly settled long enough to get a good look at the Austrian’s intent before an aerial duel between Walker-Peters and Brandon Williams resulted in a head-to-head collision and Williams being unable to continue. Solskjær used the lengthy injury stoppage to reconfigure his team, directing them into a 5-3-1 with Matic and Fred on the left side of the back five. This narrow midfield gave United limited ability to screen passes into wide areas while also protecting the middle. Southampton took the space given to them and Redmond earned what would prove a pivotal corner off of Wan-Bissaka in the sixth minute of added time.

United defended with six zonally inside the six against four Southampton attackers and were man-to-man with the three runners Southampton had around the penalty. In both this instance and an earlier corner in the 87th minute following De Gea’s save on Redmond, United seemed willing to allow Southampton to have free player in the box and prioritized the numbers they had in the six. On both these corners the two players near De Gea made the same runs, and on this second occasion Ward-Prowse’s choice or execution was better and Southampton capitalized. Bednarek came back toward the six in line with the near post to win the flick on, while Obafemi backed into Lindelöf, obstructing his movement toward the ball. Lindelöf was able to get goal side of the striker but not ball side, and as a result Obafemi was first to the flick-on and stabbed the ball home from close range to give Southampton a share of the spoils.