Three shapes, one idea. I’ve spent a bit of time this week trying to formalise how I want the team to behave in different game states without falling into the usual trap of having one tactic for normal play, another for chasing and then something completely different for seeing a game out.

The important bit for me is that I don’t want the ideology to change with the score line. If the base idea is positional play, short connections, central progression and manipulating space, then going 1–0 up should not suddenly turn us into a low-block side clearing everything into the stands. Likewise, going 1–0 down should not mean abandoning structure and throwing bodies forward at random. So I’ve ended up with three IP shapes that all come from the same basic principles.

The normal shape is still the familiar 3-4-2-1, which becomes the 3-2-5 through role movement. Bombino steps inside, the right wing-back goes high, Peralta moves into the half-space, the box-crasher advances and the front five forms naturally. That remains the default because I like the idea that the shape is created by movement rather than just placing everybody in their final positions from kick-off.

The attacking version is basically that same finished picture, but from the start. Rather than waiting for the IWB, wing-back and attacking midfield roles to create the 3-2-5, the players are already positioned much closer to it. It is less a different system and more the normal system with the rotations accelerated. The back three and midfield protection stay intact, but the attacking occupation is immediate and the mentality can move from Balanced to Positive without changing the entire way we play.

The control shape was the one I had to think about more. My first instinct was still a little too conventional: keep the back three, make the midfield more cautious, reduce movement and protect the lead. It worked structurally, but I wasn’t completely happy with it because it still felt like the team was becoming more passive simply because we were ahead. The version I’ve settled on is much cleaner. Both wing-backs become inverted, so in possession the side effectively becomes a 3-3-3-1. The back three remain underneath the ball, Müller sits centrally, Bombino and the right-sided IWB step inside either side of him, then Rotondi, Lohan and Peralta occupy the next line behind Salvo. The idea is that defending a lead becomes about controlling the ball and controlling the centre, rather than dropping deeper. There should always be close passing options around Müller, the back three have a much stronger rest-defence platform in front of them, and the three attacking midfielders still give us players between the lines if the opposition start overcommitting. Salvo stays high enough to keep the centre-backs honest as well. The trade-off is width, obviously. With both wing-backs moving inside and the attacking midfield line narrower, we lose some of the natural outside threat. But in a game we are trying to control, I’m happy to accept that in exchange for better central security and more passing connections.

The other thing I like is that none of the three shapes requires a change in philosophy. The base shape uses movement to create the 3-2-5. The attacking shape starts closer to the 3-2-5. The control shape becomes a 3-3-3-1 and uses possession as part of the defending. So instead of having three separate tactics, it feels more like three levels of risk within the same football. That is probably the bit I care about most. I don’t want the team to stop looking like San Diego just because the game state changes. I want the structure to change, the level of risk to change and the roles to change, but the principles underneath it all should stay recognisable.

Out of possession, we adapt to opponents; in possession, we adapt to game state.

A couple of academy promotions to round out the squad, and both have an interesting connection: Richard Duah and Akwasi Kwarteng originally came through the Right to Dream Academy pathway before joining San Diego’s system.

Duah, 17, comes in as Paulsen’s No.2. He is still raw, particularly aerially – despite his 6’6″ frame – and with the ball at his feet, but there is enough there to develop. More importantly, he comes onto a Reserve contract with a $0 cap charge, which makes far more sense than spending meaningful cap space on an experienced backup who may barely play. Kwarteng is only 16 and will start as the developmental third striker behind Salvo and Ríos. McGuire’s poor pre-season probably opened the door, but the promotion is about upside rather than necessity. Kwarteng has pace, dribbling, a decent first touch and enough finishing ability to make him worth exposing to senior training and selected minutes. He also has a clear target: five league goals in 2031 will move him from Breakthrough Prospect to Fringe Player for 2032.

The Right to Dream connection makes both promotions more interesting. That pathway was already something I wanted to explore as part of the longer-term recruitment model, so finding two players from it already inside the academy is a nice reminder to look inward as well as outward.

Neither changes the first XI, but both are cheap, cap-free and have genuine development value. That feels like exactly how the final roster spots should be used.

Poor trainers: how much should it actually matter?

One thing I’ve been trying to get a handle on is how much weight to give training ratings when they clash with what I’m seeing on matchday.

A few players have been poor in training recently and have had several warnings. Today, I wanted to explore a plan to set them training related targets. Simpson-Pusey and Luque both accepted the criticism but rejected formal targets, while Salvo reacted badly to the conversation altogether. That raised the obvious question: what do you actually do with a player who trains badly but still performs?

I don’t think the answer is to treat poor training as proof that the player is bad. It is information about development, consistency and personality, but it is not the same thing as match performance.

So I’ve settled on a fairly simple hierarchy.

Match performance comes first. If Salvo is scoring and doing the job I need from the role, he plays. I’m not going to drop him because he posted a 5.9 in training unless that starts to become part of a wider issue.

Training becomes important when it turns into a pattern. One poor week is noise. Repeated poor ratings over a month are much more useful evidence, particularly for younger players who are supposed to be improving.

Age matters too. A 23-year-old training badly concerns me more than a 31-year-old doing the same thing. With someone like Culbreath or Simpson-Pusey, I want the training environment to help develop them. With Luque, who is older and has already performed well in matches, training is much more about maintenance.

The response to criticism matters as well. If a player repeatedly trains badly, refuses targets, reacts poorly to criticism and then stops developing, that starts to become part of the long-term recruitment decision. But I don’t think it needs to become an immediate selection issue unless it starts affecting performances or the wider dressing-room standards. I also don’t want to start changing the whole team schedule because three players are struggling with it. The training programme has been built around the football we want to play, so the first question is whether the schedule is working for the squad as a whole. Individual load can be adjusted if fatigue or intensity is the problem, but I’m not rebuilding the week around one bad trainer.

The approach I’m going to use is:

poor training triggers a conversation; persistent poor training affects development and long-term squad planning; match performance still determines selection unless the problem becomes chronic enough to affect the team.

That feels more sensible than turning every 5.9 training rating into a disciplinary crisis.