The first proper complication of the winter has arrived, and it is a fairly significant one. I have spent the opening weeks thinking about where the squad needs improving, identifying targets and starting to put some deals together, only to discover that, at the moment, I cannot actually register any ordinary new signings.
The problem is the salary cap. San Diego are currently sitting at around $7.26m against a maximum of $7.068m, so the squad is already over the limit before I add anybody else. The club itself is extremely healthy financially, with more than $60m in the bank and plenty of room in the ordinary wage budget, but none of that matters if the MLS roster cannot be made compliant. It is a useful reminder that having money in this league and having room to spend it are two very different things.
Tahiel Peralta is unaffected because he is arriving as a Designated Player, which gives us a different route through the cap rules. Beyond that, though, everything is currently on hold. António Gilberto has not completed his move yet despite the deal being agreed in principle, and the proposed move for Juan José Rosa has run straight into the same problem. Until I create some salary-budget space, those sorts of signings simply cannot be pushed through.
The more I look at the squad, the more I think this explains some of what went wrong last season. San Diego do not have a particularly small group. Quite the opposite. There are plenty of contracted players, but too much of the cap appears to have been spread across players who are useful without necessarily being important enough to justify what they cost against the budget. A $250k or $350k cap charge does not look disastrous in isolation, but collect enough of them and suddenly there is very little room left to improve the starting eleven.
That leaves you with the worst of both worlds: a large squad without particularly strong depth. There are enough bodies to make the roster look full, but not enough difference-makers to explain why so much of the available cap has already disappeared. It is possible to be rich as a football club and still have a badly constructed MLS roster, and I am beginning to suspect that is what I have inherited.
That does not change my view of the 2030 season entirely. The underlying performance still suggests San Diego were unfortunate in certain areas, particularly scoring only 28 times from 45.12 xG. There was enough chance creation there to believe results could have been better. But bad luck on the pitch and inefficient squad construction can exist at the same time. The more expensive middle of this roster may have left the club without the flexibility or depth needed when the established players underperformed.
So the next stage of the rebuild is going to be less glamorous than finding another Peralta. I need to go through the squad properly, work out which players are genuinely worth their cap charge and identify who can be traded or moved on. In MLS, getting a player off the books can be every bit as important as signing the replacement, particularly if a trade can also bring General Allocation Money back in the other direction.
I came into the winter thinking the task was probably three first-team signings and then some depth. It may be bigger than that. Before I can improve the squad, I first need to create the space to do it.
The money is there. The roster room isn’t.
The salary-cap problem has turned out to be less of a dead end and more of an MLS bookkeeping lesson. Once I finally found the right screen, the club’s General Allocation Money became usable to reduce the cap impact of players already on the roster. San Diego had just over $6m in GAM available, so rather than immediately forcing players out for poor returns, I have used some of that flexibility on the players I am certain will be central to the team.
The first three were Philipp Müller, Jahmai Simpson-Pusey and Dylan Salvo. Müller’s cap impact has been bought down from $879k to $150k, Simpson-Pusey from $589k to $150k, and Salvo from $421k to $150k. That has cost a little over $1.4m in GAM, but it has created roughly the same amount of salary-cap space without weakening the intended starting eleven. All three remain fully paid on their normal contracts; this is purely an MLS accounting mechanism for the 2031 season, with their cap charges resetting next year.
That distinction is important. Müller is the captain and the deep-lying playmaker the midfield is being built around. Simpson-Pusey is the ball-playing centre-back on the right of the back three, and Salvo is going into preseason with the opportunity to keep the centre-forward position. These are exactly the players I am happy to spend allocation money on because the issue is not that their contracts are bad. The issue is that their cap charges were preventing me from building around them.
The result is that the projected salary-cap figure has dropped dramatically, leaving more than $1m of immediate room before the next arrivals are completed. That should be enough to revisit the deals for António Gilberto and Juan José Rosa when the game allows negotiations to reopen. Gilberto would arrive through the U22 Initiative at a heavily reduced charge, while Rosa could potentially sign as a Senior player and then have his own cap impact bought down once he is actually under contract.
The next part of the job is different. I do not want to use GAM to make every inherited contract more palatable. There are still several players whose cap hit is too large for the role I expect them to play, and those are the ones I have started offering around MLS. Lucas Ríos, Joseph Rosales and Anthony Markanich are the clearest examples at the moment, while there are other players I would listen to offers for if the trade value makes sense.
The priority in those negotiations is not necessarily cash. General Allocation Money is useful, draft assets can have value, and simply removing an inefficient cap charge can be worthwhile in itself. The important thing is that I do not replace a $400k reserve with another $400k reserve and end up exactly where I started.
That means the next phase of recruitment has to be much more disciplined. If an expensive squad player leaves, the replacement does not necessarily need to be of equal financial status. I need smaller Senior contracts, Senior Minimum Salary deals and off-budget options wherever possible, particularly for depth positions. Backup goalkeeper, reserve centre-back, another defensive midfielder and cover on the right wing should not be consuming the same sort of cap space as starters.
The model is starting to become clearer. Use GAM to protect the expensive players who genuinely matter. Trade the expensive players who do not. Then refill the gaps with contracts appropriate to the role rather than the reputation of the player.
For the first time since discovering we were over the cap, there is some room to manoeuvre again. The next challenge is making sure I do not immediately spend it all in the same way the previous roster seems to have done.
One of the quieter jobs this winter has been putting some structure around individual training. I have always tended to use additional focus fairly casually in Football Manager, usually spotting a weak attribute and assigning something that looked sensible, but with this squad I wanted something a little more deliberate.
The starting point was to define what each player is actually being asked to do in the system and then work backwards from there. Rather than trying to make everyone more complete, I have set up a development plan around the specific attributes that could stop them performing their role properly. Each player now has up to three additional-focus targets, with a threshold attached to each one. Once the first target is reached, the plan moves on to the next rather than leaving someone training the same area indefinitely.
There are some fairly obvious examples. Sergio Uribe is working on defensive positioning because he is going to be the central reference point of the back three but, whilst undertaking a leadership course, cannot add any additional foci to his daily training. Jahmai Simpson-Pusey, as the ball-playing centre-back on the right, is focusing first on his passing and will then move onto ball control and positioning. Luca Bombino is being developed more like an additional midfielder than a traditional wing-back because his role will be to move inside alongside Philipp Müller, while Dylan Salvo’s priorities are finishing, first touch and acceleration as he auditions for the starting centre-forward role.
The same principle applies further forward. Müller does not need to become more athletic at 26; I want to sharpen the qualities that make him the deep-lying playmaker, so passing and first touch come first. Rotondi’s work is centred around keeping the technical parts of his game strong rather than trying to reverse the physical decline of a 33-year-old. When Tahiel Peralta arrives, his development will be aimed at making him even more dangerous receiving inside from the right rather than spending time teaching him attributes his role should rarely require.
Traits are being treated in much the same way. I have mapped out a small number for each player rather than throwing new Player Traits at everyone. Some players already have behaviours that fit perfectly, others are currently learning one, and there are a few where the game simply will not allow the trait I wanted. Simpson-Pusey already brings the ball out of defence, for example, while Rotondi is working towards becoming more disciplined about holding the width on the left.
The important thing is that this is not really about squeezing another point out of every attribute. It is about making sure the training reflects the football. If a player reaches the level required for his role, the focus moves on. If an older player stops progressing, I am not going to leave him working on the same target for a year simply because I think he should reach 14.
It is another idea carried over from Miami. The tactic worked there partly because players were judged by what their particular role needed rather than by whether they looked complete on the attributes screen. The aim now is to apply the same logic to development: train the function, not the footballer in the abstract.
The first friendly is done, a 3–2 win over Sporting Kansas City, and there was enough in it to feel encouraged without pretending everything is already in place.
The biggest positive was Philipp Müller. The screenshots during the game already suggested he was sitting in exactly the right areas, but the numbers backed it up: 90% pass completion, 15 progressive passes and five key passes. He looked like the player the whole structure could revolve around, not just circulating possession but actually moving us through the pitch and then appearing again around the final third. The 3-2-5 build-up shape also looked natural, with Bombino coming inside and the right wing-back pushing high, which gave Müller plenty of angles to work with.
That right side is where the biggest weakness was exposed too. Alan Luque covered around 14 kilometres, made eight dribbles and got into some very advanced positions, but the end product was poor. His xA was only 0.06, and his crossing success was low despite getting repeated opportunities. In other words, the tactic is putting the right wing-back exactly where I want him; the issue is whether the current player has enough quality to make those situations count. That probably pushes RWB even higher up the recruitment list.
Dylan Salvo was the other obvious talking point. Three goals from 0.87 xG is a very good start, and his movement was lively, but the pass map showed him dropping a little deeper than I really want from the striker in this version of the system. The point of moving away from chasing another false nine is to have someone who pins the centre-backs and creates space around him, so I am going to test Salvo as a more conventional Centre Forward rather than a Channel Forward and see whether that gives us a cleaner vertical reference.
Out of possession, the structure was mostly encouraging but there were a couple of warning signs. The shape itself looked compact enough, but Müller was probably stepping out to press more aggressively than I want from the player who is supposed to anchor the midfield. I would rather keep the overall press relatively neutral and set the real triggers opponent by opponent. The other concern was defending wide areas and crosses. Kansas City created one goal from a 2v1 on the flank before delivering into the box, and Uribe only won 33% of his aerial duels. That is not enough evidence to change the system, but it is enough to keep wide defensive support and aerial defending on the watchlist.
That becomes even more important once Tahiel Peralta arrives. Vera did a decent job coming inside from the right and combining around Zone 14, but Peralta should give us a much more dangerous version of the same movement. If the RWB can hold the width, Salvo can stay higher and Peralta can receive inside on his left foot, the right side should become one of the main attacking routes in the team.
So the first game did not really change the overall plan, but it sharpened it. Müller looks exactly right. Salvo may be the striker after all, just perhaps in a different role. The right wing-back position clearly needs more quality. Out of possession, the framework looks workable but there are still questions around pressing discipline and defending crosses. And the shape itself looks ready for Peralta to drop into it.
For a first outing against a strong MLS side, that is a pretty useful place to be.
The first real player-for-player trade of the rebuild is done, and the more I look at it, the more I like it.
Nashville came in with interest in Liam Beaumont, who had looked a long way off the level I want in the opening preseason game. Rather than simply asking for picks or GAM, I went through their squad looking for someone who could solve an actual problem for us. That led to Andrew Privett, a 30-year-old defensive midfielder/centre-back, and Nashville accepted the straight swap.
On the face of it, Privett is exactly the kind of depth piece I have been trying to find. He covers two positions that were already on the shopping list, comes in on a slightly smaller contract than Beaumont, and gives us a player who can rotate with Müller or fill in across the back three without forcing me into another expensive Senior deal. At 30, this is not an upside play. It is about function, reliability and making one roster spot do the work of two.
His attributes suggested that already, but the 2030 numbers make the case stronger. In 27 MLS appearances he completed 90% of his passes, averaged 45.96 completed passes per 90, and still managed 5.72 progressive passes per 90, which suggests he is not simply recycling the ball sideways. Defensively he won 90% of his tackles, made 47 interceptions and, perhaps most encouragingly given some of the aerial issues from the Kansas City friendly, won 76% of his headers.
There is some useful security in possession too. Privett averaged 14.93 possessions won per 90 against only 5.52 possessions lost, and he made no errors leading directly to goals. That is exactly the sort of low-variance profile I want from a reserve player. He is not there to dominate matches; he is there to come in, keep the structure intact and avoid creating a new problem. There was a little more progression than I expected as well. He produced 18 key passes, eight chances created and 3.79 xA across the season. Nobody is mistaking him for Müller, but it is enough to suggest he can step into midfield without becoming a complete dead end in possession.
The limitations are obvious. He offers almost nothing as a dribbler or scorer, his athletic ceiling is not especially high, and his Determination is lower than I would normally want in this squad. But again, context matters. I am not buying a starter. I am swapping an expendable attacking squad player for a defensive utility player whose recent numbers suggest he can give us competent minutes in two different roles.
That also potentially removes the need to recruit both a backup centre-back and a backup defensive midfielder separately. One player can now cover both jobs, which means one fewer Senior contract, one fewer roster place and more flexibility elsewhere. More than anything, though, this feels like the first sign that I am beginning to understand how to use the MLS trade market properly. Instead of asking what Nashville would pay for Beaumont, I asked what Nashville had that would make San Diego more efficient.
That feels like a much better way to build this squad.
I think I’ve gone too far down the recruitment rabbit hole since taking the San Diego job.
At Knoxville, I had almost no control. Recruitment wasn’t really mine, tactical freedom was limited, and most of the challenge came from trying to get something out of the squad I’d inherited. Greenville gave me a little more room, but it was still largely about working with what was there and learning what players could actually do before deciding what needed changing. Miami was where I became much more deliberate because by then I had a clearer tactical identity, clearer roles and a better idea of the sort of player I wanted. Recruitment became more structured because the football itself was more structured.
I think I’ve arrived at San Diego and skipped a step.
Because the club is bigger, the budgets are bigger and MLS gives you so many different roster mechanisms to think about, I’ve gone straight into trying to build the finished recruitment model. I’ve been exporting data, building huge player pools, searching obscure nations and leagues for hidden value, then narrowing hundreds of names into increasingly elaborate shortlists. In theory, that sounds sensible. In practice, I suddenly looked at what I was doing and realised I was halfway to planning an almost complete rebuild before I had really coached the squad I’d inherited. That was the panic moment.
All of that has some value, but I’ve also realised I’ve started making judgements on players without really doing the things I said I wanted this save to be about. I haven’t properly used the statistical side yet. I haven’t gone back through my own players’ performances from last season to work out who was actually producing, who was underperforming and who might simply have been badly used. I haven’t really looked at the MLS market for players who are undervalued, blocked or in the wrong situation. Instead I’ve been jumping straight to attributes and deciding what I think the squad needs before I’ve properly understood what I already have.
Salvo is the obvious example. Before the first friendly I was already thinking about whether striker might need to become a DP position. Then he scores a hat-trick and suddenly that looks very different. At the same time, I’d already decided players like Ríos, Rosales and Markanich were potentially up for trade without really judging them on how they train, how they respond to the new system or what kind of impact they might have through pre-season. That feels too quick.
I want to give this group a few months to show me whether they deserve to stay. Some of them probably won’t, and there will still be difficult decisions to make, but I want those decisions to come from actually coaching them rather than from looking at a cap number and an attribute screen.
That is probably the manager I want Villalba to become now. Not someone who arrives at every new club and immediately starts recycling the squad, but someone who improves players, finds better roles for them, gives them time to adapt and only moves them on when the evidence says it isn’t going to work.
So I’m changing the approach a bit.
I’ll still use TransferRoom for specific gaps where I know I need a player, and the scouts will focus more on building knowledge of divisions and markets rather than me manually searching through thousands of players. I’ve started creating some developmental shortlists, but they’ll stay as scouting lists rather than immediate transfer lists. I’m also accepting that the squad is not complete, but it doesn’t need to be overhauled today. The friendlies will give me a better picture of who can do what, who responds to the system and where the real weaknesses are. Once I’ve got that evidence, I can sit down, analyse it properly and decide what actually needs changing.
The second friendly, a 2-0 win over Sacramento, felt much more useful than the first, mainly because I went into it with a proper pre-match plan rather than just loading the tactic and seeing what happened. The idea now is to treat the in-possession structure as the identity of the team and use the out-of-possession side more as game management. Sacramento gave me a good first test of that because their 4-2-3-1 was fairly fluid and their attacking midfielders moved enough to make some of the original matchups difficult to follow. I have beaten too many MLS sides with Miami to take a USL side for granted.
The pre-match plan was built around keeping our attacking structure intact while trying to control where Sacramento could progress. I wanted their weaker defenders to have the ball, while Moustier and Lócio were the players I was more concerned about. Lócio in particular looked dangerous between the lines and Sacramento’s left half space, with Torquato – once of Miami – and Janssen, had enough pace to cause problems if Luque got caught too high, although the former tucked inside.
For the first part of the game, the plan worked only in patches. We were 1-0 up early, but defensively we still looked a little loose and Lócio was getting involved too easily. At one point he got at Viery directly and really should have scored. The issue wasn’t really Viery, though. The bigger problem was that the ball was reaching Lócio too comfortably in the first place; that was where the in-game tweaking became interesting: Moustier had moved across into the left-central midfield position, so I swapped the responsibilities of Müller and Martínez out of possession. The principle stayed the same: one midfielder presses the source and the other protects the space behind. Even with that, Moustier was still finding Lócio too often, so the next adjustment was to move Vera inside out of possession and get him jumping directly onto Moustier – that worked much better.
Instead of asking one of the defensive midfielders to leave the screen and chase the passer, Vera became the first presser while Rosales could stay underneath and protect the lane into Lócio. It made the whole defensive picture cleaner and, in the second half especially, Vera’s OOP role looked really effective. That is probably the most useful thing I took from the game because it gives me another mechanism to use in future matches when an opponent has an obvious deeper playmaker.
The overall performance was decent. We were still not completely secure defensively and there are moments where the transitions between the IP and OOP shapes look messy, but I can see it improving. Müller and Bombino again looked important to the structure, and Vera was probably the biggest positive because he contributed at both ends and scored as well.
Salvo had a poor game after his hat-trick in the first friendly, which is probably another useful reminder not to overreact either way. One good game does not make him the answer and one bad one does not suddenly mean striker becomes a priority again. That is exactly the sort of judgement I want to get better at making over a run of games rather than from one performance.
The biggest change for me, though, was how the game felt to manage. I wasn’t changing the attacking idea every time Sacramento caused a problem. The base stayed the same and the changes came through the OOP roles, pressing responsibilities and matchups. That is much closer to how I want Villalba to evolve tactically: keep the principles, read the opponent, and adjust the defensive behaviour around what the game is actually showing me.
Leave a Reply