Late-Season MLS: When 31 Shots Are Not Enough to Win a Match

Late on Saturday night, I sat in front of a screen in Liverpool, watching D.C. United lose 2-1. When the statistics appeared, I paused. Thirty-one shots. 3.4 xG. One goal. This was not the first time I had watched a team dominate by the numbers and still lose, but the figure forced me to rewind the footage to check whether my own perception had drifted.

Data whispers; those who listen hear miracles. But the miracle does not always arrive as a victory. Sometimes it arrives as a number that has to be decoded before anyone understands what just happened.

The context of a distorted weekend

MLS is entering the final stretch of the regular season. This is a period when every metric is decisive yet simultaneously most easily distorted, because every match is shaped by playoff arithmetic, by the physical toll of months of competition, and by off-field pressures no statistical table can measure.

Three stories dominate this week. Inter Miami, a side described as objectively not a very good soccer team, keeps accumulating points through Lionel Messi. LAFC is struggling to convert two expensive attacking names into goals. And D.C. United has just spent around 10 million dollars reinforcing its attack while leaving gaps in midfield and defence.

What stands out is that all three stories share a common denominator: each club is revealing the gap between what it declares and what it actually does on the pitch. I have followed MLS long enough, since the days when the league was dismissed as a retirement home for European stars, to recognise that such gaps are not new. They simply become visible once data is collected seriously.

Core analysis: three cases, three distinct problems

Start with D.C. United, because this is the case where the data speaks most clearly. They took 31 shots and generated 3.4 xG in a defeat. I checked the trajectory of every attempt. Most came from positions any xG model must classify as good chances. D.C. United's problem is not chance creation; it is finishing quality and conversion.

To place 3.4 xG in context: an average MLS side generates roughly 1.3 to 1.5 xG per match. Reaching 3.4 means they created more than double the league average. And they still lost. Tai Baribo, signed to be the lead striker, has 14 goals, while the rest of the squad has 17 combined. Concentration of output in one individual is a double-edged sword: high short-term efficiency, but systemic risk if that individual loses form or is marked out.

What the analysis of D.C. United overlooks is that the problem lies in a multi-year recruitment structure. For nearly a decade, the club has been perpetually on the verge of a strong offseason. They buy strikers but not creative midfielders. They buy attackers but do not reinforce defence. This is a star-topping pattern rather than holistic squad-building. The 10 million dollars spent on two attackers is not the issue. The issue is where that money goes within a squad structure that is short elsewhere.

Next is Inter Miami. This is the inverse case. The team is, objectively, not a very good team, yet keeps accumulating points. Messi has 31 goal contributions in 23 matches, a rate of 1.35 per 90 minutes. By comparison, the average for a top European forward sits around 0.7 to 0.9. Inter Miami is operating an individualised attacking model, not a tactical system. This means their results depend on a non-replicable variable. When Messi plays, they win. When he is absent, the structure collapses.

This is not a moral judgment. It is an observation about systemic risk. And it raises a question the media often avoids: if a team wins through one individual, whose victory is it? Statistically, the answer is clear. It belongs to that individual, which means the system remains unverified.

Then there is LAFC. Marc Dos Santos declared a model modelled on Luis Enrique at PSG: Son Heung-Min and Denis Bouanga switchable, producing a flexible three-pronged attack. In theory, a reasonable idea. In practice, LAFC sits sixth in the West, and at least six teams have scored more. Son has 6 goals. The gap between stated model and output is LAFC's core problem, and it is not about player quality.

A side with two attackers of the calibre of Son and Bouanga that still cannot score enough has a structural problem, a problem of spacing between lines, a problem of how the ball is circulated into dangerous zones. But it could also lie in something that no metric captures: the cohesion between individuals inside a new system that has not yet been tested by time.

I once stood before a spreadsheet and felt I was witnessing a miracle at Anfield. In 2026, when I calculated Liverpool's average PPDA under Klopp at 8.2, the league's lowest, I believed I had seen something others had missed. I was right about the number. It took me years to understand that a correct number does not automatically produce a correct outcome if the context does not allow it.

A contrarian angle

Here I must self-criticise. Every analysis above rests on a very small sample: one MLS weekend. I made a similar mistake in 2026, when I confidently predicted France would win the World Cup based on a homemade xG model, then had to hide in a library for two weeks reviewing the data after discovering my model ignored corner kicks. That error taught me that correlation is not causation, and that a small sample can create the illusion of a pattern.

With Inter Miami, there is a perspective the media overlooks. Yes, they depend on Messi. But in football, dependence on an exceptional individual is not always a weakness. Sometimes it is a strategy. The real question is not whether they depend on Messi, but how long they can sustain it. For a player in his thirties, the answer lies not in present data but in biological trajectory.

With LAFC, another possibility exists. Mid-season tactical model change is one of the hardest problems in football. A new manager has no pre-season to install automatic patterns. Dos Santos's switchable model demands repetition and tactical discipline that only time produces. Judging Dos Santos on current output may be a methodological error, because we are measuring an unfinished process.

Those who are right before their time always pay with solitude. But not everyone solitary is right. That is the thin boundary I try to hold in every analysis.

And here is the point I want to stress, because it concerns the nature of this profession. The original headline asks what is next for Marc Dos Santos. But that is a media question, not a data question. Data can only say that, under the current structure, LAFC is underperforming. It cannot say Dos Santos is the cause.

The key takeaway and next-cycle signals

Among these three stories, one signal strikes me as more durable than the rest: MLS's youth development pipeline. Nine of thirteen uncapped United States internationals play in MLS. Cavan Sullivan leads in both numbers and star power. Julian Hall, 18, scored from an NYCFC defensive organisation error, in a backline containing two United States internationals.

This is the week's most important data signal: MLS is no longer a place where careers end, but a place where they begin. And that matters far more than any single win or loss.

Next round, I will track three metrics. First, D.C. United's xG conversion. If the gap between xG and actual goals persists, it is a finishing-coaching problem, not a structural one. Second, Messi's availability, because each absence is a natural test of the dependency hypothesis. Third, LAFC's playoff progression, because that is the decision trigger for Dos Santos's future.

I learned at Anfield that belief is also a variable. But belief in data must be tested with patience, because sometimes the right course is not to change the manager, but to give the process more time to prove itself. In a world of long seasons, the awakened can only rely on their own spreadsheet.