The Giants Collapse in My Spreadsheet Before They Collapse on the Pitch

The Madrid derby ended 1-2, and among the thousands of comments I scrolled through after the final whistle, nearly half invoked the same word: "decline." Not one of them offered a single concrete number. I reopened my tracking notebook — the log where I record every Premier League and La Liga match from the start of the season — and saw something quite different: the problems of the giants do not lie in the results, but in the metrics that produce those results, and those metrics had been deteriorating for several rounds before the scorelines reflected it. People remember the 1-2. I remember how that 1-2 was made.

In Round 5 of the Premier League, a mid-table side scored 16 goals, while a few traditional giants conceded 3 in a single match in almost unbelievable fashion. That number 16 appeared in none of the headlines I read. That is precisely why I sat down to write this.

Context: one round, two leagues, and a story told in the wrong order

We are in a phase where emotion is compressed. The Premier League has just passed Round 5, La Liga has closed Round 7, and the international break is about to interrupt the schedule. This is the moment the public usually calls a "crisis" — but from my experience of tracking many seasons, this is precisely the phase in which data speaks most honestly. The sample is small, but tactical trends become clearer than raw goal totals.

The public picture is drawn almost uniformly. On one side, the "giants" — the names everyone assumes must win: Manchester United, Manchester City, Liverpool, Arsenal, Chelsea, Tottenham in the Premier League; Real Madrid, Atlético Madrid, Barcelona in La Liga. On the other, the "phenomena" — Brighton, Leeds, Everton — sides producing shocks. The familiar narrative is: the giants falter, the small clubs rise, and so we have a "crisis of big football."

That narrative has one fatal flaw: it reverses the order of cause and effect. It tells the result first, then looks for an explanation. I do the opposite. I start with the question: which metrics actually changed at the giants, and when did they change? As I always tell my editors — data is never in a hurry. The one in a hurry is the one who is wrong.

Here, we must distinguish two kinds of events. First, result events: wins, draws, losses, goals. Second, process events: volume of chance creation, quality of chances, intensity of ball recovery, distance covered. The second kind always precedes the first by several rounds. When a giant is about to collapse, it collapses in the spreadsheet first.

Core analysis: three layers of evidence

I build this analysis like a court file. There is precedent, there are metrics, and only then a verdict. No step is reversed.

_Layer one: chance-creation metrics (xG)._

A goal is a raw event. It cannot distinguish a six-yard finish into an empty net from a thirty-yard strike into the top corner. xG can, by assigning each shot a probability of becoming a goal based on position, angle, type of contact, defender density, and the assist situation before it. Every shot is a hypothesis. xG is how we test it.

In the data I collected through Round 5, a mid-tier side such as Brighton stood out with 16 goals. What matters is not the 16 — it is that this inverted-Ivy number came alongside a fairly modest xG in several matches. This is where I must be extremely cautious, because my data is not sufficient for an absolute claim. But if a team scores 16 while its total xG sits around the middle of that range, the gap comes from two sources: finishing above standard, or luck. And luck, by statistical definition, does not repeat systematically.

On the reverse side, the giants conceded in a manner that raises suspicion. A side like Leeds or Everton conceding 3 in a game does not necessarily mean their defense is terrible. It may mean they allowed the opponent to create low xG, but their goalkeeper let in low-quality shots. This is one of the most common traps of the casual reader: attributing a team's collapse to the defense when the data shows the problem lies in finishing or in the keeper.

_Layer two: pressing intensity (PPDA) and distance covered._

Before the summer, I published an analysis of a major national team: its pressing coefficient (PPDA — the number of opponent passes allowed before the team commits a defensive action) dropped from 8.1 to 12.6 in just four years, with average distance covered falling 6.2 km per match. My conclusion then was blunt: that team trusted possession too much and forgot to win the ball early. People called me a statistics fanatic. Three weeks later, that team held 74% possession and was eliminated in the group stage.

What I learned from that, and am applying to the current season, is this: defensive laziness does not show up in the scoreline. It shows up in PPDA. When a giant begins to let opponents build freely from the back, that is the first sign. When their off-ball distance covered drops, that is the second. The scoreline is only the third sign — and it always arrives last.

For the La Liga giants after 7 rounds, I observe a pattern: they still control possession heavily, still have high pass accuracy, but the number of passes before the opponent commits a resistance action keeps rising. In other words, they pass more but press less. And a team that does not press will eventually be pressed.

_Layer three: squad structure and points pressure._

This is the layer where I admit I am weakest in terms of public data. I do not have exact minutes played per player, no internal medical reports, no workload data on who is overloaded. What I have is the schedule and the number of matches. And the schedule tells a story: the giants must play three competitions at once — domestic league, continental cup, national cup — while mid-tier sides focus on a single competition. This is a resource asymmetry that is sometimes inverted into an advantage for the smaller club.

But I must be clear about my limits: I cannot quantify fatigue from the schedule alone. The margin of error here is large. And so I draw no conclusion from this layer. I only note it as a variable to be tested further against actual physical data, which I do not yet have. The crowd may leave the stadium, but physical data never rests — and that is precisely why I am always cautious when I do not hold it.

Contrarian angle: the asymmetry of public opinion and the "learn from the small clubs" trap

Now let us talk about what the crowd does not want to hear.

When a mid-tier side scores 16 goals or strings together good results, the first media reaction is: "the giants must learn from them." This is a fallacy I have met many times in my career, and it has a name: correlation is not causation.

Brighton scoring 16 goals in a few rounds does not prove they have a tactical model that Manchester City or Liverpool need to copy. It may simply prove they are at the peak of a lucky cycle, where their low-quality shots are going in at a rate above what is sustainable. If that is true, their next run of matches will cool down — not because they play worse, but because probability drags the result back toward its true value.

On the other side, when a side like Real Madrid is called "in decline," I want to ask: in decline relative to what? Relative to a season when every shot went in? Or relative to an average standard they themselves set in the past? If it is the latter, then what is called "decline" may simply be a return to their own realistic level.

And here is the most sensitive part: the story of a great coach like José Mourinho. The question "has he passed his peak?" appears everywhere. But to me, that is a wrongly framed question. A coach trusts reputation. Data trusts repetition. The issue is not whether a coach is still good. The issue is: how much xG per match does his team structure allow opponents to create, and is that number stable or worsening? If it is stable at a good level while results remain poor, that is a luck or finishing problem. If it is worsening, that is a systemic problem.

I refuse to answer the question in the name of emotion. No verified data, no conclusion — a principle I set at Lạch Tray stadium in 2026, and I have no reason to break it now. That day, the home side created 1.92 xG but lost 0-1, and the whole city called it a "decline." I called it "random injustice." Two weeks later, the head coach publicly cited my numbers in a press conference.

That is why I say: people remember results; I remember the conditions that formed the results. Without the conditions, the result is just noise.

What comes next: signals for the coming round

If you want to follow this season with a data eye, here is what I will be watching. First, do not look at the scoreline — look at the gap between xG created and xG conceded. A giant winning with a negative gap is a time bomb. Second, track the PPDA of the big clubs round by round; if it rises steadily, that signals they are losing their ability to press high. Third, for teams scoring above standard, prepare for a cooling period — not because they got weaker, but because probability is self-correcting.

Data is never in a hurry. And in a season where emotion is compressed like this, the one who knows how to wait will see the truth first. I am still taking notes. And I will keep taking notes until the spreadsheet says what the pitch has not yet said.