76.98 Per Cent: The Autopsy of a Mourinho–Guardiola Number Whose Denominator Nobody Counted

Eleven minutes past eleven at night. A screenshot arrived: the all-time La Liga win-percentage table. Top, José Mourinho — 126 matches, 97 wins, 76.98 per cent. Second, Pep Guardiola — 152 matches, 116 wins, 76.32 per cent. Beneath it, a caption: 'This is why Mourinho is the greatest coach in Spanish league history.' The gap is 0.66 percentage points. Small enough that one match, one deflected goal, reverses the order. And yet the screenshot was being forwarded like a verdict, as though the hearing had already closed. I did not reply with an opinion. I replied with a question: where does 126 come from? Because the notebook on my desk says otherwise. Twenty years ago, at the 2026 US Open in September, when no vendor would sell point-level WTA data to our desk in Ahmedabad, I hand-coded all four of Sania Mirza's matches off a rented feed — 387 points, fourteen columns, fourteen nights. The sheet said she won 31 per cent of second-serve points. The studio called it negativity. Since then I have held one rule: I do not trust a number whose denominator I have not counted myself. This piece is that rule applied. The question is not whether Mourinho is a greater coach than Guardiola. The question is where every column in the table that supposedly proves it actually came from. At London 2026 I ran a pairing model on the Indian doubles combinations — hold percentages, return games won, super tie-break records across 2026 to 2026. No Indian pair reached a medal round. In the mixed zone a British colleague asked who had let 'the statistician' in; my AITA accreditation was checked twice that week. I began appending a methodology box to every analytical piece — sample size, coding rules, and what the model cannot see. That box goes on this claim today, and the conclusion is declared in advance so no correction has to be negotiated later. Win percentage is a dirtier unit in football than in tennis. Tennis hands you points, games, service and return separately, so process and outcome can be separated. League football folds draws and defeats into one bucket and ignores opponent strength, squad investment, fixture difficulty and refereeing variance. A 38-round round-robin is this sport's Grand Slam tier — highest domestic level, no discretionary entry, rotation the only elasticity. In that frame a single percentage point cannot summarise three years of work. If it has been made to, it must be opened. Start with the denominator. Mourinho managed Real Madrid across three full La Liga seasons — 2026/11, 2026/12, 2026/13 — from July 2026 to May 2026. Three seasons at 38 matches each is 114. Not 126. Twelve matches is a third of a season. If the numerator is right and the denominator is 114, the rate becomes 85.09 per cent — a figure no public league record carries. If the numerator is wrong, the entire table falls. Either way the headline does not survive. And the whole 'lead' in the comparison — that 0.66-point margin — is generated by this single denominator and nothing else. One thing must be said plainly, because this is where readers are misled. The arithmetic is internally perfect: 97/126 = 76.98, 116/152 = 76.32, gap 0.66. The three numbers reconcile. But internal consistency is not external validity. The error sits upstream, in the match count, and a perfect arithmetic built on a wrong denominator is only a tidier lie. Guardiola's 152 — four seasons from 2026 to 2026, 38 each — is arithmetically clean and matches the calendar. Mourinho's 126 matches no season count. Second, the metric's job. Win percentage is a low-information instrument in football. It suppresses draw frequency, which is the single largest determinant of who wins a league — one side can draw twelve and finish champions, another can draw four, lose three and finish second. A table without draws is not a league table. Third, the figure the author did not hide testifies against his own headline. The trophy ledger reads: Guardiola, three La Liga titles and two Champions Leagues in four seasons at Barcelona. Mourinho, one La Liga in three seasons at Real Madrid, no Champions League. A 0.66-point win-rate edge coexisting with a 3:1 title deficit does not establish that win rate is the mark of the bigger coach. It establishes that win rate is the wrong metric. Titles are the resolved output, the place where draws, defeats and pressure all get priced in. Fourth, 2026/12. That season is real, documented and an outlier: 100 points, 121 goals, a La Liga record. Mourinho's side played the fastest counter-attacking football in Europe, and that season is the cleanest sample of his career. The article itself concedes the mark is 'very difficult to surpass.' Two paragraphs later the same article reports that Barcelona has scored 48 goals in 12 rounds — 4.0 per match, a pace of roughly 152 goals across a 38-round season — alongside eleven wins from twelve, 33 points. The piece first declares 121 goals unsurpassable and then describes a team on course to pass it comfortably. That is not a debating point; it is an editorial collision. Two data blocks from different times and different sources were lifted onto the same page. Fifth, the heaviest problem of all — the player list. The article describes a 'current Real Madrid' under Mourinho and names Kylian Mbappé, Jude Bellingham, Aurélien Tchouaméni, Arda Güler, Federico Valverde and Bernardo Silva. The season given is 2026/27. Mourinho's Real Madrid tenure ended in May 2026. Valverde came up from Castilla in 2026. Tchouaméni arrived in 2026. Bellingham and Arda Güler in 2026. Mbappé in 2026. Bernardo Silva has been a Manchester City player since 2026 and has never been a Real Madrid player in any recorded source. The squad described did not exist under the coach being described, and the season described has not yet been played. This is not a typing slip. It is the load-bearing wall. When the central fact of a piece is lifted from another coach's team, auditing its numbers is no longer the task; recording the finding is. Sixth, the tactical contradiction. The technical section describes a high defensive line, aggressive counter-pressing inside the opponent's half, an inverted right-back in the half-space and a creator operating between the lines. That is the direct opposite of the deep block and transition model that built Mourinho's identity. The article never reconciles the inversion — it simply asserts a reinvention. And when you announce a stylistic inversion of this scale, you owe the numbers that measure it: pressing intensity, PPDA, possession share, xG, average defensive-line height, recoveries in the opponent's half. Not one appears. I have a 2026 sheet with fourteen columns behind it and fourteen nights of coding; this has zero columns and one paragraph. Seventh, the article's own defence. It tries to prove Mourinho's attacking credentials with the 121 goals of 2026/12. But those goals came from transitions — running, breaking from deep, moving the ball forward early. Scoring volume is not chance-creation method. Being the highest-scoring team in a league is not the same as being a proactive possession side, and conflating the two is a category error of the kind that produces 'he batted better because he hit more fours.' My notebook taught me that distinction the hard way: keep the counting stat apart from the process stat. Eighth, the inverted resource picture. The article frames Mourinho as the underdog against a dominant Barcelona. Between 2026 and 2026 Real Madrid was the most expensively assembled squad in world football, and Guardiola's Barcelona sat in the same top tier of wage spending. Both operated at maximum resource. Casting one as an underdog is reading the ledger upside down. Ninth — and this is the part that stings — the article buries its strongest evidence. Mourinho's genuine, rare claim is not a win percentage. It is portability: league titles in four countries — Portugal, England, Italy, Spain — and European Cups with two different clubs, Porto in 2026 and Inter in 2026. That profile belongs to a handful of coaches alive. It sits in paragraph eleven, beneath a fabricated percentage table. No researcher places their best exhibit at the bottom unless they were in a hurry to put the wrong thing at the top. I will register the test publicly, because a conclusion that cannot be falsified is not a conclusion, it is an opinion. My claim is that the 126 denominator does not come from any primary league record. To disprove it, someone need only produce one document: an official La Liga match register showing 126 league matches under Mourinho's name. If it appears, I will publish the correction, in this column, in this space. If it does not, the table must stop being called a table. The real problem is not that someone argued for Mourinho. Arguing for Mourinho is a legitimate journalistic project — his reputation has slid since 2026 and that slide deserves an audit, and his multi-country title record is a solid foundation for it. The problem is that the audit was hung on a number nobody counted. And beneath that sits a category problem. These two coaches do not belong on one axis. One is a knockout specialist whose method travels across borders and whose weakest link has always been domestic consistency. The other is a league-accumulation specialist whose sides are built to bank ninety-plus points. Ranking them by a single win-rate figure is like ranking a serve-and-volleyer against a clay grinder by first-serve percentage. Selection is never a story about talent — it is always a story about the frame the selection happened inside. Here the frames are different shapes. I should also record what the piece does not contain: not one line on disciplinary record, club governance, contracts or coaching licensing. An article claiming to identify the greatest league coach in history never mentions his touchline record. That is not an oversight. That is curation, and curation always reveals the conclusion the writer intended to reach. The next-round signal is printed inside the same article and the author never joined the dots. After twelve rounds: Barcelona three points clear, seventeen goals better on differential. If the purpose was to prove Mourinho superior, that is the line testifying against him — and it is in his own piece. An analyst who reads the counter-evidence in his own copy and files it separately is not an analyst; he is a lawyer. What startles me is not 76.98. It is that nobody asked about 126. Four seasons versus three is such an ordinary count that it never feels worth checking — which is precisely the signature of the best kind of falsehood. Every collapse has an entry point. My job is to find it, and in this case the entry point is a denominator. The market speaks in rumour. I answer in columns.