Tab 3 — The Probabilities

The probabilities

Four hypotheses try to explain the noise around Argentina. Each gets a speculative range, the three best points for and against, and an honest 'what would change my mind'. These are my subjective estimates — argue with them using the Bayesian explorer below.

H1Sporting thread

Deliberate FIFA-level officiating favoritism

The strong claim: FIFA-level actors are deliberately favoring Argentina through appointments and calls. On base rates and detectability, I think this is unlikely.

2%–8%speculative probability this is happening Speculative

For

  • The penalty outlier is real — 8 across two World Cups, roughly double any other side.
  • The Pinheiro appointment and an all-Argentine crew nearby were, at minimum, bad optics.
  • A Messi farewell has enormous commercial value, which is a motive if you're looking for one.

Against

  • Zero proven finals precedent, ever — the base rate is punishing.
  • Each flagged call has a defensible, rule-based explanation on the record.
  • The all-Argentine crew was on a rival's match — the opposite of what favoritism would design.

What would change my mind

Leaked comms, a VMO or appointer whistleblower, or a rigorous assignment-anomaly study linking appointments to outcomes would move this sharply upward.

Sources:Yahoo Sports — 'FIFA's Argentina Problem' ↗ · NBC DFW — the Pinheiro appointment scrutiny ↗ · Wilson Raj Perumal — how real fixes actually happened ↗

H2Financial thread

AFA financial corruption (the money thread)

The credible thread. I split it in two: that a genuine preliminary inquiry exists, and — separately and lower — that chargeable US conduct is eventually established. Note: this says nothing about match outcomes.

70%–90%that a genuine preliminary inquiry exists Speculative
45%–65%that chargeable US conduct is eventually established

For

  • La Nación plus Miami Herald sourcing — a newspaper of record and independent US corroboration.
  • Specific detail (a Florida company, named prosecutors) and a perfect FIFA-Gate jurisdictional fit.
  • Tapia's separate Argentine charges and the December 2025 raids show real regulatory heat.

Against

  • No charges have been filed, and the FBI declined to comment.
  • Key financial specifics are single-sourced or rest on anonymous law-enforcement sources.
  • The AFA denies wrongdoing and frames the domestic cases as political persecution.

What would change my mind

A grand jury subpoena, an indictment, or a cooperating witness would raise the second number; a documented AFA rebuttal of the money trail would lower both.

Sources:La Nación (Olivera & Alconada Mon) — the FBI/AFA report ↗ · Miami Herald — Florida firm linked to the FBI probe (via Tampa Bay Times) ↗ · US DOJ (EDNY) — 2015 FIFA racketeering indictment ↗ · US DOJ — Jorge Arzuaga guilty plea ↗ · La Nación — Tapia processing upheld on appeal (Argentine case) ↗

H3Sporting thread

Soft / structural bias (no conspiracy)

The quiet, well-studied one: referees, under crowd and status pressure, unconsciously shade toward high-profile sides. No conspiracy, no instruction — just human bias. I think this is a material contributor.

35%–55%as a material contributor to the perception Speculative

For

  • Peer-reviewed status-bias literature — Garicano et al. found ~2× injury time favoring home sides under pressure.
  • Argentina fits the champion-with-a-megastar profile that draws deference (cf. Erikstad & Johansen).
  • VAR-era penalty inflation gives borderline calls more chances to break one way.

Against

  • Some studies find null results (e.g., Morgulev et al. in basketball) — the effect isn't universal.
  • Argentina's key calls each have defensible, specific rule-based explanations.
  • The rival-crew assignment is inconsistent with systematic shading toward Argentina.

What would change my mind

An xG-adjusted study of Argentina's decisions versus peer teams — showing they do (or don't) get more marginal calls — would sharpen this a lot.

Sources:Garicano, Palacios-Huerta & Prendergast (2005) — 'Favoritism Under Social Pressure' ↗ · Erikstad & Johansen (2020) — favoritism toward successful teams ↗ · Morgulev et al. (2018) — a null result on referee bias ↗

H4Sporting thread

No bias: variance + a perception cascade

The boring, powerful one: three straight nervy knockouts plus a couple of viral clips plus social media equals a narrative — with no bias at all. Combined with H3, this is my best explanation of the perception.

30%–50%as the primary explanation of the perception Speculative

For

  • Three straight one-goal / extra-time knockouts naturally breed 'rub of the green' stories.
  • Correct rule-based explanations existed for the viral moments (the Cape Verde restart, the Embolo dive).
  • Collina's on-record rejection of bias is a direct, accountable statement, not spin.

Against

  • The penalty outlier stubbornly persists across two tournaments.
  • Complaints came from many independent pundits, not only the losing sides.
  • 'It's just variance' is unfalsifiable if you never look at the pattern rigorously.

What would change my mind

If the officiating pattern regresses to the mean under semifinal and final scrutiny, that's strong confirmation; if it intensifies, this weakens.

Sources:FIFA / Collina — rejection of the Egypt bias allegations (link unverified — search) · IFAB — Laws of the Game, Law 5 (bleeding player) ↗

Speculative

Posterior

My published range

The math — Bayes

Bayes in one line: turn your prior probability into odds, multiply by the likelihood ratio of each piece of evidence (how much more expected it is if the claim is true vs false), and convert the result back to a probability. Doing it in log-odds just means adding instead of multiplying, which keeps the arithmetic stable. Strong evidence has a ratio far from 1; weak evidence sits near 1 and barely moves the needle.

Pick a hypothesis, set your prior, then slide the evidence strengths. The tool combines them in log-odds and shows the posterior against my published range band. Reset returns the defaults that reproduce my ranges. It's an intuition pump: the evidence is treated as independent, which it isn't fully — so this reasons about beliefs, it doesn't compute the truth.