Home
Hobocode Counter-Hate Monitor Special report · The 2026 World Cup · July 2026

Hate at the World Cup came from the mainstream, not the fringe

We followed 33 accounts and posts tied to the 2026 tournament, from the February buildup through the July 19 final. Nine crossed the line from fandom, policy, or criticism into hate. Every one of them ran a script that existed long before the draw.

The world's biggest sporting event landed in a country mid-argument with itself, and everyone with a grievance found a match to hang it on. A disallowed Egyptian penalty became proof that a religious minority secretly runs FIFA. A riot after Morocco beat the Netherlands became a loyalty verdict on four generations of Dutch citizens. Well-behaved Scandinavian fans became a weapon for a fabricated claim about who else gets let in. None of these stories needed the tournament. The tournament gave them a stage, a crowd, and fresh material.

We checked 33 accounts and posts that engaged with the tournament, drawn from our 925-account monitoring roster plus items surfaced during the tournament itself. Nine produced hate under our standard: language that collectivizes or dehumanizes a whole group, rather than criticizing a policy, an institution, or a person's own conduct. The nine sort into five underlying scripts, and the most consequential fact about them is who was doing the talking. Three of the nine came from elite carriers with audiences in the millions: a national television host, a syndicated pundit, a sitting Dutch party leader. Only two came from what anyone would call the fringe.

The other 24 accounts are just as much the story. Nine immigration and civil-rights organizations spent the tournament criticizing ICE raids, entry denials, and FIFA's own conduct, some in very sharp language, and every one of them stayed on the protected side of the line: aimed at governments and institutions, never at a people. We read all of it in full, and it's browsable below, because a standard that only publishes what crossed the line never shows anyone where the line is.

9confirmed hate items, from 33 checked
5narrative scripts, all predating the tournament
3elite carriers with audiences in the millions
0confirmed hits from the 9 advocacy organizations checked

01 · Who carried it

Where the nine hits sit in the media ecosystem

Hover a segment for the accounts in it. Elite and institutional carriers outnumber the fringe, the opposite of where hate monitoring usually expects to find them.

02 · The five scripts

How the five scripts score

Each script is scored 0 to 5 on reach, toxicity, coordination, mainstreaming, and harm potential. Click one for its scorecard. Coordination sits near the floor across the board: these carriers do not cite each other, and the scripts spread without anyone organizing them.

03 · The nine

Every confirmed item

Click a row for the quote, the coding, and the source.

04 · Where the line held

Twenty-four items that stayed protected

Sharp criticism of governments, institutions, and named individuals stays protected under our standard, whoever it comes from. Click a row for why each item did not qualify.

05 · Method

How this was built

What counts as a hit

Language that collectivizes, dehumanizes, or calls for exclusion of a group defined by race, ethnicity, religion, or national origin, or that runs the same move against a political camp through dehumanizing or eliminationist framing. The implied claim has to be statable as "group are/do predicate." Criticism of a government's conduct, an institution's policy, or a named person's own behavior stays protected, however harsh. The nine hits and the 24 protected items above were all coded under the same test, symmetrically.

The severity score

Five factors, each scored 1 to 5: reach (R), toxicity (T), coordination (C), mainstreaming risk (M), and harm potential (H). They combine as (R·T·C·M1.3·H1.5)1/5.8: multiply all five, with mainstreaming risk raised to the 1.3 power and harm potential to the 1.5 power because those two carry the most real-world weight, then take the 5.8th root. The 5.8 is the sum of the weights (1 + 1 + 1 + 1.3 + 1.5), which brings the result back onto the 1-to-5 scale. Multiplying instead of adding means no single factor can carry a score alone. Coordination scores above the floor only on real signals: a named hub, near-duplicate phrasing, synchronized timing, or a confirmed bridge between actors.

Verification and reach

Every item rests on an independently fetchable, dated source read in full. Reach uses the first real number available: a direct post metric, then follower count, then domain traffic. Where no verifiable figure existed for an item, its card says not measurable rather than carrying an estimate. One follower figure traces to Wikipedia rather than a live platform count, and its card says so.

Support the fundraiser →
Hobocode Counter-Hate Monitor · Online Hate Narrative Monitor methodology · Window: January 1 to July 20, 2026 · Words by Jerm in collaboration with Claude
Digital Civility Research: when speech moves toward harm. Formerly the Counter-Hate Monitor, the name this research was published under through mid-August 2026.