An AI-Built ASEAN Supporter Loyalty Ranking: Vietnam First, Indonesia Seventh, and the Flaw in an Unverifiable Metric

On 17 October 2026, the channel ASEAN Football, a media outlet based in Malaysia, published a supporter-loyalty ranking of Southeast Asian national teams. Vietnam led with 85%, followed by Thailand 80%, Malaysia 78%, Cambodia 75%, Brunei 72%, Singapore 70%, Myanmar 60%, Laos 55%, Timor-Leste 50% and the Philippines 45%. Indonesia placed seventh with 65%.

Years of writing disciplinary match reports in France taught me something simple: when a number appears without a method attached, the first task is to find out what it was actually measuring. In the very article publishing the ranking, the outlet admits the figures are not official federation statistics but were compiled by artificial intelligence analysing social media posts. The same article supplies attendance data from the 2026 AFF Cup: Vietnam averaging 25,599 per match, Indonesia 24,875, Cambodia 17,169.

Placed side by side, the two data sets open a crack that anyone who has ever held a whistle has to stop and look at.

What this ranking actually measures

According to the publisher's own description, the criterion is the enthusiasm of supporters in accompanying their team on away trips abroad. The scope is limited purely to fans flying directly from their home country. Vietnamese working in Japan, South Korea, Taiwan or the Czech Republic, the group that routinely fills a corner of the stand whenever the national team plays in Asia, is not counted. Indonesian communities in Malaysia, Saudi Arabia or the Netherlands are likewise excluded.

This is a defensible definitional choice, but it shifts the result systematically. A supporter flying from Hanoi to Bangkok needs a passport, a visa, an air ticket, leave from work and a not insignificant sum of money. A Philippines-born supporter living in Los Angeles only needs to drive to the stadium. When the metric records only the first group, it is not measuring loyalty. It is measuring the disposable income and time flexibility of the middle class, two economic variables rather than two emotional ones.

I once made exactly this kind of error. In September 2026, working as a league disciplinary reporter for Ligue 1, I was sent to Groupama Stadium to cover Lyon against Marseille. In the first half I recorded Dimitri Payet committing three fouls instead of four. The disciplinary report was sent back by the organisers. I spent the following four weeks reviewing footage of twelve Marseille matches to cross-check every incident the referee had whistled. The mistake at the 2026 World Cup qualifiers taught me this: you do not write the report in advance. And a report that gets one foul wrong gets the whole substance wrong.

A twenty-point gap and a three-percent gap

This is the most analytically interesting part of the whole story. Vietnam is credited with 85% loyalty, Indonesia with 65%. A twenty-percentage-point gap. Yet average attendance at the 2026 AFF Cup differed by just 724 people out of almost twenty-five thousand, roughly three percent. One article, one editorial desk, two numbers telling two different stories.

If Indonesian supporter loyalty were only 65% of Vietnam's 85%, then in a home stadium, where no passport or visa is required and the only barriers are ticket price and distance, the gap should be far wider. It is not wider. It is almost identical.

When a metric contradicts its own data, it ceases to be a metric. It becomes an opinion packaged with a percentage sign.

Let me be clear about this to avoid misunderstanding. I am not disputing Vietnam's position. What I have observed over years covering national-team qualifiers and regional tournaments is that the reach of Vietnamese supporters is real, particularly on away trips within the region. The problem lies elsewhere. A correct conclusion can still be reached by a flawed method, and when that happens the conclusion loses its value the first time anyone checks it.

The reaction after publication is itself worth recording. The original article notes heated discussion among Southeast Asian football fans, along with the assessment that Vietnamese supporters can out-sing fanbases long known for being large and loud. For a media outlet, a controversial result is a good result. For someone writing a report, a controversial result is a file that needs reopening.

A self-feeding loop and the reporter's problem

There is a noteworthy technical detail in how this ranking came about. The algorithm aggregates social media posts to produce the figures. Once published, those figures generate a fresh wave of posts, and that wave becomes the raw material for similar rankings in the future. Input and output are the same thing.

In refereeing we call the equivalent situation a loop with no exit: the referee whistles, the players react, the stands react, and the footage only has value when it is reviewed at normal speed rather than through the crowd's reaction. My method for covering the 2026 World Cup was a net with a fine mesh: small eyes, no fish missed. I chose the fourteen lowest-scoring group-stage matches, ignored the fixtures the media chased, and found that Iran under Carlos Queiroz had the tournament's highest rate of counter-attack-stopping fouls, twenty-three in three matches. Nobody reported it, because it was not where the crowd was looking.

Applying that net to this ranking, I find three threads worth keeping.

First: the ranking measures the segment of supporters capable of international travel, not the total fanbase. Second: excluding the diaspora creates a systematic advantage for countries with smaller overseas communities. Third: the percentages are presented with false precision, with no sample size, no confidence interval and no model description.

A percentage without a sample size is a brand, not a measurement.

The contrarian angle: what is right inside a wrong ranking

Here I want to separate myself from the usual reflex of the trade, which is to reject everything once a methodological flaw is found. Total rejection is another form of laziness.

This ranking, weak as its method is, gets one point right: Vietnam and Indonesia are the two largest supporter markets in Southeast Asian football. That conclusion can be verified independently through attendance data, through ticket sales for away fixtures in the region, through the media coverage both countries generate. For federations, sponsors and broadcast rights holders, this is genuinely valuable commercial information.

What deserves caution sits one layer above. An unofficial ranking with a contentious result and a ready-made disclaimer is a familiar content pattern: high engagement, low accountability. The disclaimer mitigates risk for the publisher, but in practice it is usually left behind as the information spreads to other platforms. The ranking travels first, the small print follows, and most readers only ever see the ranking.

There is one more small detail the net catches. The publisher's name, ASEAN Football, is easily confused with the ASEAN Football Federation, the region's official governing body. To a skimming reader, a ranking by a Malaysian media outlet can easily be remembered as a document from the federation. The fault does not lie with the reader. The risk lies in the structure of the naming.

Takeaway

What I want to keep after reviewing all the data lies elsewhere, not in the placings of Vietnam or Indonesia. It is the question of whether this region lacks a trustworthy measurement system for what is called supporter strength.

Southeast Asian football has attendance data, ticket figures and a fully documented head-to-head history. That is enough to build a verifiable supporter-strength index, with sample sizes, timeframes and clear definitions of diaspora and home fixtures. Until someone does that, algorithm-generated rankings will keep being cited as fact, not because they are correct, but because they are the only thing that exists.

For someone who writes reports for a living, that gap is more worrying than any argument about placings.