In September 2025, FIFA announced the number first: $60. That would be the cheapest ticket available for any of the 104 matches at the 2026 World Cup, final included. It was a clean, headline-ready promise, and FIFA kept it, technically. What didn’t make the press release is that once you add up both teams’ allocations, that $60 tier worked out to somewhere around 1.6% of the seats in the building for a given match. Everyone else was buying into a system that had never published a price list and could move the number on them between visits to the site.
This is the next post in our series on AI and machine learning at the 2026 World Cup. The first post laid out the two ways this technology enters the tournament, on the pitch and off it. The second asked whether a model could predict who wins it. The third looked at how national teams find players nobody else was tracking. The fourth turned to keeping those players healthy through a longer tournament than any before it. The fifth went after VAR, the technology fans argue about every weekend. This one goes back to a question fans start arguing about before a single ball gets kicked: who decided what you paid to get in the building.
I went through this one with Finn McCallum again, the player and coach who’s been in every post in this series, plus his own secondhand research into the buying process itself.
Six Hours in a Queue, Then a Price That Moved
Finn didn’t buy World Cup tickets himself, but he tracked the process closely enough to describe it like someone who had. Queues for ticket drops regularly ran six hours or longer. At least one release phase went live a full day ahead of its own published schedule, catching fans off guard. During checkout, 403 errors and bank-verification timeouts were common enough that tickets people thought they’d secured got kicked back into the pool for someone else to grab, and once you lost a slot that way, the system frequently locked you out of trying again for a while. People who were actively trying to pay lost both the ticket and their next shot at one.
That queue chaos works as an input to the pricing model rather than a glitch sitting outside it. A system built to move price with demand needs a live read on how badly people want in, and a stampede large enough to produce 403 errors and early-release accidents supplies exactly that signal. FIFA’s research numbers back up the scale: more than 4.5 million people applied for a spot in just the first Visa presale window, and later sales phases logged 20 million ticket requests. A queue that long isn’t incidental to dynamic pricing, it’s the demand signal the whole system runs on.
What a Fixed Price Used to Buy You
Before this tournament, Finn said, World Cup ticket prices were fixed. Not just cheap at the bottom tier, fixed. South Africa 2010 set its lowest tier around $20, reserved specifically for local, lower-income residents. Brazil 2014 did something similar around $30, with a reserved student and elderly rate that was also capped at face value on resale. Whatever the ticket cost when it went on sale is what it cost the day before kickoff.
2026 broke both halves of that arrangement. FIFA did add a low tier, the Supporter Entry Tier, priced at a flat $60. But it applied only to fans of the two qualified teams playing that match, capped at a small share of each side’s allocation, which is how you get a tier that covers the entire tournament on paper and still amounts to under 2% of the building. There’s no equivalent local-resident carve-out this time around. And unlike previous tournaments, FIFA never published a master price sheet at all. Prices moved across sales windows without an announcement, and fans mostly learned what had changed by comparing notes with each other after the fact.
FIFA has also been careful about what to call this. Executives have repeatedly said the tournament uses “variable pricing,” not dynamic pricing, a choice of wording that sounds cosmetic until you notice what it avoids. After the backlash over Oasis reunion tour tickets in 2024, UK regulators forced dynamic-pricing platforms to show a price range before a fan joins the queue. By declining the label, FIFA positions itself outside that regulatory precedent, even though the fan experience, a price that moves and mostly moves up, looks the same either way.
Why Fans Point at FIFA Specifically
Finn’s read on the anger wasn’t just about the dollar figures. It was about FIFA spending its credibility elsewhere while ticket prices climbed. He brought up two moments unprompted, and got one detail slightly out of order. The sequence changes the story enough that it deserves straightening out.
He remembered FIFA handing Donald Trump a “Peace Prize” around the time of last year’s Club World Cup. That happened separately, at the World Cup draw ceremony at the Kennedy Center in December 2025, where FIFA president Gianni Infantino presented Trump with a newly invented award for “exceptional and extraordinary actions for peace.” Human rights groups filed a formal ethics complaint over it within days, and critics called the whole thing a favor to a sitting head of state who also happens to be co-hosting the tournament. Different controversy from the one Finn was picturing, but the same underlying pattern: FIFA using the tournament’s spotlight for something other than the tournament itself.
The second one Finn described accurately. In late July, Infantino proposed selling a roughly 20% stake in a new $20 billion commercial venture that would run FIFA’s biggest events, including the World Cup, to outside investors linked to Joshua Kushner’s Thrive Capital. All 55 UEFA member federations voted to boycott FIFA competitions over it. Infantino scrapped the plan within days, but UEFA kept threatening legal action afterward, and the vote split along lines Finn had already noticed: Europe, Mexico, and Canada opposed it, while Argentina and South America’s federation stayed supportive of Infantino, the same alignment that recurs whenever expansion plans favor South American hosts down the line.
Neither of those stories is directly about ticket prices. But they’re the backdrop fans bring with them every time a price refreshes higher on the screen. When an organization spends its political capital on a medal ceremony and a private-equity pitch in the same calendar year, a resale fee starts to look less like a rounding error and more like a pattern.
The Stadium Is Never Where You Think It Is
Finn pointed out something specific to Seattle that most fans in other host cities didn’t get to experience: the stadium sits close to downtown, so getting there barely counts as a cost. Most 2026 venues aren’t built that way. He mentioned Boston’s stadium sitting roughly 20 minutes outside the city center by his own account, with local transit charging fans a steep premium for the ride.
The clearest documented version of that same problem played out in New Jersey. NJ Transit confirmed a $150 round-trip fare for the 15-minute train ride from Manhattan’s Penn Station to MetLife Stadium, about 12 times the regular $12.90 fare. Shuttle buses ran $80 round trip, and advance parking at the nearby mall went for $225. New York’s governor publicly called the fare “awfully high,” and FIFA itself objected that it hadn’t been consulted on the plan, which didn’t stop it from taking effect.
Add that on top of a ticket that was already priced above anything fans had seen at a previous World Cup, and the real cost of attending a match is two separate line items stacked together: what FIFA charges to sit in the building, and what a third party charges just to reach it. Neither number appears anywhere on the ticket itself.
What a Model Built for Fairness Would Have Done Differently
Post 2 in this series covered a lesson that carries directly into this one: a model optimizes exactly the objective it’s handed, nothing more generous than that. The same idea, explained in more general terms here, applies directly to pricing. FIFA’s dynamic pricing engine produces an $8,680 final ticket next to a $60 one nobody outside a qualified team’s fan base can buy because that’s precisely what a revenue-maximizing objective does with real-time demand data. The frustration fans describe centers on the fact that revenue was the only thing anyone ever pointed the model at.
I asked Finn what he’d optimize for instead, if he were the one setting the objective. His answer was immediate: cheapest possible, with the caveat that FIFA is nominally a non-profit and did spend real money giving every one of the 48 teams equal access to the same AI scouting and analysis platform covered in Post 1, so some revenue target isn’t unreasonable on its face. His real objection was narrower than that: fairness never got written into the model as a goal competing alongside revenue.
He had two concrete ideas for what that model could look like instead. The first was his own: if a category 3 seat is selling for $100, a category 1 or 2 seat at the same match shouldn’t be allowed to resell for three or four times that gap, a tiered cap tied to the ticket’s original category rather than a single flat rule. The second he found while researching for this conversation, not something he’d thought of himself: an AI chatbot he asked suggested capping resale at face value plus a flat 5% administrative fee, which would leave zero profit margin on any resold ticket and remove the incentive to buy purely to flip.
FIFA’s actual resale marketplace runs in the other direction. Outside Mexico, where regulators forced a face-value cap, there’s no ceiling on what a reseller can list a ticket for, a real departure from Qatar 2022’s flat cap. FIFA then charges a 15% fee to the seller and another 15% to the buyer on every transaction, a combined 30% cut that’s about triple what the organization took in the last tournament with a price cap in place. Rather than curbing scalping, that structure collects a bigger fee every time scalping happens.
The Numbers That Made It Real
A few stories from the tournament meant more to Finn than the aggregate figures did. A group of Scottish fans in Boston, facing a roughly $100-per-person round trip on local transit for a 40-minute ride each way, chartered school buses instead and split the cost across the group. Finn estimated the total savings somewhere in the tens of thousands of dollars for everyone involved, though he flagged that he was working from what he’d heard rather than a receipt. In Atlanta, a father buying tickets for himself and his son found that two of five seats bought through a resale platform never arrived, because the seller had listed tickets he hadn’t secured yet, betting prices would fall before kickoff so he could buy in cheap and deliver. When that bet didn’t pay off, the buyer was short two tickets until a stranger who’d seen the story online stepped in to help.
Neither of those is a data point FIFA reports anywhere. But by the time resale listings for the final crossed $38,000, and group-stage nosebleed seats were averaging $500 to $800 on the open market by Finn’s own rough tracking, the aggregate numbers and the individual stories were describing the same tournament from two different angles. One measures what the model optimized for. The other measures what it felt like to be inside it.
The final post in this series comes once the tournament is fully behind us, looking at what the fan experience turned into once the games were played, the broadcasts ran, and every technology this series has covered got tested in front of an audience that never asked for any of it. Ticket pricing is the clearest example so far of a model doing exactly what it was built to do. Whether that’s the same thing as doing its job well depends entirely on who you ask, and FIFA never really asked.
