Apple's Sports App and the Quiet Fight Over Sports Data

Apple's Sports app does almost nothing, and that is the point. Slate's reviewer, writing on 10 June 2026 as the World Cup kicked into gear, counted what it offers: team names and logos, records, game times, live and final scores, a standings page, and a toggle that switches off the betting odds sitting on the scoreboard. No news headlines. No streaming guide unless you tap a game. No design feature built to keep you scrolling.

That restraint is unusual enough to be newsworthy. It also lands in the middle of a much bigger argument about sports data: who collects it, who pays for it, and who gets to decide which version of a number is the true one.

A minimal app in a maximal market

Apple announced Sports in the winter of 2024. According to Slate, the pitch was one sentence from an Apple executive: "We created Apple Sports to give sports fans what they want, an app that delivers incredibly fast access to scores and stats." The app covers World Cup matches plus soccer in MLS, the NWSL and roughly 30 other leagues, including Ecuadorian Serie A and second-division Bundesliga, and adds men's and women's tennis and the LPGA Tour. Tap a game and you get a basic box score, leaderboard or rosters. Live Activities can push a team's scores to the lock screen.

The broader market around that simplicity is anything but quiet. MarketsandMarkets, in a report carried by Yahoo Finance on 22 September 2026, projects the AI in sports market at $2.61 billion by 2030, growing at a 16.7% compound annual rate. Recent coverage in the same window includes a Databricks partnership with Cal Athletics, an Atlantic Council sports centre, and a New York Post report alleging DraftKings used AI to target losing bettors. None of that friction shows up inside Apple's app, which is precisely why the reviewer liked it.

The counter-argument is that score-checking was never the interesting part. The interesting part is what happens to the data underneath: the event feeds, the tracking systems, the integrity checks. That is where the money and the controversy sit.

Where the data actually comes from

The Guardian's Sean Ingle made a version of this case on 15 July 2025, after Wimbledon's second year of electronic line calling. The headlines that fortnight were about failures. The system missed a Sonay Kartal ball that landed long against Anastasia Pavlyuchenkova, in a case where an official had accidentally switched the system off. A Taylor Fritz forehand was called out despite landing four feet inside the baseline, with a ballboy still on court when the American began his serve.

Ingle's point was that the outrage buried the base rate. Wimbledon was running a souped-up version of Hawk-Eye, in use there since 2007. Researchers have estimated line judges get around 8% of close calls wrong. Players do worse. Of the 1,535 challenges across men's and women's singles at Wimbledon in 2024, just 380, less than 25%, were overturned. When a player thought the ball was out, they were wrong three times out of four.

"No system is 100% perfect, but they are demonstrably more accurate than relying purely on human decision-making," said Matt Drew, who founded the integrity department at StatsPerform, in comments published by the Guardian.

Ingle also cited two studies on human bias. One asked 40 qualified football referees to judge 47 incidents from a Liverpool-Leicester match; half watched with crowd noise, half in silence. The group with crowd noise awarded significantly fewer fouls against Liverpool, 15.5% fewer. Another, in Norway, found successful teams were more likely to get favourable penalty decisions. Psychologists call it conformity.

That is the case for the machines, and it is a data case, not a sentimental one. Accuracy is measurable. Bias is measurable. The two can be compared.

The analytics bill nobody budgets for

Then there is the layer most fans never see: the plumbing that turns raw events into a number someone trusts. A post on the Cassis blog, published 23 August 2026, argues that cheaper SQL does not make analytics cheap, because every answer carries a human cost that gets redistributed rather than removed. The post breaks that cost into three parts: indexing, the upfront work of defining metrics and documenting logic; querying, the effort at retrieval; and reliability, whether the number that comes back is semantically right even when the SQL runs fine.

The example is a CFO asking how many active enterprise customers renewed last quarter. Before anyone answers, someone must decide which table holds customer data, what "active" means, what counts as "enterprise", and what "last quarter" means in the fiscal calendar. Cassis describes one senior data manager running a weekly rotation where a single person handles ad-hoc questions, consuming about 20% of a team of five or six people, permanently.

Sports organisations have the same problem with different nouns. What counts as a completed pass? A possession? A tackle? A fan's session on a club app? The definitions are the product. Get them wrong and the dashboard still renders.

Builders are shipping the boring parts

Two open-source projects published in September 2026 show how much of this work is now commodity infrastructure. Adaca Analytics, a self-hosted dashboard for Google Analytics 4 built on Cloudflare Workers, stores daily rollups from the GA4 Data API or a BigQuery export in a D1 database the operator owns, and keeps realtime on Google. It ships six dashboards and a four-step builder, with drill-down pages for sources, pages and countries, read-only share links that can pin a filter, and weekly or monthly summaries sent by email or Slack. It is MIT licensed and comes from Adaca, a software consultancy.

Plainoldanalytics, posted to Hacker News on 13 September 2026, takes the opposite approach: a bolt-on analytics middleware for Go applications, similar in spirit to Umami, Plausible or PostHog but embedded in the app. Its core package is storage-agnostic, with backends including DuckDB and an in-memory store, a built-in dashboard mounted at a path of your choice, session recording, custom events and key-value properties. Traffic flushes to disk roughly once a second.

Neither project will tell you who won last night. But they are the layer that decides whether the number you are shown is the number you think it is.

The World Cup will keep producing arguments about calls, and the NFL will replace its chain gang with Hawk-Eye technology from September, as the Guardian reported. The screens will keep getting faster. The definitions underneath them are where the fight actually is.