Why football stats on TV don't always add up

The number on screen is a guess dressed as a fact

Every big broadcaster leans on a data provider to generate the numbers that appear during a match: possession, expected goals, distance covered, sprint counts. Those figures are not measured with the precision the graphic implies. They come from tracking systems and event-logging teams working in real time, often assigning judgement calls, like whether a shot was blocked or off target, whether a pass was intended for the man who received it, whether a tackle was a foul or a clean challenge. Those calls get made in seconds, live, by humans watching a feed. Mistakes happen constantly, and most get quietly corrected after the final whistle, by which point nobody in the stadium or on the sofa is paying attention any more.

Different providers, different numbers

A big source of confusion is that broadcasters and clubs don't always use the same data company. Sky Sports might display one expected goals figure, a rival broadcast or a club's own website might show another, and both can be defensible. xG models vary depending on which shots a provider weights more heavily, how they treat headers versus footed shots, and how much they factor in defensive pressure. None of that makes either number fake, but it does mean a single television graphic is one interpretation, not a settled scientific reading. Fans who screenshot two different broadcasters' stats for the same match and treat the mismatch as proof of an error are usually just looking at two different models.

Live graphics break in more mundane ways too

Separate from the analytics layer, there's the plain mechanical stuff: on-screen graphics packages relying on manual input from production staff, timing boards, substitution boards and other in-stadium tech that occasionally glitches under matchday conditions. A malfunctioning electronic board can throw off a manager's planned changes or leave a broadcast graphic frozen on old information, and when it happens it isn't a data inaccuracy in the analytics sense. It's the same underlying truth though: the technology layered over a football match is running live, under pressure, operated by people, and it breaks.

Why the errors rarely get corrected on air

Once a stat is wrong on screen, there's little incentive to go back and fix it live. Broadcasters are producing a fast-moving show, not compiling a reference document, and a wrong distance-covered figure or a slightly-off possession split doesn't change the outcome of the match being shown. The correction, if it happens, tends to arrive in the underlying data feed rather than back on the broadcast, which is why post-match stat pages and databases often look different to what viewers saw during the game itself. That gap between the live number and the eventual settled number is the single biggest reason fans feel misled: they are comparing two different snapshots of the same evolving dataset.

Useful shorthand, not forensic evidence

None of this means broadcast stats are worthless. Distance covered, expected goals and possession are still useful shorthand for what happened in a match, and the big swings, a team dominating territory, a striker racking up huge underlying numbers, are almost always real. What isn't reliable is treating any single figure as forensic. Live football data is provisional by nature: assembled fast, revised later, and dependent on which provider is doing the counting. Sky Sports isn't uniquely careless here, it's simply the most visible face of an industry-wide process built for speed and entertainment first, precision second.