Home Advantage, Explained

· 9/17/2026

"Home advantage" gets treated as a fixed fact of football — teams just do better at home, roughly the same amount everywhere. The real picture is messier and more interesting than that, and it's worth understanding because it's a real input into every prediction we publish.

It's real, but it isn't constant

Home teams score more than away teams on average, across essentially every league in the world. But how much more varies a lot — by league, by season, and over time. A league with tightly packed, similar-quality teams might show a small home edge; a league with long travel distances, hostile atmospheres, or big quality gaps between top and bottom clubs can show a much bigger one. Treating home advantage as one universal number flattens out real differences between competitions.

How we actually measure it

Instead of assuming a fixed boost (the textbook figure often cited is around 30%), we calculate a home-advantage factor per league directly from that league's own finished results this season: the ratio of average goals scored by home teams to average goals scored by away teams. If home teams in a league are, on average, outscoring away teams by a wide margin, that shows up as a larger factor for that league. If a league is genuinely balanced, the factor sits close to neutral.

Why this matters for a prediction

This factor feeds directly into each team's expected goals for a match, scaled to the specific league the fixture is being played in. A team gets its home boost calculated against what home advantage has actually looked like in its own competition this season — not against a number borrowed from a completely different league with different crowds, pitches, and travel patterns.

What we don't do

We don't invent a home advantage figure for a league that hasn't played enough games yet to measure one. Early in a season, or for a competition with too little finished-match data, the model uses a neutral, no-adjustment value rather than guess — an honest "we don't know yet" instead of a plausible-looking number with nothing behind it.