Expected goals, usually written as xG, has moved from analyst blogs to television broadcasts because it answers a better question than shots or possession: how good were the chances? Each shot is given a value between 0 and 1 based on how often similar shots in the past became goals. A tap-in from six yards might be worth 0.70, while a hopeful strike from thirty yards might be worth 0.03. Add up every shot and you get a team's xG for the match, a measure of chance quality rather than chance volume. For bettors, that distinction is the whole point, because scorelines lie far more often than chance quality does.

Models weigh several inputs when scoring a shot: distance from goal, angle, body part used, the type of pass that created the chance, defensive pressure, and whether it came from open play or a set piece. A penalty is usually rated around 0.76 to 0.79, which is why penalty-heavy teams can post flattering goal totals without creating much from open play. Different providers build different models, so 1.4 xG from one source will not always equal 1.4 xG from another. The practical rule is to compare figures from the same source over time, and to treat non-penalty xG separately when you want a cleaner picture of open-play attacking quality.

The signal that matters is rolling xG over five to ten matches, never a single game. One match tells you almost nothing: a side can post 2.5 xG and lose 1-0, or post 0.4 xG and win 2-0, and neither result rewrites what the team is. Over eight to ten matches, though, a pattern emerges. A team averaging 1.8 xG for and 0.9 against is genuinely strong, even if recent results were unlucky. A team scoring freely from 1.0 xG per game is probably overperforming its chances and may be due a correction the market has not yet priced. Split home and away figures where you can, because many sides create far better chances at home than on the road.

To turn xG into a betting assessment, combine each team's attack with the opponent's defence. A simple method is to average one side's xG created with the other side's xG conceded, giving a rough expected goals figure for each team in the fixture. From there, totals and both-teams-to-score markets become easier to judge: if the combined expectation is around 2.8 goals, Over 2.5 may be worth a look at the right price, while a combined expectation near 1.8 points the other way. Convert the bookmaker's odds to implied probability and compare. If your estimate puts an outcome higher than the implied probability, there may be an edge; if not, pass, however attractive the match looks.

Finally, respect what xG cannot tell you. It does not know about a missing striker, a goalkeeper in poor form, a tactical change, or a team resting players before a bigger game, so always check team news alongside the numbers. It also describes past performance, not the next result, and it carries a margin of error of roughly 0.2 xG per team per match. Use it as one layer in a wider assessment, keep stakes small when the edge comes from a model gap rather than confirmed news, and never treat xG as a prediction. Bet responsibly, within a budget you set before the weekend begins.

A weekly xG routine need not be complicated: update rolling five-to-ten-match figures for the teams you follow, split them home and away, note any side whose results are running well ahead of or behind its chance quality, and flag fixtures where the market price looks out of line with the combined expectation. Then, and only then, check team news, because a missing striker or a tactical change can move an expectation more than a whole month of xG data. Keep the stakes small when the edge comes from the model alone, record the bets and their xG context, and review after fifty or more bets whether the model is genuinely adding value or simply adding confidence to decisions you would have made anyway.