If you have watched football in the last few years, you have heard commentators say xG. It appears on broadcasts, in match reports and in every online argument about whether a team deserved to win. Yet most fans have never had it explained properly. Here is the plain version: expected goals measures the quality of a chance, not the number of shots.

Expected goals, or xG, is a single number attached to a single shot: the probability that a shot taken from that position, in that situation, becomes a goal. It runs from 0 to 1. A tap-in from two yards might be worth 0.9 xG, meaning nine times out of ten it is a goal. A hopeful effort from thirty yards might be worth 0.03, meaning you would expect it to go in about once every thirty-three attempts. A penalty, as a fixed reference point, is typically worth around 0.76 to 0.79 xG, because penalties are scored roughly three times in four.

xG models are built from historical data on thousands and thousands of shots. Data scientists look at every shot ever recorded and measure the factors that determine whether it goes in: distance from goal, angle to goal, body part used, type of assist, whether it came from open play or a set piece, and how much defensive pressure the shooter was under. More advanced models add goalkeeper positioning, the speed of the attack, and whether the shot was first-time. The model then compares a new shot to similar historical shots and assigns it a probability.

Add up every shot a team takes in a match and you get the team's xG: the number of goals an average finisher would have scored from those chances. If a team finishes a match with 2.40 xG, then based on the quality of its chances, an average team would be expected to score between two and three goals. This is why xG tells a different story from the scoreline. A 1-0 win can hide a team that was out-created, and a 0-0 draw can hide a team that missed chance after chance.

The 2022 World Cup final is a useful illustration. Argentina and France drew 3-3 before Argentina won on penalties, but the underlying numbers were less of a coin-flip than the scoreline suggested. Argentina out-created France 2.76 to 2.27 in expected goals and out-shot them two to one. France's evening was a monument to finishing: three goals, including two penalties, from 2.27 xG and ten shots. The single biggest chance of the match was a penalty worth 0.78 xG. That is xG in one game: not a verdict, but a fuller account of the story than the final score.

xG has limits, and honest analysts say so. Different models give different values for the same shot, because they do not all use the same factors. A single match is a small sample, so good or bad luck plays a major role. And xG says nothing about the quality of the finish itself: a brilliant save or a terrible miss looks the same in the pre-shot number. Post-shot xG exists to capture that difference, measuring where the shot actually went, but the basic number you see on television is pre-shot.

Used sensibly, xG is a powerful tool for watching football. Compare a team's xG over several matches rather than one game, and you get a sense of whether its attack is genuinely creating good chances or just riding a hot streak. If a striker keeps scoring more than his xG, he is finishing well or due a dry spell. If a team keeps losing while winning the xG battle, it may be unlucky, or it may have a finishing problem. Either way, you are now arguing about the game with better evidence than the scoreline alone.