Generate custom courses on any topic — with hands-on practice, AI guidance, and visuals built in.
Already have an account?
Pick any World Cup 2026 match and the first weird thing you will notice is that the final score can look clean while the chances underneath were messy. xG means expected goals. It is the total of shot probabilities, where each shot gets a value between 0 and 1 based on how often shots like it become goals. The point is not to replace watching. The point is to separate finishing from chance creation fast. Use the explorer to look at score, total xG, non-penalty xG, shots, xG/shot which is average chance quality, and a shot map which shows where shots came from.
xG is a model of shot difficulty, not a rating of a player’s talent and not a verdict on who played beautifully. A shot’s xG usually rises when it is closer, more central, and struck with a better body part. It also depends on how the chance was created, like a through ball versus a long cross. What xG does not see well is everything before the shot, like the pass that breaks the line, the run that drags a defender, or a team that never shoots because it cannot enter the box. Use the simulator to change one shot feature at a time and watch the xG value respond.
Model, not truth xG is an estimate from past shots, so it summarizes patterns rather than proving what should have happened.
In a single game, the same total xG can come from totally different stories. A team that scores early may take safer shots and allow low quality attempts, while the team chasing may rack up many low probability shots from distance. That is why xG is not a deserved win meter. It is a chance quality ledger. Use the comparison to see two matches with similar xG but different game state, meaning whether a team is leading, level, or trailing.