Expected goals (xG) measure the likelihood of a shot resulting in a goal based on several factors, including:
Each shot is assigned an xG value between 0 and 1, where 1 represents a 100% chance of scoring. For example:
By aggregating xG values for a team over a match or season, you can evaluate their attacking and defensive strengths.
xG offers a more objective way to assess team performance than traditional statistics like possession or shots on target. Here’s why it’s invaluable for predicting correct scores:
xG provides insights into how many goals a team should score or concede based on the quality of chances, not just the actual goals.
Example: A team may win 1-0 but have an xG of 0.3, indicating they were lucky and not likely to repeat the result consistently.
By analyzing a team’s average xG per game, you can estimate how many goals they’re likely to score or concede in an upcoming match.
Example: If Team A’s average xG is 1.8 and Team B’s average xG conceded is 1.5, the likelihood of Team A scoring 2 goals is high.
xG allows for adjustments based on opposition strength, playing style, and recent form.
Example: A strong defensive team with a low xG against might limit high-scoring matches, making low correct scores (e.g., 1-0 or 0-0) more likely.
To leverage xG effectively, follow these steps:
Attack: Look at the average xG created per game.
Defense: Examine the average xG conceded per game.
Tools like FBref, Understat, and Wyscout can provide detailed xG statistics for teams and players.
Calculate the expected number of goals for both teams based on their xG and xG against.
Example: If Team A’s xG is 2.0 and Team B’s xG against is 1.5, Team A might score 2 goals. Similarly, if Team B’s xG is 1.2 and Team A’s xG against is 0.8, Team B might score 1 goal. Predicted score: 2-1.
Consider the league’s average goals per game to fine-tune predictions.
Example: If the league average is 2.6 goals per match, scores like 1-1 or 3-1 are more probable than 4-3.
Adjust predictions for factors like:
Let’s apply xG to an example match:
Using these metrics, we can predict:
While xG is a powerful tool, it has limitations:
For best results, combine xG with other metrics and qualitative insights.