Expected Assists (xA) extends the expected goals framework to measure the quality of chances created by a player or team, regardless of whether those chances were converted. While xG evaluates the shot-taker, xA evaluates the chance-creator, providing insight into a team's creative quality that is independent of finishing skill or luck.

How xA Is Calculated

Expected Assists assigns a value to each pass or action that leads to a shot, based on the xG value of the resulting shot opportunity. If a through ball creates a one-on-one chance worth 0.45 xG, the passer receives 0.45 xA regardless of whether the shot is scored, saved, or missed entirely. This separation of creation from finishing allows analysts to evaluate playmaking quality independently — a crucial distinction when assessing team and player performance.

xA as a Prediction Feature

Teams with high xA are creating quality chances consistently, which is a more sustainable source of goals than individual moments of brilliance. Our AI models use team-level xA as a key input feature, with higher xA indicating a team's structural attacking quality. When a team's actual assists significantly exceed their xA, it suggests that their chance creation may be less prolific than their goal output suggests — a potential warning sign for future performance regression.

Identifying Creative Quality

xA helps identify teams and players whose creative contributions are undervalued by traditional statistics. A midfielder who consistently creates high-xA chances may not top the traditional assist charts if teammates fail to convert those chances, but their creative output indicates a team that is generating quality opportunities. Our models recognize this distinction, using xA alongside xG to build a complete picture of each team's attacking capability.

Combining xG and xA for Complete Analysis

The combination of xG and xA data provides a comprehensive view of team attacking quality. A team with high xG but low xA might be generating shots from individual skill rather than collective play — a pattern that is less sustainable. A team with balanced xG and xA is creating chances through structured teamwork, which tends to be more consistent over time. Our AI models analyze both metrics together to assess the sustainability and reliability of each team's attacking output.