What is xG in hockey?

Published on 20 Nov 2023 | Last updated on 01 Jul 2026 | Reading time: 6 minutes

What is xG in hockey?

While expected goals (xG) is more commonly associated with football (soccer), a similar concept is increasingly being applied to hockey to provide a quantitative measure of the quality of scoring chances. In hockey analytics, this concept is often referred to as "Expected Goals For" (xGF) or "Expected Goals Against" (xGA). The goal of applying xG to hockey is to provide a more comprehensive assessment of a team's offensive and defensive performance beyond traditional statistics like goals scored and goals against. This analytical approach can contribute to a deeper understanding of a team's strengths and weaknesses and help in strategic decision-making for coaches, managers, and of course punters.

How is xG in hockey calculated?

The calculation of Expected Goals (xG) in hockey involves assessing various factors associated with a shot attempt to estimate the likelihood of that attempt resulting in a goal. The exact formula can vary between different analysts, models, and platforms, but here are some common factors considered in the calculation of xG in hockey. The first three are more commonly used and the rest are dependent on the specific model used:

  1. Shot Location:

    • The position on the ice from which the shot is taken is a crucial factor. Shots from high-scoring areas, such as in front of the net or close to the goaltender, are assigned higher xG values.

  2. Shot Type:

    • Different shot types have different conversion rates, so analysts may assign varying xG values based on whether the shot is a wrist shot, slap shot, deflection, etc.

  3. Angle of the Shot:

    • Similar to soccer, the angle at which the shot is taken is considered. Shots from more favorable angles are assigned higher xG values.

  4. Defensive Pressure:

    • The level of defensive pressure on the shooter may be factored into the xG calculation. A shot taken under high defensive pressure might be assigned a lower xG value.

  5. Rebound Opportunities:

    • Shots that create rebound opportunities may be given additional consideration, as rebounds often lead to high-quality scoring chances.

  6. Power Play and Penalty Kill Adjustments:

    • Adjustments may be made for shots taken during power play or penalty kill situations, as the dynamics of the game change in these scenarios.

  7. Historical Data:

    • xG models are often trained on historical data to understand the relationship between various factors and the likelihood of a goal being scored. Machine learning techniques, such as logistic regression, may be employed to develop these models.

  8. Goaltender-Specific Factors:

    • The historical performance of the goaltender facing the shot, such as their save percentage, may be considered in the xG calculation.

  9. Passing Sequences:

    • The quality of the play leading to the shot, including passing sequences and assists, may influence the xG value.

It's important to note that the precise formula and the weightings assigned to each factor can vary, and some models may be proprietary. Analysts often fine-tune their models based on extensive datasets and use advanced statistical techniques to improve the accuracy of xG predictions. The goal is to provide a more nuanced understanding of a team's offensive and defensive performance beyond traditional statistics.

How to take advantage of xG in hockey predictions?

Expected Goals (xG) in hockey can be used for predictions by assessing team and player performance based on the quality of scoring chances created and faced. Here's how you can utilize xG in hockey predictions:

  1. Team Analysis:

    • Evaluate teams based on their xG metrics over a series of games. Teams consistently generating higher xG values may be considered stronger offensively, while those with lower xG against values may have a solid defense.

  2. Comparison of Teams:

    • Compare the xG metrics of two competing teams to gauge their relative strengths and weaknesses. A team with a higher xG difference (xG for minus xG against) may be expected to perform better in upcoming matches.

  3. Player-Specific Analysis:

    • Assess individual player performance by examining their contribution to xG. Players who consistently contribute to creating high xG opportunities may be key to a team's offensive success.

  4. Goaltender Performance:

    • Consider the historical save percentage of goaltenders to understand their effectiveness in stopping shots. Teams facing goaltenders with lower save percentages might be expected to have higher xG values.

  5. Power Play and Penalty Kill Predictions:

    • Use xG to assess the effectiveness of teams during power play and penalty kill situations. Teams with a high xG for during power plays or a low xG against during penalty kills may have an advantage in these scenarios.

  6. In-Game Situations:

    • Consider the context of the game, such as scoreline, time remaining, and special teams situations. Teams trailing in the score might be expected to generate higher xG values as they push for goals.

  7. Home/Away Performance:

    • Assess how teams perform at home versus away using xG. Some teams may have a higher xG at home, which could influence predictions for their home games.

  8. Lineup Changes and Injuries:

    • Take into account any recent lineup changes or injuries, as they can impact a team's xG performance. Losing key offensive players or having a star goaltender injured may affect the xG dynamics.

  9. Game-By-Game Analysis:

    • Analyze xG on a game-by-game basis, considering recent form, head-to-head records, and other contextual factors to make more informed predictions.

As with any predictive metric, it's important to use xG as part of a broader analysis and consider other factors that may influence the outcome of hockey games. Additionally, hockey is a dynamic and fast-paced sport, and unexpected events can significantly impact the results.

Example how to use xG in hockey predictions

Let's consider a hypothetical example to illustrate how expected goals (xG) might be applied in a hockey game:

Scenario:

Suppose there's an upcoming match between Team X and Team Y in a professional hockey league. Analysts have been tracking the xG metrics for both teams over the past several games.

Team X:

  • Team X has been consistently generating high xG values in recent games, indicating strong offensive performance.
  • Their top forward, Player A, has been contributing significantly to the xG with precise shots and playmaking.

Team Y:

  • Team Y has a solid defense, as reflected in their low xG against values.
  • The goaltender for Team Y has a high save percentage, suggesting effective goaltending.

Previous Head-to-Head Meetings:

  • In their previous head-to-head meetings, Team X has generally had higher xG values, but Team Y has managed to secure some victories through strong defensive play and effective goaltending.

Injury Update:

  • A key defenseman for Team X is currently injured and will not be playing in the upcoming game.

Prediction: Based on the xG analysis and considering the context:

  • Team X might be expected to have a higher xG in the upcoming game due to their strong offensive performance.
  • However, the absence of the key defenseman could impact Team X's defensive stability, potentially leading to a higher xG against.
  • Team Y, with its solid defense and effective goaltending, may be well-positioned to capitalize on counterattacks and defensive plays.
  • The outcome could be influenced by factors such as special teams (power play and penalty kill), individual player performances, and the overall flow of the game.

This is a simplified example, and in reality, various additional factors would be considered in a more detailed analysis. The use of xG provides a quantitative measure to inform predictions, but it's important to integrate this information with other relevant aspects of the game for a comprehensive understanding of the potential outcomes.

Conclusion

In conclusion, expected goals (xG) is a valuable metric that can be applied to hockey to assess the quality of scoring chances and predict game outcomes. By analyzing xG metrics for teams and players, analysts can gain insights into offensive and defensive strengths, individual player contributions, and potential outcomes of upcoming matches. However, it's crucial to consider xG as part of a broader analysis, taking into account contextual factors, recent form, special team performance, injuries, and other game dynamics.

Real-world examples, such as evaluating Team X and Team Y in a hypothetical scenario, demonstrate how xG can be used to make informed predictions. While xG provides a quantitative measure of scoring chances, the dynamic and unpredictable nature of hockey means that other factors, beyond statistical metrics, also play a significant role in determining game outcomes.

As the field of sports analytics continues to evolve, the integration of metrics like xG contributes to a deeper understanding of the game, enabling teams, coaches, and analysts to make more informed decisions and predictions.