Thank you for this, thank you. You've talked in your videos about the importance of understanding math and statistics. And it's really nice how you broke down this idea and explained it in a simple way.
Something that I've found with building my own xG model is that many data points need to be scaled by distance to be useful. The best example of this would be headed vs kicked shots. Overall, they have similar rates of success, and headers may actually be more likely to score. But at 10m, or 20m, or any specific distance, shots taken with the foot are more successful than headers.
Based on the xG formula variables, it seems that it could help referees to determine whether a foul is considered to stop a promising attack (SPA) or denying a goal score opportunity (DOGSO). With the current technology provided in most of the top leagues, do you know if this information would be available to the referee crew immediately or it will still need considerable processing time to collect the stats for all the formula members to calculate the xG on a specific play?
Thank you for this, thank you. You've talked in your videos about the importance of understanding math and statistics. And it's really nice how you broke down this idea and explained it in a simple way.
Thank you! Glad you enjoyed it
Something that I've found with building my own xG model is that many data points need to be scaled by distance to be useful. The best example of this would be headed vs kicked shots. Overall, they have similar rates of success, and headers may actually be more likely to score. But at 10m, or 20m, or any specific distance, shots taken with the foot are more successful than headers.
Based on the xG formula variables, it seems that it could help referees to determine whether a foul is considered to stop a promising attack (SPA) or denying a goal score opportunity (DOGSO). With the current technology provided in most of the top leagues, do you know if this information would be available to the referee crew immediately or it will still need considerable processing time to collect the stats for all the formula members to calculate the xG on a specific play?
Possibly it would be difficult in my opinion to deliver that information in a reliable way