Adding a comment here because I had more thoughts after I sent the post. This can also be useful for self-scouting. In the scatterplot, consider UMass and Tennessee. UMass's estimated effect is better than their raw explosive rate. That means that they weren't getting a lot of "explosive-friendly" situations and still were creating big plays. Tennessee, on the other hand, had a higher explosive rate than their estimated effect, suggesting they had more "explosive-friendly" situations but weren't capitalizing at much. UMass can then ask themselves: "how do we set ourselves up to highlight our explosiveness more?", while Tenn can ask "what can we do to capitalize on those big play situations?"
Awesome insight from top to bottom. This may be one too many layers on the cake, but would it be possible to determine the point at which a team's Explosive Rate Over Expected tends to stabilize throughout the course of the season, similar to what you did with the reliability score and success rate? Would you expect it to be about 5 games / 350 plays like it is for SRate?
Adding a comment here because I had more thoughts after I sent the post. This can also be useful for self-scouting. In the scatterplot, consider UMass and Tennessee. UMass's estimated effect is better than their raw explosive rate. That means that they weren't getting a lot of "explosive-friendly" situations and still were creating big plays. Tennessee, on the other hand, had a higher explosive rate than their estimated effect, suggesting they had more "explosive-friendly" situations but weren't capitalizing at much. UMass can then ask themselves: "how do we set ourselves up to highlight our explosiveness more?", while Tenn can ask "what can we do to capitalize on those big play situations?"
Great stuff!
Awesome insight from top to bottom. This may be one too many layers on the cake, but would it be possible to determine the point at which a team's Explosive Rate Over Expected tends to stabilize throughout the course of the season, similar to what you did with the reliability score and success rate? Would you expect it to be about 5 games / 350 plays like it is for SRate?