You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
If you take a given play and then look at plays that happened at a similar game state, you should be able to construct a distribution of WPA from that game state (you could imagine this like a markov process, or just binning up nearest neighbors and averaging). Once you have that distribution, you could first compute the WPA, then convert that into a Z-score. Overall this is probably not a huge change, but there could be some issues with unusual distributions in situations near the end of the game. Not sure if it will make a difference or not, but it's something I can keep an eye on.
The text was updated successfully, but these errors were encountered:
If you take a given play and then look at plays that happened at a similar game state, you should be able to construct a distribution of WPA from that game state (you could imagine this like a markov process, or just binning up nearest neighbors and averaging). Once you have that distribution, you could first compute the WPA, then convert that into a Z-score. Overall this is probably not a huge change, but there could be some issues with unusual distributions in situations near the end of the game. Not sure if it will make a difference or not, but it's something I can keep an eye on.
The text was updated successfully, but these errors were encountered: