looking at OPR across events

[Lil' Lavery]

Originally Posted by Michael Hill

I think OPR is as worthless statistic in this game. It works ok games where there is a lot of individual contribution, not so much with teamwork. 2013 was a great year for OPR. 2012 wasn't so bad either, but 2014 is not a good year for OPR.

Ideally, this is exactly the type of game you would want to use a OPR or CCWM metric for. With a variety of potential roles within an alliance, a team's contribution to offense or winning a match isn't always obvious to more traditional scouting. However, I don't think a single event comes anywhere near the sample size required to normalize the data, especially given the alliance driven nature of the game. Games with very discreet roles are much easier to get a normalized/accurate OPR, but OPR is less useful in those years because they're much easier to get meaningful data from scouting. District model teams, or any teams who compete in 3+ events and 50+ matches, may have enough input data for OPR to be useful, but in-season improvement (of both them and the average level of play) once again throws a wrench in how predictive OPR will be for future contributions.