Hi, I am currently trying to figure out team and individual ratings per Dean Oliver's book but I am having trouble understanding the correct way of common usage of those ratings vs. possessions. The thing is that in a game each team uses a different number of possessions which leads to a different team's off and def ratings so one team's def rating doesn't match other team's off rating which is understandable but if I take a look at a box score for a random game at basketball-reference site the pace rating (which is same as possessions) consists of average of both team's possessions which then creates a match between both team's off and def ratings.
So now, my question is, which way is more correct and should be used when calculating team's ratings? Why would you calculate an average of possessions of both teams when you create an error in ratings with this approach that can be pretty significant with just say four poss difference? Thanks for the answer.