At the risk of repetition ...
There are 200 minutes played in each regulation game, 5 players on the floor times 40 minutes. Thus, for any single game, the total is exactly 200. However, when one looks at multiple games, it is possible that at least one player on a roster of more than 5 does not play at all in one game or more. That game does not count in calculating that player's minutes per game. As a result, summing the average minutes per game across the entire team yields a number greater than 200. The math that does work exactly is: the sum of (minutes per game times number of games played) for each player equals 200 times the number of games.
When one is predicting the minutes per game of players, as we are here, there are 2 approaches one can take--force the total to exactly 200 or let it be slightly higher. If one forces the total to 200, it is assumed that no players miss any games or, alternatively, it's impossible to know who might miss games, so no accounting of that is taken. The second approach--letting the sum exceed 200 by a small margin (perhaps up to 15 or so)--actually is more consistent with empirical data from past years.
If folks want average minutes to sum to 200, they can scale someone else's numbers down from 210 or 215 to exactly 200. The relative numbers are the same. In reality, there is very little absolute difference. And that difference is probably much smaller than the error in the prediction.
Peace!