This is exactly why you need a statistician. I think if you look at the last 16 games of teams their winning percentage might be very close to what it is the first 66 times on average. I just don't know to be sure one way or the other.
There will be a degree of variability but what's just as much of an issue is how much of an impact that variability has in terms of a percentage. 72-10 is practically 90%. And so, you can't really afford to deviate from 90% much if you lose a game or two off of your previous pace during the last 16 games. Additionally, the variability is problematic in that, as stated, 90% is a ridiculously high win rate and the odds of the minus occurring is greater than the plus occuring. In other words, winning less than 14 of the last 16 games is a much more likely occurrence than winning 15 or 16 of 16 games.
As I've said before. Nevermind injuries. The wider you expand the scope of the season, the more you increase the possibility of a team going through a "slump". Look at what happened to the Spurs last year. They had a high win % well into the season and then really fell off at the end.
All statistics would do is project a degree of variability based on the 66 games. This is hardly fool proof even with "statistical validity". And again, you have to look at the variability in terms of winning % too. When you're winning 87% of your games, it's much easier for the % to come down than it is to move it up.
