How can per-48 statistics be applied to ANYTHING, they are theoretical and only exist in a mathematical fantasy land. They are statistics that can NEVER be achieved.
Okay, I'll try again.
Per-48 stats are just the same as per-minute stats. The only difference is a factor of 48. If Player A averages 18 rebounds "per 48 minutes," that doesn't mean that he is "potentially" someone who could actually average 18 rebounds per game. That isn't the point at all.
What it means is, if Player B averages only 13 rebounds per 48, then if each player were to be on the floor the same amount of time, Player A would be expected to collect more rebounds. That's all.
If you want, think of it like a batting average. What does it really mean if someone hits .322? Technically, it means that if he had 1000 at-bats over the season, he would be expected to get 322 hits. Of course he will never get 1000 at-bats in a season, but the number is still meaningful. He collects hits at a "rate" of 322 per 1000, no matter how many at-bats he winds up with. 500 at-bats means 161 hits. And so on.
Imagine if baseball had a statistic called "hits per game." You could have a terrible hitter, who has to play regularly because the team doesn't have a decent second baseman (or whatever), and he'd be a .220 hitter with about 1 "hit per game." Then there'd be the guy who's your fourth outfielder and top pinch hitter, and he wouldn't get nearly as many plate appearances, so he might be a .290 hitter with, whatever, 0.4 "hits per game." The "hits per game" statistic wouldn't tell you as much about the quality of his hitting as his batting average would.
That's the problem with looking at "per game" statistics for part-time NBA players. Obviously someone will get bigger total numbers if he plays starter's minutes. But there are bench players who are very effective during the time they get.
So that is the use of per-48 stats. It's not that someone will play 48 minutes in a game. It's a rate. If he gets 20 rebounds per 48 minutes, then that means 5 rebounds in 12 minutes, or 7.5 rebounds in 18 minutes, or 10 rebounds in 24 minutes. Whatever. If you don't like calling it 20 rebounds per 48 minutes, then you can call it 0.43 rebounds per minute. It's the same thing.
With per-48 stats, sometimes you'll have players whose numbers are inflated by garbage time, and sometimes you'll have players whose stamina (or propensity for fouls, or whatever) simply doesn't allow them to play many minutes -- like Arvydas Sabonis, who is fantastic per-minute (or per-48) player. So it's far from a perfect indicator. No statistic is.
But, in my opinion, when you're talking about Mutombo, it's still valid. Mutombo could play 30-35 minutes for any team in the league right now and do just fine. The Nets are so much of an open-court team, they can't use him, and you're right that his numbers are down since he came back from his injury.
That, however, was not the topic of discussion. Hcsilla said that any Suns center could put up similar numbers to Mutombo "if given the time." The only way to evaluate that claim statistically is to look at per-minute (or per-48) figures.
Again, if you want to reject the statistic, go ahead. There are a lot of statistics out there, and depending on how you want to look at the game, not all of them are meaningful.
