Perhaps the BEST B1G Forum anywhere, here at College Football Fan Site, CFB51!!!

my point is simple

JUST adding a few more carries per game or even doubling the carries doesn't assure the average per rush will go down.  It's not a mathematical certainty.

there are many variables

But a strong trend is not undone by 1 or 2 exceptions.  Regression to the mean as volume increases is a statistically valid thing.  Yes, there are tons of variables, but even still, it's valid.
And I'd like some evidence on the John Riggins claim before I believe it.  All he had to do was break a long run in a few different 4th quarters to get a reputation like that.  I'd be stunned if the stats backed it up.

The way you're comparing season to season, you're allowing the variables to dominate the data.  Why would a RB with a different OL blocking for him and a different OC calling plays provide valid data to compare???  Plus, when bringing up an NFL career, the normal productivity aging bell curve to boot.

Look within a season for one team.  Just use that to wrap your head around the idea before you start with multi-year, multi-team stuff.

This is more macro-YPC analysis here.  No one in college football history averaged over 8.2 ypc with 400+ carries.  And no one averaged over 7.8 ypc with 700+ carries.  Do you see the trend line?  The carries go up, the highest ypc average goes down.  It is a trend line - a mostly-horizontal, slanted line.  There would be dots all around it, but the center shows a definite trend.  

Now, if you look at the above post, you can pick one guy out - let's say Ricky Williams.  Just as an example.  He has the highest ypc average for any RB with 800+ carries.  Great.  Now, if you wanted to make the claim he was better than Archie Griffin, you'd suggest that with 100 more carries, Williams would maintain a ypc lead on Griffin WHILE ACKNOWLEDGING it would decrease from 6.3.  His ypc would not magically have increased with more carries during his career.  And even if his ypc only dropped to 6.2, you could still claim him superior to Griffin, citing his ypc advantage (6.2 to 6.0, given the same number of carries (900+)).  

Now, these 100-even carry lines are totally arbitrary, and we may get different players listed if we used 350+, 450+, etc.  That doesn't matter, the point still holds true.  The player with the highest YPC ever at 350+ carries WILL 100% HAVE A HIGHER YPC AVERAGE THAN THE RB WITH THE HIGHEST YPC at 850+ CARRIES.

And THAT is the point.