Quote:
Originally Posted by TwistedChief
Well, first, yes, his average/catch would be 9.9. That’s a fact.
“And that’s why you get rid of the outlier data points.” Uhm, wrong. Because who defines outlier? Did Pachecho rush for 12/95yds (7.9ypc) or do we just take out his 39yd run and then it’s 4.7ypc? What if it were a 29yd run? Would that then be okay to add in?
I guess Barry Sanders was just a really average running back if you just took out all of his explosive plays. That’s definitely the right way to look at it.
|
Lol, nope. We did that in a class. Didn't change his average rush yds/att by that much. The thousands of reps marginalize the few dozen or whatever big rushes. Leave them in, take them out, it's roughly the same.
When looking at small datasets, taking out the outliers is SOP. Sorry, but if you think player X in my example averaged 9.9yd/cth, well, you don't know what you're talking about.