Doctor MJ wrote:Hmm.
First and foremost, what keeps getting said again and again and yet people never seem to get it: The issue with +/- is one of reliability. There's a lot of noise involved. This says nothing about the stats' validity, and reliability always improves with sample size. So when we take a multi-year sample, we're giving the stat good reliability to go with the validity.
In other words: We're going away from one year samples for a reason.
but all-defensive teams are only for the one season. if they're noisy why are using them to act like his all-d teams a joke? how do you know kobe's D in 2008 was actually deserving based on+/-, but say 2007 not? shouldn't u be using more defensive stats than just +/- then?
also, its funny how you always brought up kobe's poor single-season +/- in 2011 as a basis to drop him in MVP voting even when numerous posters brought up the noise of that stat. but i guess it doesn't matter to further a agenda...
Second, looking at RAPM, here's how Kobe's defensive numbers look going back to '02-03 (which is all we have):
Year Score
'11 -1.0
'10 +1.2
'09 +0.8
'08 +1.0
'07 -0.2
'06 -0.9
'05 -1.1
'04 -0.6
'03 +0.1
There isn't a single year in the bunch that would warrant All-Defensive selection if we used that metric. It's actually kind of remarkable.
and again, if you use +/-, 2008 and 2010 are deserving.
you're also missing his best defensive years--2000,2001, 2002
exactly what SG was playing kobe-type minutes and did better? what SG had a better combo of +/-, defensive PER, NET ,"eye -test", versatility etc at comparable minutes?
that's right. none!
and if +/- is so noisy, you probably shouldn't be using its as a single judge for yearly awards, especially for kobe who's defense clearly has peaks and valleys depending on the year in question.