Comments on: TrueHoop Network 2009-2010 Preview: In Defense of the Undersized guard

By: Crow

Crow — Wed, 28 Oct 2009 07:49:12 +0000

Alright Matt. I look forward to seeing what else you do on the issue.

If the small guards hurt you on defense argument had merit in general it would likely be in the Adjusted +/- … but I take it from what you said it isn’t there.

Still you could press the research further if you want.

I wonder what the differences between the (mostly team) Defensive Ratings on/off the court are by size.

And maybe you can protect a small guy by putting him with your best defenders? Do the small guys with good defending teammates more of the time do better on Defensive Adjusted +/- than those without? If so when you play the small guard and with whom becomes key.

It might also be that the overall Defensive Adjusted is ok for the small guards due a combination of being able to guard their own man well offsetting any shortfall in ability to handle switches (and guarding your own man being a more frequent event and therefore more important) and bigger guards having the opposite problem? Checking both the team and counterpart level stats would give you a chance to try understand these issues in a way the overall team defensive stats alone would not.

Finally it might be interesting to break guards down by both height and weight. If you are short and light it might be different than if you are short but of adequate weight and presumably strength. I’ve divided PGs into three piles before, maybe it was >180, 180-195 and above 195. But you could divide into just 2 piles say 185. Does being light in general hurt defense? Short and light? It would add an appropriate 2nd variable to the “small” discussion. On the surface small guys of adequate weight & strength would seem to have more a basis to hold their own on switches. But as always worth checking the numbers, if you have time & desire.

By: SM

SM — Tue, 27 Oct 2009 20:16:43 +0000

Matt, Lee’s comment about the selection bias is 100% correct and would need to be adjusted for to make a solid case on size in general. I also think it would be better to have an overall 2 guard baseline and compare production as a distance form that baseline.

By: Matt Moore

Matt Moore — Tue, 27 Oct 2009 18:51:40 +0000

Lee: I think part of the idea is that too often a player gets pushed out of a starting spot because coaches think they need size, when in reality, there’s not conclusive evidence that that’s the case. The evidence (and chart) certainly show that a MUCH smaller number of players reach the NBA contributor level who are smaller, so it weeds out those players in essence. My rebuttal would be that if you look at the vast majority that the middle height section takes up, it’s pretty clear that there’s not evidence to suggest they’re doing much better. The idea here is not that smaller guards are better, but that trying to say you should start a less talented tall guard at the 2 rather than a small guard is faulty thinking. The reason I started on this was because someone said “Yeah, he’s good, but who’s going to guard Kobe?” But the problem is… NO ONE’S GUARDING KOBE. IT’S KOBE FREAKING BRYANT. He’s going to score regardless. To that end, the evidence submitted here suggests, from a small and recent sample size, that taller guards are not performing better than smaller guards.

By: Matt Moore

Matt Moore — Tue, 27 Oct 2009 18:45:17 +0000

Crow,
Agreed on all counts. I used Off/Def because I need the best and easiest thing to look at with as few caveats as allowable, and also that included net overall team performance, because the idea is that small guards hurt your team through the lack of size.

For what’s it’s worth, I ran the analysis on both Adj. +/- as well as a number of other measures and we see the same result, with almost no correlation (PER was especially bad, but that’s only offense).

I’ll be revealing some further results in the coming weeks. Consider this the “for the masses” piece.

By: Crow

Crow — Tue, 27 Oct 2009 18:35:07 +0000

If you used the Offensive and Defensive ratings at Basketball Reference based on Dean Oliver’s formulas the Defensive rating is still mainly derived from team (on average 80% of the shots are being taken against other people) and isn’t that precise or good a measure of personal defensive performance. Counterpart PER and points has flaws too but I’d look at both and probably weight the counterpart data heavier.

Defensive rating is largely based on the idea the defense is a team thing and then it runs counter to that and gives the guy who made the steal or the block all the credit for that play- when it works- but only 1/5th the blame when going for a steal or block is unsuccessful and proves harmful. It is not the perfect answer.

Defensive Adjusted +/- may not be perfect either but it is probably better for this kind of study.

By: lee

lee — Tue, 27 Oct 2009 17:52:16 +0000

these stats are biased. most of the evaluation of “small” players takes place before they are drafted. the players whose stats are being used in this analysis (current NBA players) clearly were good enough to get the “too small” label removed. once you get below a certain level, i wonder if size becomes more of a factor (like second round/undrafted players). thoughts?

By: David

David — Tue, 27 Oct 2009 01:57:28 +0000

I realize this doesn’t have much to do with your post, but it drives me absolutely insane / confuses the heck out of me when I see you describe shooting guards as “small guards.” I think you’re talking about point guards (who are, in fact, small) when I read that.

Other than that, love your work =P

By: TrueHoop Network 09-10 NBA Season Preview

TrueHoop Network 09-10 NBA Season Preview — Mon, 26 Oct 2009 17:39:25 +0000

[...] bonus, Henry Abbott will have some more stuff up later in the day over at TrueHoop, Matt Moore has his piece on undersized guards at Hardwood Paroxysm and Haubs has his “10 NBA Questions for the 2010s” at The Painted [...]