Basketball Prospectus blasts Hansbrough's NPOYs

[BigTedder]

article

Thoughts?

sooooooo, does that mean he thinks Love is better? I couldnt tell...the article was so subtle.

[sandinmyshoes]

I think the fact that a substantial number of Duke fans are defending the choice of a UNC player as NPOY is more telling by far than and contortion of stats.

[Duvall]

I think the fact that a substantial number of Duke fans are defending the choice of a UNC player as NPOY is more telling by far than and contortion of stats.

Yeah, it tells us that they've watched thirty UNC games this year.

[mus074]

seems like a silly argument to me. 18 & 10 ain't 24 & 10 and the other stuff is a wash, and if anything is in favor of hans: (i.e. they both play for top teams so thats a wash, but hansbrough stepped up his game when lawson went out and carried the team...add in the fact that hansbrough is a junior and has done this for 3 straight seasons now, and its not even close)

the real argument is between Hansbrough and Beasley, and I would have given it to beasley a month ago, but the way hans carried his team w/o lawson, I have to give it to Hans now...

The whole point of tempo-free stats is to take the number of possessions out of the equation. UNC plays about 74.5 possessions per game and UCLA 65. That's why Love's %age numbers are at least as good, if not better. I agree that the numbers favor Love.

[moonpie23]

All of us get to watch Tyler a lot more than other folks around the nation. I've tried to catch as many UCLA, Kansas, KS, Memphis, Tx, Louisville, as possible. Of course, I don't see as many of the other players as I do ACC players, but I've got to give it to Tyler. (cleaning the yack off my lip after saying that)


However, Hansbrough will be a buried "interesting stat" if he can't make a final four or win a championship.

[Troublemaker]

I think the fact that a substantial number of Duke fans are defending the choice of a UNC player as NPOY is more telling by far than and contortion of stats.

The stats aren't contorted. Actually I'm not sure what you mean by that. Also, I disagree that 4 or 5 Duke fans supporting Hansbrough is telling. Or if it IS telling, I don't think it is telling us anything more significant than "Duke fans watch lots of UNC games."

[sandinmyshoes]

The stats aren't contorted. Actually I'm not sure what you mean by that. Also, I disagree that 4 or 5 Duke fans supporting Hansbrough is telling. Or if it IS telling, I don't think it is telling us anything more significant than "Duke fans watch lots of UNC games."

Perhaps "contorted" is not the correct word. Let me say that I found their stats selective. And their use of Hansbrough's stats from his first and second year an obvious sign of an agenda or bias. As others have pointed out, what do they have to do with this year's award? Why are they introducing something that is irrelevant?

And as for the number of Duke fans who think Hansbrough is worthy of the NPOY, four or five on this thread, from perhaps a dozen or so is a pretty substantial number. Heck, any number of Duke fans willing to concede something like this is telling, and it would be astoundingly so if UNC fans were doing it in reverse.

It doesn't mean it is a slam dunk that the award should go to Hansbrough. But it is evidence, unless you think basketball is ONLY about numbers. I think most of us, as Duke fans, are well aware of the all important intangibles. Our teams often thrive on those.

[Troublemaker]

Perhaps "contorted" is not the correct word. Let me say that I found their stats selective. And their use of Hansbrough's stats from his first and second year an obvious sign of an agenda or bias. As others have pointed out, what do they have to do with this year's award? Why are they introducing something that is irrelevant?

As I explained above, the author never claimed the previous seasons' stats to be relevant to his conclusion, which was based only on this season's stats. Why does every single word in a long article need to be relevant to his topic statement? Sometimes writers provide us with an aside that they believe the readers will find to be interesting (and I DID find that aside to be interesting). Sometimes they crack a joke. How is that a sign of agenda or bias?

Regardless, as I explained above, even if there IS an agenda, why would that agenda include a bias against Hansbrough? I think if the author had concluded that Hansbrough's efficiency stats were better than Love, then he would support Hansbrough for NPOY. That Hansbrough comes out on the short end of those stats is not a matter of bias. It just is what it is.

And as for the number of Duke fans who think Hansbrough is worthy of the NPOY, four or five on this thread, from perhaps a dozen or so is a pretty substantial number. Heck, any number of Duke fans willing to concede something like this is telling, and it would be astoundingly so if UNC fans were doing it in reverse.

It doesn't mean it is a slam dunk that the award should go to Hansbrough. But it is evidence, unless you think basketball is ONLY about numbers. I think most of us, as Duke fans, are well aware of the all important intangibles. Our teams often thrive on those.

No, it's not telling. I'm being very strict here but obviously of all the evidence in support of Hansbrough, one of the weakest and most ridiculous would be that 4 or 5 Duke fans out of 12 on a random thread on a random message board support Hansbrough for NPOY. It's not really evidence of anything. And even if it were 400,000 Duke fans out of 1.2 million Duke fans, I'm not sure it's indicative of anything more than Duke fans watch lots of UNC games and so we're more exposed to Hansbrough than to other candidates.

As I mentioned above, I support Hansbrough for NPOY. But that doesn't mean I believe the Prospectus author has a bias against Hansbrough (or for Love) or that this thread is telling of anything.

[sandinmyshoes]

As I explained above, the author never claimed the previous seasons' stats to be relevant to his conclusion, which was based only on this season's stats. Why does every single word in a long article need to be relevant to his topic statement? Sometimes writers provide us with an aside that they believe the readers will find to be interesting (and I DID find that aside to be interesting). Sometimes they crack a joke. How is that a sign of agenda or bias?

Regardless, as I explained above, even if there IS an agenda, why would that agenda include a bias against Hansbrough? I think if the author had concluded that Hansbrough's efficiency stats were better than Love, then he would support Hansbrough for NPOY. That Hansbrough comes out on the short end of those stats is not a matter of bias. It just is what it is.



No, it's not telling. I'm being very strict here but obviously of all the evidence in support of Hansbrough, one of the weakest and most ridiculous would be that 4 or 5 Duke fans out of 12 on a random thread on a random message board support Hansbrough for NPOY. It's not really evidence of anything. And even if it were 400,000 Duke fans out of 1.2 million Duke fans, I'm not sure it's indicative of anything more than Duke fans watch lots of UNC games and so we're more exposed to Hansbrough than to other candidates.

As I mentioned above, I support Hansbrough for NPOY. But that doesn't mean I believe the Prospectus author has a bias against Hansbrough (or for Love) or that this thread is telling of anything.

Ah, then what we have here is a disagreement on many points between thee and me. In my book, Duke fans overcoming their bias against a UNC player IS indicitative of how much that UNC player impacts the game.

As for the bias of the author, why argue for objective use of stats, but then include irrelevant stats?

[Troublemaker]

Ah, then what we have here is a disagreement on many points between thee and me. In my book, Duke fans overcoming their bias against a UNC player IS indicitative of how much that UNC player impacts the game.

...or that Duke fans are more exposed to Hansbrough than to other candidates, if you're comparing them. No argument from me that Hansbrough makes an impact, though. That's obvious.

As for the bias of the author, why argue for objective use of stats, but then include irrelevant stats?

Been explained.

[calltheobvious]

Again, this all depends on your definition of "performance" and what variables factor into your evaluation. In the writer's definition he factors in efficiency/pace of play/etc. but not everyone factors those things into their evaluation of "performance" and thats part of MY point. Just b/c you call it antiquated doesn't make it right or wrong. What it is is "simple." The Big 3 stats (pts, rebs, asts) have always been the big stats and they are easy to monitor.

I don't call the practice of not using tempo-free stats "simple," I call it "absurd." And I will call it "wrong," if by "right," you mean using a set of statistics that can give us the most accurate picture possible of what's going on so that comparisons become intelligible.

How can you argue that ignoring tempo is okay? What if Grinnell jumped to DI tomorrow and continued their style with the same success? Would you vote a Grinnell player worthy of NPOY solely because he averaged 34.8 ppg, 9.2 rpg, and 8.7 apg? Of course not, because you'd realize that the numbers were inflated by the number of possessions.

And at the end of the day guys, tyler scores 24ppg to Love's 18ppg (and I already mentioned that the other significant factors are pretty much a wash, imo). Thats a significant difference and PPG is obviously the most looked at and heavily weighted stat for basketball players. Shelden averaged 18/game. JJ averaged 24. Think about that. If thats an antiquated viewpoint, then fine, but basketball is still about outscoring your opponent so PPG is kinda important.

Seriously, when you discount for tempo, it's not a significant difference. And the other stuff is a wash? Let's say you get to pick Hansbrough or Love for your team. You know with certainty that Love is going to get the rebound a much higher percentage of the time the ball comes off the rim than Hansbrough will. You also know that he's going to find other players for baskets more often, and block a lot more shots. All that stuff's irrelevant? Please don't be the GM of my team.

As for PPG being the most heavily weighted stat, you'd get not one whit of disagreement from the BP writer. The point that Pomery makes seriously, and www.firejoemorgan.com makes comically, is that this is an unbelievably parochial way to look at things, especially given all of the superior metrics we have available now.

[wumhenry]

I don't call the practice of not using tempo-free stats "simple," I call it "absurd." And I will call it "wrong," if by "right," you mean using a set of statistics that can give us the most accurate picture possible of what's going on so that comparisons become intelligible. How can you argue that ignoring tempo is okay?

Why do you ignore the point made in #27?

[calltheobvious]

Why do you ignore the point made in #27?

Didn't mean to. Here's my response to your point that the validity of temp-free stats assume Love's ability to play at a higher tempo.

I think this question turns on the issue of framing. The way you pose the question, I have to justify the idea that Love could play equally well at a quicker pace. But what if the tables are turned and you have to assess what Hansbrough's numbers would look like with 11 fewer possessions per game? After all, Hansbrough gets a statistically significant number of points because of his speed and endurance, and if UNC slowed down the pace to UCLA's level, the impact might be more than linear.

Ultimately, players and teams play how they play. Tempo-free stats are a way to try to assess the impact of players at the margins of individual possessions, and I think that's the most complete way to assess the picture.

Just curious, but would you feel as strongly if TH's numbers were deflated by 5.5 possessions per game and Love's were inflated by the same factor?

[wumhenry]

Here's my response to your point that the validity of temp-free stats assume Love's ability to play at a higher tempo.

I think this question turns on the issue of framing. The way you pose the question, I have to justify the idea that Love could play equally well at a quicker pace. But what if the tables are turned and you have to assess what Hansbrough's numbers would look like with 11 fewer possessions per game? After all, Hansbrough gets a statistically significant number of points because of his speed and endurance, and if UNC slowed down the pace to UCLA's level, the impact might be more than linear.

Sure, it stands to reason that Hansbrough's PPG would be lower if UNC played at UCLA's slower pace. But it's not self-evident that Love's PPG would be as high as, or higher than, TH's if UCLA played at UNC's pace. So it seems to me that your tempo-adjusted statistics are worthless in the absence of a convincing reason to assume that Love could play at UNC's pace with no loss of efficiency.

[Troublemaker]

So it seems to me that your tempo-adjusted statistics are worthless in the absence of a convincing reason to assume that Love could play at UNC's pace with no loss of efficiency.

Worthless? Absolutely not. Tempo-free stats are better indicators than non-tempo-free stats. But your point was indeed a good counterpoint to the article. Keep in mind that there's a counterpoint to your counterpoint as well. Does Hansbrough's efficiency drop if he is forced to play at Love's pace? One of Hansbrough's biggest strengths is arguably his conditioning, and a slower pace would reduce the advantage that his superior conditioning gives him over opposing big men, thereby perhaps causing a decline in his efficiency stats if he were to play at Love's pace.

[tbyers11]

Sure, it stands to reason that Hansbrough's PPG would be lower if UNC played at UCLA's slower pace. But it's not self-evident that Love's PPG would be as high as, or higher than, TH's if UCLA played at UNC's pace. So it seems to me that your tempo-adjusted statistics are worthless in the absence of a convincing reason to assume that Love could play at UNC's pace with no loss of efficiency.

In my opinion, you are overestimating the effect that UNC's fast pace would have on a particular player from a slower paced team. The difference (according to Ken Pom) is 11 possessions over a 40 minute game. Basically, UNC gets slightly more than 1 extra possession every 4 minutes than UCLA. So if UCLA ran one extra fast break every 4 minutes, they would roughly have an equivalent amount of possessions/game.

Kevin Love is definitely not in as good of shape as Tyler Hansbrough, but I don't think that type of increase in pace would adversely affect him much and his lesser conditioning wouldn't keep him from posting numbers at his current pace. While UNC plays fast, they do not play a non stop full-court pressing scheme like some of Nolan Richardson's old Arkansas teams or Pitino's old Kentucky teams.

[bdh21]

Sure, it stands to reason that Hansbrough's PPG would be lower if UNC played at UCLA's slower pace. But it's not self-evident that Love's PPG would be as high as, or higher than, TH's if UCLA played at UNC's pace. So it seems to me that your tempo-adjusted statistics are worthless in the absence of a convincing reason to assume that Love could play at UNC's pace with no loss of efficiency.

Why is it relevant whether Love could play at a faster pace? We're not picking teams for a pick up game. UCLA and UNC aren't trading those two players for the tournament.

Tempo-free stats are just there to explain production within the context of the offense. And within the context of UCLA's offense Kevin Love is as productive as Hansbrough is within the context of UNC's offense. What's wrong with Ben Howland implementing a strategy that gives his team the best chance to win games even if it involves 11 fewer possessions per game than UNC?

[calltheobvious]

Sure, it stands to reason that Hansbrough's PPG would be lower if UNC played at UCLA's slower pace. But it's not self-evident that Love's PPG would be as high as, or higher than, TH's if UCLA played at UNC's pace. So it seems to me that your tempo-adjusted statistics are worthless in the absence of a convincing reason to assume that Love could play at UNC's pace with no loss of efficiency.

Apparently I didn't make myself clear. Of course everyone agrees that TH's numbers would be lower with 11 fewer possessions. I'm saying that deflating his numbers by 11 possessions out of an "average" UNC game might actually underpredict the drop-off. In other words, I think there's an increasing marginal return to Hansbrough on Carolina possessions. He's able to get fast-break points partly because his opponents aren't as fit as he is. If I'm right--and I'm not certain I am, I'm just certain that you haven't proved I'm not--Hansbrough does not get short-changed in the tempo-free discussion.

[killerleft]

If Hansbrough is more fit than Love, isn't that just another feather in his NPOY hat? If they play head-to-head wouldn't Hansbrough have a built-in advantage?

I confess that I have seen Love play only a couple of times. But this point-counterpoint stuff can be argued forever.

[Troublemaker]

If Hansbrough is more fit than Love, isn't that just another feather in his NPOY hat? If they play head-to-head wouldn't Hansbrough have a built-in advantage?

I confess that I have seen Love play only a couple of times. But this point-counterpoint stuff can be argued forever.

Conditioning is indeed part of being a good basketball player. But it's only one part of many, many parts constituting the whole. Love could still be a better basketball player than Hansbrough despite being less fit (IF he's even less fit).