Hollinger Rankings: C's #11

[Employee8]

Ah gotcha.  Makes sense to me.  I would think statistics play a role in trading/acquiring/releasing players as well as evaluating their personality and the ability of his skills to mesh in with the team.  Seems to me both of you guys are dug in deep on both sides of the trench when there should be a fair balance between the two. 

I think people take offense to Hollinger because his predictions/rankings are purely on a statistical approach and personally, I don't think that works.  And that serves to explain why he is pretty much 50/50 with his predictions in the playoffs.

No they aren't - he just maintains one ranking list based on a stats system as an exercise. He also writes scouting reports of every NBA player and offers his personal opinion frequently beyond "just the numbers".

He just gets more attention for his numerical systems, because that's his unique schtick.

Also, is there ANYONE consistently good on playoff predictions, beyond a little pure luck (exhibit: all NCAA pools).

Right, but the topic of this thread is his ranking, not a "What do you think of Hollinger as a writer?" topic.

[action781]

I don't see what metrics can rank the Celtics above the Mavs and Lakers.

Both the Mavs and Lakers, have better records than the Celtics and that is with stronger strength of schedules.  Dallas actually has a higher margin of victory too.

This makes no sense to me.

[MBunge]

I'm interested in the statistical side of the 2007-2008 Celtics and the recent Mavs squad.  Let's hear it.

What I'd like to see is the statistical analysis that projected the Mavs over the Heat as decisively as it turned out.  I believe that Hollinger had the Heat over the Lakers in last year's finals and predicted a Dallas loss in the 1st round.

Mike

Wow, his stats must be awful considering the Heat came within 2 games of the championship (and 3 points of a 3-1 lead), and the Mavs played a close 6 game series in the 1st round.

Do you call people stupid for betting on red at the casino, when it comes up black?

And right here, ladies and gentlemen, is what drives people up the wall about stat heads.  Whenever statistical analysis works, like forecasting a victory for a team with a worse won-loss record but a better point differential, it's absolute proof that stats are the be-all, end-all of understanding sports.  But when statistical analysis proves as flawed and fallable as anything else, it means absolutely nothing.

Mike

[fairweatherfan]

I don't see what metrics can rank the Celtics above the Mavs and Lakers.

Both the Mavs and Lakers, have better records than the Celtics and that is with stronger strength of schedules.  Dallas actually has a higher margin of victory too.

This makes no sense to me.

Stronger weighting of recent performance - Cs have been winning and have a couple of blowouts (Magic and Raptors) and most of their losses have been fairly close.  Over the last 10 games the Celtics average margin is +9.6, the Mavs are +2.1.  The Mavs have played better teams but that's a big gap.

There is some validity to this approach - how a team played in its most recent games is a better predictor of future success than what it did in the early season.  That's why Orlando is #19 despite being 4 games over .500 - they started hot and have completely collapsed.

[byennie]

And right here, ladies and gentlemen, is what drives people up the wall about stat heads.  Whenever statistical analysis works, like forecasting a victory for a team with a worse won-loss record but a better point differential, it's absolute proof that stats are the be-all, end-all of understanding sports.  But when statistical analysis proves as flawed and fallable as anything else, it means absolutely nothing.

Mike

That's exactly what stat heads DON'T do, and it's not what I was doing. I have stated multiple times that they are not the "be all" or "end all". They are just a part of the equation. The point is that there is NO reliable method of picking playoff results known to man, or else said man would be making a ton of money in Vegas.

You are making nothing but strawman argument after strawman argument here. The "stat heads" actually understand the limits of their methods, and still value scouting and everything else. YOU are the one who is trying to make it black & white and dismiss stats.

[fairweatherfan]

Whenever statistical analysis works, like forecasting a victory for a team with a worse won-loss record but a better point differential, it's absolute proof that stats are the be-all, end-all of understanding sports. 

I've never heard of a single "stathead" who's ever said stats are "the be-all end-all of understanding sports".  They pretty much unanimously take the moderate position that stats are a useful tool for evaluating players and teams, but subjective analysis plays an important role as well.  Advanced stats give us info we can't get elsewhere; but nobody pretends they're Scripture except for people who feel threatened by them.

And no one who knows anything about statistics would cite a single case as absolute proof of anything. 

But when statistical analysis proves as flawed and fallable as anything else, it means absolutely nothing.

There's lots of evidence that advanced statistical analysis, while certainly not perfect, is significantly less flawed and fallible than conventional metrics (although ironically the most popular one of these - W/L record - is itself a statistic).  You probably wouldn't care for that evidence, though - it's based in stats.

[Who]

I don't see what metrics can rank the Celtics above the Mavs and Lakers.

Both the Mavs and Lakers, have better records than the Celtics and that is with stronger strength of schedules.  Dallas actually has a higher margin of victory too.

This makes no sense to me.

Stronger weighting of recent performance - Cs have been winning and have a couple of blowouts (Magic and Raptors) and most of their losses have been fairly close.  Over the last 10 games the Celtics average margin is +9.6, the Mavs are +2.1.  The Mavs have played better teams but that's a big gap.

There is some validity to this approach - how a team played in its most recent games is a better predictor of future success than what it did in the early season.  That's why Orlando is #19 despite being 4 games over .500 - they started hot and have completely collapsed.

I got the impression that the only reason for the recent performance angle is to give some added movement to the rankings to make it more interesting for viewers over the course of a season.

A statistical equivalent of Marc Stein's power rankings. Lots of movement on a regular basis to keep readers engaged in it throughout the year.

[fairweatherfan]

I don't see what metrics can rank the Celtics above the Mavs and Lakers.

Both the Mavs and Lakers, have better records than the Celtics and that is with stronger strength of schedules.  Dallas actually has a higher margin of victory too.

This makes no sense to me.

Stronger weighting of recent performance - Cs have been winning and have a couple of blowouts (Magic and Raptors) and most of their losses have been fairly close.  Over the last 10 games the Celtics average margin is +9.6, the Mavs are +2.1.  The Mavs have played better teams but that's a big gap.

There is some validity to this approach - how a team played in its most recent games is a better predictor of future success than what it did in the early season.  That's why Orlando is #19 despite being 4 games over .500 - they started hot and have completely collapsed.

I got the impression that the only reason for the recent performance angle is to give some added movement to the rankings to make it more interesting for viewers over the course of a season.

A statistical equivalent of Marc Stein's power rankings. Lots of movement on a regular basis to keep readers engaged in it throughout the year.

I can't rule that out as a consideration, but there is data to show that recent performance is a stronger predictor than season-long performance as a whole.  It's not woven from thin air or anything, there is a rational basis to it.

[BudweiserCeltic]

I don't see what metrics can rank the Celtics above the Mavs and Lakers.

Both the Mavs and Lakers, have better records than the Celtics and that is with stronger strength of schedules.  Dallas actually has a higher margin of victory too.

This makes no sense to me.

Stronger weighting of recent performance - Cs have been winning and have a couple of blowouts (Magic and Raptors) and most of their losses have been fairly close.  Over the last 10 games the Celtics average margin is +9.6, the Mavs are +2.1.  The Mavs have played better teams but that's a big gap.

There is some validity to this approach - how a team played in its most recent games is a better predictor of future success than what it did in the early season.  That's why Orlando is #19 despite being 4 games over .500 - they started hot and have completely collapsed.

I got the impression that the only reason for the recent performance angle is to give some added movement to the rankings to make it more interesting for viewers over the course of a season.

A statistical equivalent of Marc Stein's power rankings. Lots of movement on a regular basis to keep readers engaged in it throughout the year.

I don't buy that. How can you not weight how you're performing recently as opposed as to what you were performing earlier in the year? Heading to the playoffs, for example, wouldn't you be concerned more with a team that is hot and got plenty of momentum vs. a team that started hot but throughout the year hasn't been playing as well, and they merely have a higher seed because of the cushion of earlier in the year?

Recent performance should be a better predictor of where a team is heading going forward.

[Who]

I don't see what metrics can rank the Celtics above the Mavs and Lakers.

Both the Mavs and Lakers, have better records than the Celtics and that is with stronger strength of schedules.  Dallas actually has a higher margin of victory too.

This makes no sense to me.

Stronger weighting of recent performance - Cs have been winning and have a couple of blowouts (Magic and Raptors) and most of their losses have been fairly close.  Over the last 10 games the Celtics average margin is +9.6, the Mavs are +2.1.  The Mavs have played better teams but that's a big gap.

There is some validity to this approach - how a team played in its most recent games is a better predictor of future success than what it did in the early season.  That's why Orlando is #19 despite being 4 games over .500 - they started hot and have completely collapsed.

I got the impression that the only reason for the recent performance angle is to give some added movement to the rankings to make it more interesting for viewers over the course of a season.

A statistical equivalent of Marc Stein's power rankings. Lots of movement on a regular basis to keep readers engaged in it throughout the year.

I can't rule that out as a consideration, but there is data to show that recent performance is a stronger predictor than season-long performance as a whole.  It's not woven from thin air or anything, there is a rational basis to it.

Ah yes, that sounds familiar after reading it.

I read something about that before. I had forgotten it though.

-------------------------------------------

I have a somewhat vague memory of a Hollinger interview on ESPN where he was explaining his power rankings when they began and he talked about the recent schedule idea and he did mention that it was, at least in part, in an effort to give his power rankings more movement to generate more fan interest (viewership) in them throughout the season. Ala Marc Stein.

I can't track that interview down but I looked up the power rankings explanation instead and there is a short note on it saying that it is for both reasons mentioned:

Recent performance
Another key variable in the formula is recent performance, which I included for two reasons.

First, it stands to reason that more recent games are more valid indicators of how strong a team is currently.

Second, I wanted these rankings to follow the model of Marc Stein's "human" power rankings, on the site each Monday, in which a team's recent play is a huge factor.

To accomplish this, I weigh a team's full-season results by two-thirds and its most recent games by another one-third, so the overall ranking gives greater weight to recent games.You're probably wondering at this point what I mean by "recent." It varies depending on where we are in the season.For the first 33 games of the season, it means a team's past 10 games.

From that point forward, however, it means the most recent 31.8 percent of a team's schedule. The net result is that, after the first 33 games, a team's most recent 31.8 percent of its schedule will account for 40 percent of its ranking.

[MBunge]

YOU are the one who is trying to make it black & white and dismiss stats.

Now, THAT is a strawman argument.  I haven't dismissed stats.  All I asked was "Where's the statistical analysis that predicted the Mavs over the Heat" and pointed out stat head Hollinger's incorrect forecast of last year's playoffs.  What, exactly, is unfair or objectionable about pointing out the failure of one stat head in particular and statistical analysis in general when it came to last year's NBA playoffs?

YOU'RE the one who took issue with those statements and if it wasn't out of typical stat head sensitivity, why was it?

Mike

[Fafnir]

Please calm it down everyone, caps "YOU"s and posts are getting too heated.

[MBunge]

There's lots of evidence that advanced statistical analysis, while certainly not perfect, is significantly less flawed and fallible than conventional metrics

Which is all well and good, but doesn't explain why stat heads like you freak out when the failings of statistical analysis are pointed out.  As I already said, what was wrong or unfair with asking "Where's the statistical analysis that predicated the Mavs over the Heat in the Finals"?  How is that some sort of impermissable attack?

Mike

[FallGuy]

You know what I don't get at all?

Stats, particularly advanced stats, are fact based.

They can tell you that e.g. Paul Pierce is or is not a good shooter from the midrange, what the Celtics best five-man unit is, that the C's struggle to score in times x or y, etc. etc.

Why would anyone dismiss that evidence?

Why wouldn't you want to be smarter and more comprehensive in your knowledge of things?

[FallGuy]

There's lots of evidence that advanced statistical analysis, while certainly not perfect, is significantly less flawed and fallible than conventional metrics

Which is all well and good, but doesn't explain why stat heads like you freak out when the failings of statistical analysis are pointed out.  As I already said, what was wrong or unfair with asking "Where's the statistical analysis that predicated the Mavs over the Heat in the Finals"?  How is that some sort of impermissable attack?

Mike

Butting in:

1) Perhaps there is predictive evidence on record. Certainly, the Mavs had some in-house figures that suggested something one way or another. It's certainly reasonable that of those that predicted the Mavs to win, some used statistical analysis in making those predictions.

2) Whether certain analytical approaches accurately or did not accurately predict a single series is in and of itself a single occurrence and not a valid argument for or against the use of evidence in analyzing basketball.