"The Wages of Wins"

But I guess he's an ESPN personality rather than an economist, so he doesn't get a book.

Just sayin'.

This season's TSP regular season leaders were Nash, Curry, Jefferson, Billups and Iguodala.

Hollinger's Player Efficiency Rating (basically a 'Win Score') leaders were Nowitzki, LeBron, Kobe, Flash,KG, Elton Brand and Iverson.

his explanation of it:
The Player Efficiency Rating (PER) is a rating of a player's per-minute productivity.

To generate it, I created formulas -- which I've outlined in tortuous detail in "Pro Basketball Forecast" -- that return a value for each of a player's accomplishments. That includes positive accomplishments, such as field goals, free throws, 3-pointers, assists, rebounds, blocks and steals, and negative ones, such as missed shots, turnovers and personal fouls.

Two important things to remember about PER is that it's per-minute and pace-adjusted. It's a per-minute measure because that allows us to compare, say, Drew Gooden to Donyell Marshall, even though there is a wide disparity in the minutes they played. I also adjust each player's rating for his team's pace, so that players on a slow-paced team like Indiana aren't penalized just because their team's games have fewer possessions than those of a fast-paced team such as Phoenix.

Bear in mind that this rating is not the final, once-and-for-all answer for a player's accomplishments during the season. This is especially true for players such as Bruce Bowen and Trenton Hassell who are defensive specialists but don't get many blocks or steals. What PER can do, however, is summarize a player's statistical accomplishments in a single number. That allows us to unify the disparate data on each player that we try to track in our heads (e.g., Danny Fortson: great rebounder, high-percentage shooter, turnover machine, fouls like crazy, etc.) so that we can move on to evaluating what might be missing from the stats.

I set the league average in PER to 15.00 every season.

Among players who played at least 500 minutes in 2004-05, the highest rating was Kevin Garnett's 28.35. The lowest was Theron Smith's 5.10.

Not to get all John Kruk here, but common sense doesn't support it. If you've watch Hoiberg play with any frequency, you know the guy is a pretty good player when it comes to draining a wide open outside shot, a la Kyle Korver, Steve Kerr, etc. But you'd also know that unlike those guys, he brings nothing else to the table at all. When he's guarded, he doesn't shoot. Period. But his stats don't express that level of nuance, and neither do Hollinger's. Reggie Evans is a perfect example. If you let a relatively tall, athletic guy play for 15 minutes a game, and tell him "Concentrate on nothing but boards," he's going to collect some. But Evans has neither the endurance to collect rebounds over the long haul nor the skill set to contribute anything else to a team.

The problem with per minute stats is that they're the equivalent of saying that Mark Sweeney is a better hitter than Todd Helton because Sweeney hit .320 as a pinch hitter but Helton only hit .310 over the course of the season. They give you a solid picture of what a player is accomplishing during his short bursts of productivity, but they don't countermeasure that against what a more talented player could do in a similar burst role. If Elton Brand was strictly a role player and only played two 10-minute sessions per game, and was told specifically to concentrate on rebounding within a zone concept, OF COURSE he's going to pile up on rebounds, and more effectively than anyone of Evans's ilk.

That doesn't mean that Hoiberg / Evans aren't good players, but I don't think it's comparable to compare the shooting of someone like Hoiberg with someone like, say, Ray Allen, who is largely scoring off of double teams, creating wide open looks for set shooters like Hoiberg. Etc.

Maybe I'm just reacting to some ESPN lackey's headlines, but I distinctly remember reading an article called (paraphrase) "Who is the best shooter in the NBA? It's not who you think", and without much of a caveat in the article. I may be wrong on this point.

PER, for me, is a mumbo jumbo stat like Quarterback Rating or Win Shares. Per minute stats at least have some kind of practical use. QR, WS and PER, are totally arbitrary and largely not useful from a greater statistical perspective.

Nothing in Hollinger's analysis or statements would, necessarily contradict with that. That's why the article you're talking about called him the best shooter in the league at that time - and not the best player.

In terms of PER, being the best shooter in the league, if Hoiberg could have kept that up as a full-time player (something Hollinger never argues), he still wouldn't be a top-10 or top-20 player. If, as one would expect, his skills regressed with more playing time, he'd barely be a top-50 player.

Win Shares (if not Win Scores) and PER are hardly 'arbitrary.' What they do is measure how each minor contribution adds up in terms of player points or team wins. A better baseball measurement would be WARP or VORP (Wins Above Replacement Player, Value Over Replacement Player) - like PER, they rely on how often each individual act in baseball - a strikeout, a double - leads to a run being scored or a team winning. How often do possessions where Steve Nash takes the ball up the floor end up in a field goal? (etc.)

The QB rating is, as I understand it, fairly meaningless.

The problem with per minute stats is that they're the equivalent of saying that Mark Sweeney is a better hitter than Todd Helton because Sweeney hit .320 as a pinch hitter but Helton only hit .310 over the course of the season.
That's a valid point, if anyone did that. But no statistician in either basketball or baseball does so, to my knowledge. Hollinger says that PER is a per-minute rate state, yes, but never that it makes equivalences.

For me, draining a wide open shot != best shooter.

... based on a completely arbitrary set of stats and ratios.

I don't like these stats either. Basically, any stat that combines defensive metrics with offensive metrics, or hitting/fielding/pitching, etc.... I am surpremely skeptical of their merits.

I don't think you're really making a good case against 'advanced statistics' here - what is there to be suspicious about. What, exactly, renders, say, VORP meaningless - what would be more meaningful?