Sad news for those who don't want their eyes to burn out from overexposure to monitor-emitted gamma rays: John Hollinger's work will now only be found online. No longer putting out a yearly Pro Basketball Forecast in dead-tree format, the write-ups and projections you'd normally find in that book will only be found on ESPN.com, behind the formidable 'Insider' wall.
As alerted to me in the diaries, This year's edition was unveiled today, and to us 'outsiders', just a couple sentences of each player profile and projected PER were revealed. As you'd expect, the Bulls have a few slightly above-average players in terms of PER (where the average is 15) and no great ones:
|Player||Projected 2006-07 PER|
note: no PER projections available for rookies or those with statistically insignificant minutes totals last year. And while there are some projections based on Euroleague stats (Bargnani, etc.) there is no projection for Sefolosha, unfortunately. Hollinger has written about the impressive accuracy of these Euroleague translated stats, I remember he pegged Nocioni as a productive player using Euro stats before Noc set foot on an NBA court.
All the players in the NBA are also ordered by position, which is fun to check out before writing your email to Hollinger calling him a 'hater'. (judging by his online chat today)
While I like any system that projects Deng ahead of Nocioni, you'll notice it also 'ranks' the Rockets' Chuck Hayes ahead of Dwight Howard. As a plea to sensible readers I will rant (and gush) like so:
It is not meant to be an ultimate 'ranking', despite ESPN's attempts to convince us as such (to get some hot blog buzz, no doubt!). It does not factor in defense outside of blocks and steals. It's also a minute-based stat...so a guy like Hayes who grabs a bunch of rebounds in limited minutes is rewarded for efficiency, even if that doesn't mean he'd average 45 rebounds a game if you played him every minute.
The method of projections for the coming year is even more dicey, as Hollinger said in his chat today that there is zero subjectivity in the projections system: the projections are based off how similar players (in age and production) fared over the course of NBA history. Some skills have shown to improve with age (and at certain ages for that matter), others don't. Deng, Hinrich, and Gordon are all projected to improve, yet a dropoff is 'expected' for Nocioni. I'm guessing because he is relatively older than the others, and when players reach a certain point in their careers it's rare to get magically better at a particular aspect of their game. Doesn't it mean it's not possible, just relatively unlikely.
These types of numbers (and trends off those numbers) are meant to be a tool towards a discussion, not a commandment set in stone. For those who have insider access, go into the component numbers (rebound rate, true shooting percentage, etc) for each player and find strengths and weaknesses: a player's inability to get to the line or propensity for turnovers can effect overall efficiency more than you'd think. Using this and other statistical tools, plus (gasp) our very eyes, we can compare players with greater accuracy, normalizing variables that can skew 'traditional' statistics (::cough:: pace ::cough::).
ESPN is almost doing Hollinger a disservice to only publish PER for each player, which suggests to those without insider that he's a one-trick pony: just plugging in stats and getting a single rating. Every year, his book has an in-depth profile of each player and explains the tendencies that lead to the stats, as well as the numbers themselves. Plus he's a funny and interesting writer, although due to his new employer I doubt he can rip Dick Vitale for embarrassing himself on draft night like in books' past. My old pal Knickerblogger's been posting his PER calculations for a couple years now, but getting access to Hollinger's writing means more than just the 'official' numbers. Guess it's time to break out the credit card.
(So basically, it'll be like paying for the book, but I'll get other insider crap I likely don't want. And it'll cost more. Maybe during the SBNation holiday party someone can give me 'Insider' for secret santa.)