John Hollinger

Prediction Update: Blazers Still On Target for 50 Wins by: torridjoe Fri Feb 06, 2009 at 15:46:13 PM PST Just under a month ago, with the Blazers entering January looking good at 21-14 and hoping to fatten up on lesser competition in the new year, I took a look at the team's year-end prospects by estimating wins in their remaining games (47 at the time). We're now down to 34 games left in the regular season, and the Blazers sit at 30-18, 12 games over .500 and an improvement of five games from our last look. In order to estimate the likelihood of a win in any given game, I used John Hollinger's Power Rankings for NBA teams, and divided the competition into four tiers of difficulty: Elite (30% chance of a win in any one game); Peer (50% chance); Lesser (60% chance); and Chum (80% chance). The tiers and figures are all eyeballed, based loosely on the idea that a team with a .700 winning percentage would be 70% likely to win any one game. Cleveland, Boston and LA are over .800 right now, but when you include Orlando, Denver, and Utah the average gets closer to .700. (And a month later we'd likely swap San Antonio for Utah, but the principle would remain the same).  Obviously a "peer" team is one generally equivalent, and that fairly well describes most of the other West teams Portland is competing with. Then there's a group of teams who aren't quite on that level but remain dangerous--we'll call them the Unpredictables, and given Portland's .600+ record so far, I apply the reverse of the elites calculation, giving the other teams 40% win chance against the Blazers. The "chum" represent the bottom half, against whom--by reputation and also good Blazer efficiency against such teams in the past--one can expect a win 4 out of 5 times. So, a month later, are the Blazers still on track? {see below for the answer)  There's More... :: (0 Comments, 421 words in story)

The NBA's Best Rebounder of 2008 is...Joel Przybilla? (Joel Przybilla!) by: torridjoe Fri Jan 02, 2009 at 08:00:00 AM PST I confess to being rather a stats geek, whether it's election statistics or fire/EMS or basic demography or sabermetrics. As much as I love football as a sport, it's a big downer that so few games means a small sample and much more variability in football stats. And other than Rachel Maddow and of course Barack Obama, who in politics came out of 2008 with their ticket punched more loudly than former Baseball Prospectus genius turned polling genius Nate Silver? It was a good year for geeks. And then there's basketball, which is kind of in between baseball and football in stat utility. Since the last time I was heavy into the NBA however, folks like ESPN's John Hollinger have injected the stat geek's spirit into it, and several good ratio statistics have been introduced. Some of the better basketball stats in terms of really seeing how well teams and players stack up are the points scored and allowed per 100 possessions, which compares efficiency against an even standard; and point differential (points scored minus points allowed). The differential can also be extrapolated in a quick formula to get what in baseball are called Pythagorean Wins: the expected win-loss total based on point differential. The excess is written off to luck and expected to be balanced out in the future as luck naturally turns. These and other stats at the player level are combined by Hollinger into a PER stat, which estimates total production per minute, pace adjusted. One of the component stats that I think really gets to the essence of a particular skill is the suite of rebound stats. What Hollinger does is compute the percentage of all possible rebounds (offensive, defensive or both) that the player actually got. It stands to reason that every time a shot is missed and you're on the floor, in theory you should have a chance at a rebound and the best rebounders get the highest percentage of available boards. I've spoiled the surprise, of course, but what I discovered while messing around with Hollinger's stats is that when it comes to rebounding, especially on the defensive end, there's no doubt that the best bound-for-bound glassman in the NBA is none other than the Vanilla Gorilla, Joel Pryzbilla. {the amazing truth, below}      There's More... :: (0 Comments, 733 words in story)

Return to LO home...!