Using his adjusted +/- "Points" rating, he briefly discusses which players have played well or poorly for every team in the month of January: http://waynewinston.com/wordpress/?p=402 Selected commentary from his Points analysis: Denver This team is totally keyed by the great play of Nene and Birdman, both of whom had a +23 Points rating in January. Billups and Melo are not nearly as important as the media believes. Houston David Andersen (+7 Points Rating) played very well. Chuck Hayes (-9 Points Rating , -11 Offense Rating) has killed the Rockets’ offense. Indiana The Pacers have two stars . Danny Granger and Mike Dunleavy both have +10 points ratings for January and the season. Solomon Jones (-25 Points rating) and Troy Murphy (-14 points rating) have really hurt the team in January and all season. Memphis The wrong Gasol got chosen for the all star team: Marc Gasol had a +22 point rating in January with a -18 Defense rating. Zach Randolph continued his great play (+10 Points Rating) while rookie Hasheem Thabeet slipped a bit( Points Rating of +9 through December and -1 in January). Sam Young (-15 Points rating) really hurt the Griz. OKC KD with his +26 points rating was the league’s best player in January. Jeff Green (-10 Points Rating) has hurt the team all season. Toronto The Raptors had a great month. Amir Johnson (+20 Points Rating -23 Defense Rating) and Maro Bellinelli (+14 Points Rating) were amazing, Calderon (+11 Points Rating) and Bosh (+9 Points Rating ) and Antoine Wright (+8 Points Rating) were also solid. Only rookie Demar DeRozan (-7 Points Rating) struggled. Washington Despite the off court turmoil, the Wizards kept soldiering on. Butler and Jamison (+5 Points Rating) were solid. Brendon Haywood (+11 Points Rating, -17 Defense Rating) has been great all season and gets no respect. Foye (-12 Points Rating) and Boykins (-22 Points Rating) killed the Wizards in January.
Pretty much my reaction when I read it. If you look at the most successful lineups, Andersen does tend to be part of them. And, amazingly, the Rockets have actually been 4 points per 100 possessions better on both offense and defense with Andersen on the floor. That's without adjusting for teammates or opponents, but Winston's method does that and he thinks the data shows that Andersen deserves a large share of the credit. Can't say I agree though. I think Andersen has been pretty awful, and this is just a big fluke.
His system produces lots of unexpected results. Is his systems is really revolutionary or is it just flawed? I don't know the answer to that but I suspect the latter. I wish he would provide explanations on his ratings in terms of on the court action, like what is David Andersen doing on the court to grade out so well.
I don't think the system is necessarily flawed, but the inferences he makes (and the confidence with which he makes them) might be. Its ok to have a statistical model that tries to estimate players performance and doesn't get it exactly right. If it can distinguish the really good players from the so-so players from the really bad players the majority of the time, that's still helpful. I think Winston's approach is very interesting in that sense. But it should be treated like a supplemental evaluation tool.
Code: # Unit Min Off Def +/- W L Win% 3 Brooks-Lowry-Budinger-Landry-Andersen 109.4 1.20 0.95 +52 11 5 68.7 4 Lowry-Battier-Budinger-Landry-Andersen 106.1 1.19 1.21 -9 8 11 42.1 5 Lowry-Ariza-Budinger-Landry-Andersen 95.9 1.08 0.95 +23 12 11 52.1 10 Brooks-Lowry-Ariza-Landry-Andersen 48.7 0.97 0.98 +3 8 8 50.0 http://www.82games.com/0910/0910HOU2.HTM This guy is nothing but an idiot who doesn't watch Rockets games if he gives Andersen the credit for the +/- the team posts when he is on the floor. Essentially ALL of Andersen's minutes come on the floor with Lowry and Landry, two guys who HAVE been playing legitimately awesome.
Andersen's minutes come against the 2nd unit and he is playing with two guys that the defense accounts for, leaving him wide open for his little Juwan Howard imitation. All you gotta do is flip him and Hayes and watch his +/- numbers plummet. Adelman is playing it right. When the opponent brings in their 2nd unit bigs, who usually have less offensive skills than the starters, then Adelman goes to Landry to create more spread, because if the opponent's bigs aren't as skilled offensively then Adelman doesn't have to worry about Hayes post defense over Landry's bad post defense. He can get by with Landry slopping around in the paint defensively and then get more spread on the other end of the floor. These guys are professionals. It's not like they don't do their own number crunching. That's why next year is going to be very interesting..........if we do not make a trade and bring back Scola. Because with Yao's defensive presence to go with Chuck's defensive presence, we may actually be better off starting Landry alongside Yao and bringing in Chuck and Scola with the 2nd unit. Scola's hustle...and his moves....going against second units should make him very, very efficient. And Landry's efficiency is just going to continue if he can play against the first unit alongside Yao being a big eraser to clean up his defensive gaps. David Andersen??? Dat dude is, ugh, permanent bench material.
Winston's method does account for the other 9 players on the floor (it is adjusted +/-). It doesn't consider their statistical production in terms of points, rebounds, assists, etc. Based just on the lineups and the changing score, his systems spits out Andersen as the biggest difference maker. It could just be a fluke. Or maybe Adelman cuts short Andersen's minutes when he's playing poorly, which makes him look more effective than he actually is on a per-possession basis.
I'm no expert, but it seems to me that +/- is much more effective in evaluating lineup combination than individual players. I mean, to say that Billups and Melo are overvalued in Denver is quite a reach, to me. Sure, Nene and Birdman could be the "x-factors", but that is because when they perform well usually Billups and Anthony are performing at their usual rate too- giving them four players being very productive. If you take Melo and especially Billups out of that lineup, I think you'll see Nene and Birdman have a big drop in +/- And like durvasa said, if the coach has a short leash on a player (David Andersen), his efficiency will be inflated.
