I was taking a look at Houston's ORtg/DRtg on/off splits. We're better on offense with Harden, no surprise, and much better on defensive with Asik. But look at these: http://www.basketball-reference.com/players/r/linje01/on-off/2013/ http://www.basketball-reference.com/players/r/parsoch01/on-off/2013/ Apparently, we're worse on offense when Lin is on the court, and MUCH worse when Parsons is on the court, which shocked me. Most would agree these are our best offensive players besides Harden, so this is really surprising. Any idea why this might be happening?
As an addendum, we're quite a bit better when our 3-point shooter Delfino is on: http://www.basketball-reference.com/players/r/delfica01/on-off/2013/ and he usually spends more time with the second unit (i.e. without Parsons or Lin). Harden, because of his heavy minutes, isn't as affected by this. This might explain part of it, especially Parsons' ORtg difference since he and Delfino play the same position.
Too much player interaction effects. If Regularized Adjusted Plus Minus (RAPM) were being kept season-to-date I'd bet a lot of money it tells a different story. http://stats-for-the-nba.appspot.com/ When it was still going in January, it had Lin as the 3rd highest in net RAPM, after Harden and Asik, as you'd expect (or at least as I'd expect).