This may be my pantyhose riding up into the unknown, but I think Hollinger's "luck" premise stems a lot from Rob Neyer's and Bill James' explanation of deviance of the actual from the Pythagorean projection for wins based upon run differential in baseball. They claim (paraphrased) that over a 162-game season, managerial decisions and close-game factors generally wash out, so if a run differential analysis says a team should win 93 games and they win 96, it's considered "luck" and unable to be counted upon going forward. I think Hollinger used that as part of his projection analysis to show that, even though we won 41 games last year, we played above our heads (in terms of point differential) and close-game results.
But I think a few things discount that "luck" aspect for us. One is that Nate's coaching is considered about the best in the league coming out of timeouts (which are by-and-large in late-game situations). So our record in close games wasn't determined as much by "luck" as by an "off-the-court asset"...which still counts for wins. I think a projection spanning a few years will show that our point differential is generally deflated due to things like pace and lower-than-average minutes from our elite players; and that our record is inflated due to good coaching out of timeouts and "clutchness" of our players.
That said, I'm having a hard time finding John's preseason projections...