Back when I started the "Project Lottery" countdown, I took the record of the worst team in the league every year for the last decade, added them all together, then divided. Nothing advanced about the math, and it served my purposes for calculating a starting point for the countdown. Today, ESPN's resident hoop-stat guru John Hollinger has basically done the same thing, but he gets paid a lot more money so he used an 11-year sample.
Hollinger crunches the numbers and tries to calculate whether it's worth it for teams to tank the season just for a better shot at Greg Oden, or Kevin Durant. It's worth a read, but Hollinger really gets bogged down by numbers, and I think he needs to add in a little perspective. He bases his whole argument on the benchmark of 67 wins. What he doesn't take into account is the state of the league this year.
For years, the West has been head-and-shoulders above the East. That's just a fact. Even though 2 of the past 3 championships have come from the East, it's been obvious that the West has a huge advantage on the whole. This year, the East seems to be even weaker. With a couple of teams vying for the #1 seed, and everyone else mired below .500. The fact of the matter is that the league is full of horrible teams this year. Teams that in any other year would have a legitimate shot at losing 70 games. But not this year, and the simple reason is that these horrible teams play each other, someone has to win.
I was thinking about re-adjusting the countdown to take this into account, but at least 65 is based on something other than an informed, but subjective look at the state of the NBA. Hollinger, much like Billy Beane and a growing number of MBA-toting, Bill James-worshiping young baseball executives, is a slave to statistics. For me, stats are a good way to predict what someone will do over a long period of time, but completely useless when relied upon without context. Individual games are decided by sprints, not marathons. This is why "Moneyball" doesn't translate into short playoff series. It's also why Hollinger's statistic-driven power ranker is laughable most of the time. Statistics alone can be bent to show any kind of trend, it's only when you add perspective that they have real meaning.