Over the last 10 games, Portland has gone 7-3. During that period our ranking among NBA teams has been right at the top. The following list shows where the Blazers rank for a number of key team statistics. * Point differential: #1. The most important measure! * Opponents' points per game: #3. Defense wins in the playoffs. * Assists differential: #3. Offensive efficiency exceeding opponents' offensive efficiency. * Rebound differential: #3. Hustle, toughness inside. * Block differential: #3. Hustle, toughness inside, again. * Opponents's rebounds allowed: #2. Keeping opponents from doing good things is critical. Here, don't let them get the ball. * Opponents' assists allowed: #3. Another measure of keeping opponents from doing good things. Here, disrupting offensive efficiency. These are some really outstanding stats. And they show things that we've known about the Blazers all year -- good offensive efficiency, good rebounding. But they show some surprising defensive effectiveness: keeping opponents points, assists, and rebounds down. As well as hustle and toughness. These are not how league observers characterize the Blazers. You hear a lot about slow pace and poor defense. These stats show that we ARE peaking at the right time. If we can keep this up into the playoffs, we can go farther than people think!
Disagree. Denver is 9-1 over that span. I dare say, wins are a MUCH more important measure than point differential. (Of course, strength of opponent comes into play in both measures.) That's the gist of the "distorted" thread regarding point differential... It isn't necessarily a measure of likelihood to earn wins.
point differential is a great statistic for making long run predictions for how a team should perform over the long haul of a season, not so much for a short run, small sample. As for the OP, I agree with your underlying premise that this team is playing it's best ball of the season at exactly the right time; this all bodes well for the post season assuming they can secure a top 4 seed, or if they at least can demonstrate an ability to beat a top 8 team on the road -- either Houston or San Antonio will do.
The problem is, point differential is a measure of what has already happened, not what is going to happen. Recent win-loss record is more meaningful for the latter, in my opinion. Point differential is one of those things you claim how great it is when your team is doing well in that category, and disregard when it doesn't reflect well, i.e. a lot of close wins...
I don't agree. I think point differential is the best measure of team ability. Using "recent won-loss" last year during the winning streak would have greatly overrated the team, since they were winning a lot of close games. Their point differential over that span didn't suggest they were a good enough team to legitimately win so many in a row...they had some luck. That was borne out by how weak they were the rest of the season. Depends on the person. I always think it's the best measure, even if it isn't perfect.
Yes it is a measure of what's already happened ... just like recent win-loss records, no? And "I" don't disregard point differential when it's not so great; I remember thinking at the time last year that the big winning streak might be a bit of fool's gold not only because of the competition we were facing and where we were playing them, but their overall point differential always pointed to the Blazers likely being around a .500 team -- especially because they were such a jump shooting team with fairly porous defense. Maybe you should define who "you" is?
One is the measure of an entire season to date, while the other shows a more recent trend. Point differential over the last 10 games would be a different matter... I meant "you" in the general sense.
Recent trends are also very subject to variance. Teams go through good stretches and bad stretches...they're generally not very predictive.
True. I just think they're a better indicator than point differential, which tells you next to nothing about recent play, just season averages. Averages by definition are not very indicative of recent trends.
