God god heyp, you're so dense. So many people have explained to you why your belly-aching is silly but let me try one more time. Hollinger's power ranking is trying to find a predictive measure/the best teams based on recent trends and weighted components like strength of opponents, point margin, and # home and away games. I'm not saying it's perfect or even that accurate, but it is better than just some dude's opinion. RW/HL is just a statistical quirk that shows high correlations between two obviously related things: record and seeding. Oh it's so novel that a measure derived from the record correlates to standings!! It's like being amazed that level of income correlates to size of house bought or being amazed that smokers have higher chance of lung cancer.
All you are saying is complexity means the stat must be better, when in fact we all now Hollingers stat is not very accurate. People point to Pt Differential all the time. Are they dense too. "RW/HL is a statistical quirk" is about the most dense statement you could possibly make. It works every year. Pat Riley first brought this to my attention in like 1993. For nearly 20 years, I've been using this stat to follow accurate seeding throughout the year (vs W/Ls), and it works. Just because you have never used it doesn't mean it doesn't work.
Losing road games to mid-ranked opponents isn't that harmful to your rating as long as you don't let them run the score up. None of those losses were blowouts (well, a lot of them were blowouts until the 4th quarter effort by Harden, but the formulas don't know the difference between that and a close game). But another big part of it is Clippers, Knicks, Golden State, etc all having a lot of slippage over the last few months as well.
We did. I believe we got as low as 16th, 17th or 18th, I can't remember. The formula is heavily weighted in several different areas. Two of which are margin of victory in the last 25% of the team's games (which we're fourth) and strength of schedule - only New Orleans has played a tougher schedule than the Rockets so far this year (based on opponents winning percentage). Still not sure how we're rated so high, though...lol.
Because we show really well on strength of schedule, margin of victory, and recent performance in each. Not that complicated. Not saying we're truly the 4th best team in the league, but we're not 14th best either (if you only went by our record). We're somewhere in between the two, in my opinion. Somewhere in the top 10 but outside the top 5.
Don't look across conference wrt to W/Ls. Stick to the West. We are not #3 in the West. Hollinger says we are. Our records says we are #8. We might make #7, but we are nowhere close to #3. So, regarding just the West, we are not closer to Hollinger's rankings than true rankings...not even close.
What's really weird is we are 2-14 against the top seven teams in the NBA (based on record). Yet, Hollingers flawed computer ranking doesn't account for that fact.
Didn't Hollinger just land a job in a NBA front office while you are posting anonymously on a basketball forum?
I'm a better computer scientist than he is. I know when simplicity is better than complexity. Occam's Razor. RWs / HLs is a better stat than his. And our 2-14 record against the top teams proves that.
After 3 weeks without Patterson, the Rockets are now down to 8th in Hollinger Power Ranking. Their Hollinger playoff odds are down to 90% whereas it was close to 98% before. http://espn.go.com/nba/hollinger/playoffodds
