Jordan Syndrome wrote:Odinn21 wrote:Jordan Syndrome wrote:
They are over a large enough sample size.
I won't bother with the rest since I already explained the fundamental issue about what you see. You're still arguing over Jordan or some other superstars better than Duncan on offense not being the first while what those numbers represented was clearly stated more than once.
Sample size is not big enough. 30 is the threshold to have a somewhat reliable baseline and st. deviation for a measurement. Even for linear calculations.
 
I have clearly moved the conversation away from "Jordan vs Duncan" and onto "O-VORP" and Team Offensive Ratings and how they can interact and compare.
The sample size I am speaking about is a players entire career, which for most superstars we can compare, have over 30 games in the post-season.
Would you like to progress this conversation along? If so do you still have the numbers you posted from the deleted thread?
 
Sure, we can go on. I still have the numbers. Just what I posted though. Not the MS Excel spreadsheet. Can recreate the formula though.
BTW, you might want to take a look at how well BPM and +/- data correlates.
https://public.tableau.com/profile/dsmok1#!/vizhome/BPMvs_RAPM/BoxPlusMinusvs_14YearRAPMThough it was done with BPM 1.0 version, not the current one.
Edit;
Here are the OVORP numbers I deleted.
Here are the rORtg numbers by your calculation method;
There're massive issues with that approach;
- Most spikes are caused by 1 series out of 3 or 4. There were so little amount of teams consistenly posted +7 or +8 offense over their opponents.
- The early '80s has massive spikes due to more drastic / bigger changes in pace.
- It's inclusive of injury related effects, such as Magic in '89 and '91 Finals or DRob in '02 second round, etc.
- There's no way to account for three pointers being way more used. The '80s has insane numbers because with less utilization of three pointers, the teams were stuck within a closer range in regular seasons. The Knicks had -8.3 rDRtg in 1993, the Spurs had -8.8 in 2004, the Celtics had -8.6 in 2008 and there are seasons in the '80s which the top DRtg and the top ORtg had less gap than those numbers. Then playoffs time, the winners in not so competitive series get a massive boost due to that. Because they stayed close in regular season but the gap became more real in postseason.
* Just look at this DRtg distribution over time;
https://i.imgur.com/Zta4CLH.png* And also look at this rDRtg distribution over time;
https://i.imgur.com/Z0epqEF.png- There's no way to eliminate impact of expansion teams causing significant changes in distributions. 
I didn't create a database for those numbers BTW. Just recorded the results. If you have any questions, I should run related numbers again before answering.
Most of those issues are related with Rtg philosophy though. I have no solution for it. If I were getting paid for these things, I would go more in-detail but I don't see no point for doing more than this. Heck, even this is a bit more for a casual fan.