KingDavid wrote:HotRocks34 wrote:Ok, here's the On/Off advanced stats from Game 3. These are the real stats, not typos. Based on ORTG/DRTG/NRTG
Durant On ---------> +03.4
Durant Off ---------> +09.5
Durant On/Off ----> -06.1
Curry On ------------> +12.3
Curry Off -----------> -33.5
Curry On/Off -------> +45.8
LeBron On ----------> +08.0
LeBron Off ----------> -204.1 (yes, 204.1)
LeBron On/Off -----> +212.1
For the game:
Durant ----> +00
Curry ------> +11 (GSW +05)
LeBron -----> +07 (CLE -05)
I really don't care for this argument, but that stat does not take teammates on the floor into account. It's one thing to have Curry drawing the attention of defenses and playmaking accordingly. But when it's Livingston; who doesn't create offense outside of post ups, Mccaw; who doesn't create offense and/or Clark who doesn't create offense, AND Draymond being in foul trouble limiting his impact and playmaking, then KD is stuck up **** creek without a paddle with that stat. His best skill and job is finishing. The end of the 4th defined that.
It's not the perfect stat, for the reasons you mentioned. RPM is the version of plus/minus that tries to control for the most 'noise,' but it is also imperfect.
But the stat is useful, particularly when it's a consistent finding over the entire regular season, playoffs and even Finals. Of course who is on the court matters. But that is more controlled with the RPM version of the plus/minus stats.
So we can look at this stat, and versions of it, for the entire regular season, playoffs, and the combination of both.
2016-17 REGULAR SEASONDurant On --------> +16.0
Durant Off -------> +07.7
Durant On/Off ---> +08.3
Team --------------> +12.1 [KD -03.8 relative to team baseline]
Durant BPM -------> +08.0 (one stat Durant 'wins' over Curry [or LeBron] in these comparisons)
Curry On -----------> +17.2
Curry Off ----------> +01.0
Curry On/Off ------> +16.2
Team ---------------> +12.1 [SC +04.1 relative to team baseline]
Curry BPM ---------> +07.3
LeBron On --------> +07.7
LeBron Off --------> -08.5
LeBron On/Off ----> +16.2
Team ---------------> +02.9 [LJ +13.3 relative to team baseline]
LeBron BPM --------> +08.4
****
2017 PLAYOFFSDurant on ----------> +19.0
Durant Off ---------> +11.0
Durant On/Off -----> +08.0
Team ----------------> +16.1 [KD -08.1 relative to team baseline]
Durant BPM ---------> +08.9
Curry on ------------> +23.3
Curry Off -----------> -04.9
Curry On/Off -------> +28.2
Team ----------------> +16.1 [SC +12.2 relative to team baseline]
Curry BPM ----------> +11.9
LeBron on ----------> +14.3
LeBron Off ---------> -13.2
LeBron On/Off -----> +27.5
Team ----------------> +10.2 [LJ +17.3 relative to team baseline]
LeBron BPM ---------> +11.5
*****
2017 FINALSDurant on ----------> +14.4
Durant Off ---------> +09.3
Durant On/Off -----> +05.1
Team ----------------> +13.3 [KD -08.2 relative to team baseline]
Curry on ------------> +19.6
Curry Off -----------> -09.3
Curry On/Off -------> +28.9
Team ----------------> +13.3 [SC +15.6 relative to team baseline]
LeBron on ----------> -08.6
LeBron Off ---------> -45.4
LeBron On/Off -----> +36.8
Team ----------------> -13.3 [LJ +50.1 relative to team baseline]
****
2016-17 REGULAR SEASON + PLAYOFFS COMBINEDDurant RPM ----------> +5.78
Curry RPM ------------> +7.57
LeBron RPM ----------> +8.16
Conclusions- None of these stats suggest that Durant has overtaken LeBron
- The report of Curry wanting to step back for Durant in the regular season seems accurate
------------> Read article here:
http://basketball.realgm.com/wiretap/245813/Stephen-Curry-Catered-To-Whole-Theme-Of-Pleasing-This-Season- Curry clearly seems to be the most important Warriors player
- Curry is the only Warriors player, in the playoffs and Finals, who has a negative "Off" rating. Meaning the team loses when he is not on the court
If this information were only true over a certain period of time, or like one stat, then I would doubt what the stats seem to be saying as regards LeBron, Curry and Durant.
Because almost all the information seems to say the same thing, I doubt the conclusions less. Even though none of these stats, by themselves, is perfect.