OhayoKD wrote:[\quote]Owly wrote:OhayoKD wrote:Which still isn't on par with the 2001 Lakers(+18.39 PSRS) or the 2020 Lakers(+11.70) despite that sample coming against a weaker first round opponent and the warriors having better support around their big two than the lakers had around bron/ad.
Not an expert here ...
On this ... is PSRS points dif plus opponent SRS?
If so doesn't that assume opponents are at RS levels in a small sample in the playoffs?
And if so why are we assuming that one team is exactly at RS levels but another can be vastly better ... does it not simultaneously require strong belief in the RS (for the opponent) and a willingness to cast it aside (to focus on this playoff SRS)?
Well the reason you use the rs is sample size(and you need something as a base). That said, I’m not sure making it more playoff-focused helps. E-balla did some derivation where every playoff was scaled based on their playoff performance and he ended up with the 16 warriors scoring at a +8. Though there was some dispute about where to score the raptors…
I would concur that a smaller sample baseline probably isn't the solution.
So my questions then would be
1) Is there inconsistency in focusing so much on the RS for one team and not at all for the other? (Say, if the '01 Lakers being 15 SRS points better than their RS selves is plausible, isn't a larger gap possible perhaps in the other direction over a single round and doesn't that then undermine PSRS).
2) Won't it oversell weak RS teams who by giving themselves tough routes, need to eke out every last PSRS point just to advance, whilst a dominant 1 seed will have 1 round, perhaps to a lesser extent a second, where their long-term interests might be better served by not seeking to demolish opponents? Indeed wouldn't an otherwise identical [elite] playoff team and performance be better served by this measure putting up a 0 SRS RS than, say, a 9 or 12, as they will not only be required to maximize PSRS more consistently but also mildly (artificially, if they really are that good) inflate all other teams including their opponents?
3) Do we "need" a baseline?
I don't have easy answers. Playoffs are messy ... still in my limited interactions, there seem like significant issues.
