LA Bird wrote:Odinn21 wrote:It's the distributions. That San Francisco team without Chamberlain is quite literally in the bottom 1% in the '60s and you can't expect an individual's impact to act like as if it is in 1σ range. It's not linear you know.
This is the scatter plot of all WOWY runs with at least 10 games missed in ElGee's spreadsheet:
Where is the evidence it is harder to impact a bad team than an average team? That line isn't flat (linear) and it isn't upward sloping for x < -3 (which is what you are claiming). It is downward sloping because it is generally harder to add the same value to a better team. Besides, even if it was harder to lift a really bad team, a +2 on a -6 team is still not a big deal - West, Oscar, Thurmond all have done a lot more with similarly bad teams in that era.
Those were definitely on similarly bad teams. That graph you added is a fact, because it's done through observations but doesn't mean it's complete and reliable enough to build on. There's a lack of estimations for roster quality and structure, also a lack of predictions about how much they could be improved, etc. There are some teams, the very worst of the worsts, some of that + value a player can bring is spent on making the team improvable. My own personal stochastic model applied on MoV numbers with prior weightings is saying that '65 Warriors team was one of those very worst teams.
LA Bird wrote:You asked me what if we pretend Wilt was in Philly the entire year. But the answer is that 65 Wilt wouldn't be any more impactful in a hypothetical scenario where he wasn't traded midseason. He would be the same player but we just won't have any WOWY numbers to show him being only +2 or +3 that season.
Wilt was leading the Sixers to a pretty solid record before big injuries crippled the team, that's not any less legitimate than you mentioning Olajuwon's W-L record in '95. I still stayed close to 18W-17L record (which was basically ruined with piling injuries in the 2nd half) in my assumption.
It's hard to see such insistence on Chamberlain being a +2 or +3 player in that season given the nature of his season.
LA Bird wrote:Sorry but the only person calculating WOWY with a disregard for injuries is you in post #34.
I already stated I missed the injuries and provided further numbers in post #38.
LA Bird wrote:I assume you haven't seen ElGee's spreadsheet because it says right in the sample controls that he only focused on the games with Greer and Costello in. You didn't realize it but the <3 SRS change I said in my very first post already took into account their injuries.
I saw ElGee's spreadsheet, I take it that the players ElGee put in Sample Controls were excluded players. Maybe I missed some detail in there about '65 Sixers.
I didn't realise it? You said nothing about injuries in that post. No wonder.
I mean Chamberlain was playing on a historically bad team before, that team was in the bottom 1% without him and they were in the bottom 35% with him, (on a scale that's normalised for jumps on distributions, that's a lot more than a +3 player, this is what I was basing on my argument), then Chamberlain gets traded to a decent team, that team does very well at first, then goes to sh.t with injuries and it's similar to the situation in SF without the missing games to compare, and with all of these we
know from WOWY, a questionable +/- substitute, that Chamberlain was a +2/+3 player in 1965. Because he quite possibly had the most wild variance throughout a season for a player...
I think it's all there is for this done comparison though. Time to move on.
