@Vox Rationis is correct.
You're mixing apples and oranges: you're comparing an NTRP
level [4.0] with the NTRP
number tied to that level [which is defined as 3.50 - 3.99]. So your statement "You are 4.5 when you reach 4.5." is incorrect: you are an NTRP 4.5 when you reach a rating of 4.00. If you reached a numerical 4.50, you'd be an NTRP 5.0.
So the SoCal player with an estimated 3.97 means he's at the very upper end of the 4.0 NTRP range and is within a hair's breadth of being an NTRP 4.5 [defined as 4.00 - 4.49].
Your comparisons are off by one level.
The "actual" level is the sum total of
all results. If the algorithm threw out the bottom and top 10%, then I'd agree with you. But there's no evidence that this is the case.
Could you come up with an explanation that justified those stellar results without him being significantly under-rated? Sure. I'd prefer the simpler explanation, a la Occam's Razor: he's under-rated.
The math didn't seem fuzzy to me at all. And your analogy is shaky: one's age increments in a predictable fashion. One's estimated NTRP is all over the map: you could play down a level one day and up a level the next.
You seem to be going to great lengths to avoid the possibility that there are significantly above-level players at Nationals [not necessarily a lot but enough to skew the predictions; just ask
@schmke.
I get mine from tennisrecord.com [t e n n i s r e c o r d.c o m, in case the filters kick in].
BTW: I enjoy facing stiffer competition. Winning is not the most important thing to me. If a closet 5.0 sneaks into my 4.5 league, I want to be the one to play him and if I get double bagelled, so be it; I'll try and enjoy the "lesson" and learn something from it. So I'm not making my arguments because I'm whining about losing to sandbaggers. I'm making my arguments to correct what I see as a flaw in your logic.