The Spurs are a better team than last year. I don't think it's reasonable to argue otherwise. Pop will eventually realize that starting a bunch of non-shooters doesn't work, and the offense will get back to being respectable. The defense will improve from better personnel and chemistry. Indeed, the West is better now, so maybe they'd miss the playoffs. But 538 recently has projected the Spurs low for a bit now, and they keep having to ratchet up their rankings "as they get more information". While every model will get better with more data, folks are too inclined to give popular models like these a pass and just assume they've already been rigorously tested. Have RAPTORS or any other other stupid acryonyms correctly predicted non-obvious results this far in advance? If not, maybe they shouldn't release them until mid-season or whenever they become accurate.
And no, even the player rankings aren't correct. These advanced stats have yet to appropriately contextualize role-players. It should be part of the brute quan ative results that stars are more important to winning than role-players. You shouldn't look at a list and wonder, "Who's more important, Aldridge or Poeltl." It's ing Aldridge, by a mile. And you should be like, "Is DeRozan less helpful than Mills?" Of course he's not. And that may seem like a trivial complaint, but if they factored in the "star boost", then their models would be more accurate.