O_6 wrote:Doc, I'm not some blind raw On/Off follower. I know it's limitations and I know there are much better ways to measure offensive impact through advanced metrics. But I was responding to someone who was trying to use raw On/Off as a pro-Curry point, I just took it the other way and showed how it could be a plus for Harden.
Rarely is a team's offense completely tied into one player like it is with the Rockets this year. And the raw On/Off just backs it up which is why I mentioned it. Just look at that post I made, Harden has a +16 ORtg and there's not a single teammate of his who played significant minutes who has more than a +1.5 ORtg difference when it comes to On/Off. When the On/Off is THAT obvious, it absolutely belongs in the discussion.
What Harden is doing this year with Bev/Ariza/oldmanTerry as his backcourt support is incredible. He IS his team's offense in a way that we haven't seen since first term Cleveland LeBron.
Also, the 40W-to-60W Curry and 30W-to-50W Harden comment was definitely a ballpark statement.
You're absolutely right that raising a 40 win team to 60 wins is more impressive than raising a 30 win team to a 50 win team. Those extra wins from 50-60 have huge value in "Championship Probability". So even though both raised their teams win totals by 20, the 20 is more impressive the higher the final win total.
But let's say Harden raised his teams raw win total by more? Let's say 25 vs. 20 instead of 20 vs. 20. Would you still give the edge to Curry because his team had the higher record?
Fair enough about the stuff that led up to what I responded. I'll acknowledge that your post was part of what I've been reading, but it's only part of a larger trend. It seems quite clear that if Harden wins the MVP, it will be based on the old "they wouldn't win a game without him" argument. Harden's a great player and a worthy MVP if Curry wasn't there, but that type of argument is basically never right. And I'm not just calling out the hyperbole here, I honestly can't recall a time I've heard that argument used for an MVP candidate where that player in question actually has order-of-magnitude better +/- than the other candidates.
On to the meat of the matter:
1st, the raw number differences in Houston you've noted are possible largely because of the injuries Houston has faced. If a team is healthy, then a starter playing next to a megastar is going to have strong raw +/- as there's no way to shake him. None of this means you can't use the fact that Harden has carried an injured crew as part of the argument for him, but it makes it very problematic to use those raw numbers to compare Harden to Curry. This is in other words, a textbook example of precisely why people should be using some form of Adjusted Plus Minus.
2nd, I'll not that with your used of the raw offensive +/-, you aren't simply using un-adjusted data, you're specifically ignoring the defensive side of the ball. Houston's defense has done considerably better with Harden off the floor. There's literally no reason at all to be pushing this to the side that I can think of.
There's also the matter of saying "He is his team's offense" when the defense is the team's bigger competitive advantage and the offense is mediocre. Doesn't mean it's not true, but being a one-man band raising an offense to mediocrity is something that can be done by a guy who is utterly incapable of leading an elite offense. Not saying that's what Harden is, but it has to be stated how little this whole claim means on the face of it given how much its weighing on you.
What I will say though is this: The injuries in Houston "inflate" the raw +/- edge he shows over his teammates from one perspective, but they also give us more sample size with him functioning with a wider variety of teammates. As such, while all the advanced metrics still have Curry coming out ahead, it is not utterly crazy to say, "I think the team would fair better without Curry if they missed him for an extended period of time than Houston would do without Harden". It is understandable to be more persuaded of Harden's status as absolute keystone
Re: How big of a lift would Harden need to surpass Curry given Curry plays on a better team? I'm with you, no debate here as to the point you're driving at. Obviously Curry can't win this only because he's on the better team. A big enough lift from Harden, and he'd surpass Curry. I don't have a specific number in my mind for this - I don't use these stats quite that literally. My point really is that by all advanced +/- metrics I know of, Curry has the better lift AND does so on a much better team. To see that happen and have people talking about him like he's just a lucky guy with a badass supporting cast is frustrating.
As I said: I really have always liked Harden, but this comparison isn't close for me right now.