But what about increasing three point efficiency (one of dozens of possible scenarios). If usage is held the same how can an increase of 3 point efficiency not increase productivity?
Several weeks ago, I said Lin would get 26 as backup, 32 as starter. I'm sticking with that prediction and extending it to Bev: 32 to the starter, 26 to the backup. I still think it's a mistake not to have either Lin or Harden (or both) on the floor at all times: Have one rim protector on at all times; have one attacking guard on at all times.
But there is no law against it. One cannot logically say that there is not a strong correlation between efficiency and productivity. In fact you can drop usage, increase efficiency and still increase productivity. In fact for many players you can drop usage to 0 and increase productivity.
A player who scores 2 points by going 1 for 1 from the field is not more productive than a player who scores 40 by going 20 for 25 from the field, even if he is more efficient. It's not the same thing.
the question is - will there be enough minutes without harden for him to up his usage rate? if harden plays like how he did last year, the answer is no. but i guess he had to play that many minutes last year because there was no viable backup, but now, theoretically, there is. not sure why they never figured out to stagger lin and harden's minutes last year but that's another discussion.
keep in mind, this is not the nyc offense where the pg brings up the call, call for the pick, and starts the attack and decide where to shoot, drive, dish to roller, and kick out for 3. in this offense, anyone can bring up the ball and the attack usually doesn't happen until the ball is passed off to someone coming off of a screen, and that's usually harden or parsons. without harden on the court, it would be lin. but of course, you can't expect every single possession to start with lin pnr - who would want to play with that?
good point. probably because he saw how explosive they can be when they're clicking (50+ pt, 20+ ast between the two), and wanted to give that combo as much pt as possible. this year, with dwight and much stronger D, they can live without the offensive potential and go with what they know they can get consistently.
jtr, my point is simply that the studies relating Usg vs. Efficiency do not necessarily translate to Usg vs. Productivity or Effectiveness. What Dean Oliver argued in his book is that players tend to lose efficiency as their usage goes up, but its not linear and for some highly skilled players (e.g. Jordan) the dip doesn't become pronounced until the usage exceeds a particular threshold. As for Usg vs Productivity, yes after a certain point we can all agree that a player's productivity/effectiveness/value will start decreasing if you keep forcing the ball into him. But before that point, a player's productivity can actually go up with usage. Trivially, a player whose usage is close to 0 is producing practically nothing. His production can't help but go up with increased usage (up to a point). Hope that's clear.
Thank you for the explanation Durvasa. It looks like a terminology convention as opposed to mathematics, and I can respect and adhere to it.
The Rockets actually looked really good with the Lin-Beverley lineup for most of the preseason, with Lin being the primary initiator on offense in that lineup. When McHale didn't alternate starters in the final game, I think it went to Beverley's head a little and he started initiating the offense more and played a lot more keep away from Lin in the few minutes they had together. The Rockets did not look good at all doing that. It was just one game, but I don't think it would be wise to summarily dismiss this concern as pure paranoia. If what happened in that stretch continued into the regular season, I would give this experiment until January before McHale would be forced to make a change again.
The relationship between usage and productivity depends on the type of player. If he is the type that is more effective with the ball in his hands, then the correlation can be positive to an extent. If it isn't an 'on ball' player, then his usage will have nothing to do with his productivity. For instance, if you increase Asik's usage, his RAPM will likely be hurt. Beverley has an usage that is much lower than Lin, yet he produces the same RAPM, because he impacts the game without the ball.(that's why he is starting) For this thread, first I'd like to track Jeremy's usage rate this season, to see if indeed he will play a different role off the bench. Second, I'd like to see if there is merit to the narrative that "he is better with the ball in his hands". Is it true, and by how much? His RAPM rating this year should reflect that.
Most of the elite PG's cannot be shut down by one guy alone. Bev's domination over Holiday in the pre-season was a freak-occurrence. So long as Beverley keeps pestering them, that's enough, he doesn't have to take them out of the game. That's not very relevant though, because that's not why he's starting. Anyone that watched the Rockets last year will agree that Lin and Harden's minutes need to be staggered so at least one of them is on the floor at all times. The offense stagnates when they're both sitting and their overlapping skill-sets limit their potential effectiveness when they're playing together. That's not to say we shouldn't play them both at the same time, cause we'll be doing that a lot, but it's pretty clear that our best strategy is trying to always have at least one of them fresh and fit to run the offense. Lin's "benching" is simply McHale implementing this rotation with efficiency instead of marketing on his mind. People seem to forget that McHale himself spent a lot of time as a 6th man despite being considered one of his team's big 3. Lin isn't being a b****, he's just intelligent enough to realize that being utilized effectively is more important than a meaningless (outside of marketing reasons and whiny people who don't know basketball) starting spot. Is coming off the bench and playing 30 minutes worse than starting and playing 25? Did anyone in Oklahoma consider Sefolosha a better SG than Harden? Will Manu be remembered as a random bench-warming scrub when this Spurs core is done? Everyone gets to sit on the bench during a game. As long as it's not the final minutes of a close one, why would anyone care specifically how their bench minutes are allocated?
I am not as skeptical as you are that the Rockets will do better putting the ball more in Lin's hands. I have already seen how much better the Rockets do when they put the ball in Lin's hands last year in the 15-game stretch after the San Antonio loss when the Rockets routinely blew out playoff teams doing just that, and that was not even staggering minutes with Harden, but playing alongside Harden. I am, however, just as skeptical as you are that a bench role will mean that McHale will put the ball in Lin's hands more. We will have to wait and see if that is actually true. Judging from the Casspi pairing, I would say that McHale is actually trying to avoid putting the ball in Lin's hands even when Harden is not on the floor. Had it been D-Mo/Smith on the first substitution in with Lin, I would have been more convinced otherwise. Lin is much more of an interior passer than a kickout passer. If McHale is running a small Beverley-Lin-Casspi-Parsons/Garcia-Howard/Asik lineup, then I am inclined to think that McHale wants to optimize for taking the ball out of Lin's hands and put it in Beverley's hands, as Beverley is much more of a kickout passer than Lin is. Casspi isn't much of a screener or interior/roll target.
This actually in my opinion holds more weight than people give it credit When Lin is on, hes very dangerous, drives heavily to the basket and forces the action. When he isn't, well, he seems to ghost especially when he isn't the primary ball hander. He just needs to be aggressive in all of the time he gets on the court.
I would say its not a matter of mere terminology. The name "offensive rating" is not actually rating how productive a player is on offense. Players who are asked to not do much of anything on offense can have a very high ORtg while still not helping their team much on that end. RAPM is actually capturing total contribution to the team's offense, while ORtg is capturing efficiency. Conceptually, it is different. You can convert ORtg/DRtg into a production stat by combining it with Usage and considering the ORtg/DRtg and Usage of other players on the team (that is essentially what Win Shares does), but (again) by itself it is just efficiency.
http://www.thedreamshake.com/2013/1...ould-start-over-jeremy-lin-at-point-guard-for Not agreeing with the choice to start Bev, cause I don't (although I am keeping an open mind), but a pretty good run down of the reasons why Lin could thrive with the ball in his hands.