Taking out the outliers makes sense in the sense that if substituting one player with another can change a 30-point blowout to a 25-point blowout, is that 5-point difference meaningful at all? I would say no. However, in a close game, if replacing one player leads to a 5-point difference, it could very well lead to winning or losing the game.
It's been posted before -- there is very little difference in opponents on average. I agree that is a factor, but this is the very best side-by-side comparison that can be made. Unless you want to rip open the space–time continuum and play them in the exact same moments in time, then no better comparison can be made. I'm well aware that the Beverley bench lineup is a smaller sample, but it shows the absurdity of you injecting a bench lineup (which also was a small sample) to say that should be our starting lineup. We know Dwight, Harden, Parsons and Jones are starting. Who fits best there is the question, and the numbers so far say Beverley. That's irrefutable. And like I said, Lin still has time to show he is the better fit. You just admitted the only point that is to be made -- Beverley is the better fit next to Harden. Why do I care if Beverley is not as talented of an offensive player if his five-man unit is outproducing on both ends the same four players with a "better" offensive player? What the Rockets are doing in starting Beverley is going with what they feel and see to be their best five-man unit, and having Jeremy Lin off the bench maximizes the time where he can thrive without Harden. It should be a win-win, but we haven't seen this fully healthy lineup very often (only twice since TJones became a starter). For McHale to go with Lin as starter for some elevated "status" in your mind would be simply irresponsible. I do think Lin could post better numbers on another team where he is the show. I know you don't like McHale, but maybe now you get a better understanding of why so many of us are frustrated with fans who don't care what is best for the Rockets but only care about Jeremy Lin.
It sure would be convenient if Toyota Center was built on an ancient indian burial mound... those are always good for that purpose. If the numbers have limited validity due to sample size, then their irrefutableness is a bit beside the point. The numbers seem suspicious to me because the five man unit is performing better offensively when objectively it seems like it shouldn't be. Is it possible that this is a real effect? Yes. But with such a small number of games other effects like injuries, schedule, random chance, etc. could be causing a lot of what is seen. There are a lot of reasons to start Bev; in fact, the starting unit does not have to perform better in total point differential for the decision to make sense as long as the total point differential for the team is better. But I do get irked by strength with which conclusions are expressed based on the games played thus far.
Exactly. It is pointless to use data to justify the decision at this moment. It is what the team or the coach wants. So be it. You can make arguments for Bev to start, but you can arguments for Lin to start as well. Does wanting Lin to start equal to not caring about the team?
I was saying all these stats are not accurate and cannot be used as hard evidence to prove who's better with the starting line up. I'm quite aware as well the bench lineups is a small sample, that is precisely why I used it to prove small sample data aren't accurate. I agree Lin still needs to show he's a better fit, but so does Bev. My opinion was Bev is a better fit on paper, but realistically speaking, we haven't seen any meaningful stats that is large enough to prove it. My opinion is based on eye test. I dislike McHale, yes, because he's stubborn, because he used TT for 10 regular games where from preseason we could have seen it wasn't working. I dislike him because he doesn't have Xs and Os, no set play, so we become a streaky team. I dislike him for overplaying our starters and risk injuries. I dislike him for never admitting it was his fault after losing a game, setting a bad example for Harden to take on leadership responsibilities. I dislike him for allowing Harden to do whatever he wants on court without discipline. BUT, I do not think he's a racist. I like Lin, and I think McHale was harsh on him with a short leash, but I'm not one of those insane Lin fans that goes out of way just to justify Lin's performance.
I apologize, AvgJoe ... I quoted a different poster thinking (or apparently not thinking) it was you. 100% my fault there. Sorry about that.
Not really following why "objectively" the 5 man unit with Beverley should not be performing better offensively.
Because I "object" to that thought? Mostly because based on individual numbers and the eye test, Lin appears to be a better offensive player. It is perfectly possible that adding Beverley to the starting unit results in the unit as a whole performing better offensively, but I need to see a lot bigger sample size when the results are running counter expectations.
Objectively, Lin is going to require more possessions than Beverley, which in turn takes possessions away from Harden/Dwight/Parsons who are the ones who should be taking the shots. This is the most likely reason why it performs better offensively.
I think what you meant to say is since Lin is a better offensive player than Beverley, subjectively, the unit should be better offensively. Eye test isn't a more objective measure of line-up performance than the actual stats itself. And there are a lot of possible explanations as to why the unit performs better offensively. For example, Rockets are a fast paced team, and one of their primary goals is to exploit the defense before it sets. Teams that miss shots or turn the ball over usually struggle more to set their defenses than if they scored the basket. Thus, it's possible that since the line-up is better defensively, they force more misses and turnovers, allowing the Rockets to take advantage of unset defense leading to more points scored. There is a diminishing return as you add more scoring-centric players on the line-up. At some point, having players that specialize in other areas would help the team a lot more. Right now, as much as people like to complain about stagnant offense or Harden ISO's, the team doesn't have a lot of trouble scoring. Their biggest problem is still on defense and rebounding, and that's something Beverley shines at.
You are both right, and those are explanations I would lean to if the numbers hold up over the long term. I don't believe the numbers yet, because there are still other reasonable explanations (e.g. Rocket's brutal schedule, Lin coming back from injury, etc.) More numbers saying the same thing would change that. That said,
As I think about it, Bev has been taking a similar number of shots, and Lin has a pretty high efficiency on his shots. So if the numbers hold up, the more likely explanation is turnovers and transition points from improved D.
At per 36 I have Bev at 9.5 per game, and Lin at 11.5, while Lin is also shooting 4 more free throws which is the equivalent of another 2 shots. There is also 2 extra turnovers than Bev which could be argued is another 2 shots. That makes for almost 6 extra shots (or possesions that could have been shots) for Lin than Bev. That's 6 less shots/possessions for Harden/Parsons and Howard. Now this doesn't factor for their averages as starters, but it is significant.
Your logic has so many flaws that you should really think before you write something. 1. Lin has the highest TS% and 3rd highest eFG% on the team. You really think giving the shots to others would be beneficial to the team? 2. Why is Lin shooting FTs a negative for the team? Can you count how many expected points that yields? 3. Harden and Dwight have a higher number of TOs per36, so how is it obvious that giving Lin's possessions to them reduces team TOs? 4. By your logic, they should take away all of Bev's shots and give them to everyone else in the starting lineup since he is by far the least efficient offensive weapon.
The free throws work out to over 1.5pts/shot, so those are clearly a plus-- but I'll definitely give you the turnovers. Still, given the overall efficiency of Lin's shots this year I would have to think the bulk of any difference would have to be accounted for elsewhere than Lin's shots depriving the other players of shots; probably in the efficiency of the other players rising with Bev as opposed to Lin (via more transition points, for example) combined with turnovers. I'm not saying this is not happening, I'm just saying I don't yet believe it is more than an artifact of small numbers.
Well we cannot be definitive, all we can do is look at the data, understand it's limitations and formulate theories as to why it may or may not be the the case. All we know is the starting lineup has produced better with Beverley in it than it has with Lin.
