Analytics and a Summary of Clutchfans Being Completely Right

I think they win both Portland games with the right coach and last night's game...

MAYBE the Miami game too...

Those 3-4 games are huge though...

 

I agree with most of your statements here. Yet, for the defense, it seems to me more like a lack of team effect. The Rox man-man defense is about average, but the PnR defense is just terrible. Man-man switching and help defense are always late, 3 men switching is no where to be found. In general it seems to me that they are inexperienced and very confused with their teammate position...
I'm not a big fan of the current NBA stats because ofits systemic error and lack of validity. For example, even the stats show that our SF is not performing well at defense, but really! is it the problem with Parsons >_>. The fact is that he is usually defending or help defending the unstoppable in the other team. When he is helping his teammate while leaving his own man open, the help defend and switch as, I've mentioned before, are no where to be found...

 

You did a halfway decent job with most of your analysis. And then you went and completely screwed it up by torching the coaching staff based just on your own opinion. Before you make assertions like that you really should go back and find coaching records of other teams with as little NBA experience. Their W/L records are really really bad.

 

Ok, I crunch numbers for a living (mostly sport stats) and my pet peeve is people drawing conclusions from small sample size.

No where in the OP's post did he mention that the extremely important caveat that the data sample should be consider too small to draw any meaningful conclusions. That's just statistically irresponsible.

I was just looking at the 82games data the other day and saw that there was too little data on the various 5v5 permutations to make any sort of conclusions, let alone definitive type ones, but had a feeling that someone would make some sort of interpretations from it. And here it is.

That being said, Harden is pretty good. The rest of the team, not so hot.

 

These are not considered high value stats for rating players. In my opinion, the current state of the art is RAPM. You can read a review article of all the major rating systems available today:

http://www.goldenstateofmind.com/2011/12/6/2602153/advanced-stats-primer

Here is the current RAPM for the Rockets players:

http://stats-for-the-nba.appspot.com/teams/HOU.html

Name MP Off. On Def. On Off. xRAPM Def. xRAPM
James Harden 659.08 1.073 1.059 4.66 -0.70
Chandler Parsons 607.08 1.075 1.064 -0.26 -0.25
Jeremy Lin 569.70 1.057 1.072 1.15 -0.13
Omer Asik 548.88 1.059 1.055 -1.82 3.90
Patrick Patterson 479.23 1.061 1.062 -1.77 -0.63
Marcus Morris 344.93 1.103 1.065 -0.40 -0.92
Toney Douglas 260.88 1.110 1.063 -0.56 -1.77
Carlos Delfino 236.77 1.149 0.982 0.24 0.77
Greg Smith 158.70 1.153 1.035 -0.29 -0.50
Daequan Cook 111.47 1.107 1.176 -0.89 -1.39
Terrence Jones 72.72 0.965 1.089 -1.67 -1.27
Cole Aldrich 69.05 0.919 1.113 -1.69 -1.22
Donatas Motiejunas 8.93 0.993 1.644 -1.32 -1.69
Scott Machado 2.57 0.850 1.816 -1.37 -1.43

It should be clear from the list that it's more accurate. Literally no one has every said in seriousness that Toney Douglas has as big an effect on winning as Jeremy Lin. Of course, that doesn't necessarily mean it's wrong. There are lots of cases where the majority's "eye tests" were completely wrong and stats have sussed it out. However, in this case, better analysis IS available, and it suggests Lin has had the 3rd largest effect on winning of Rockets players, which is somewhat astonishing given how he's currently being used by the staff.

 

The problem with the current form of RAPM... or rather, xRAPM, is that it is prior-informed. It means that it will take into account numbers from previous season(s). How much exactly, well, I don't really know. You really need a pretty large sample size this season for it to be meaningful.

 

We need JVG

 

For an example of effective coaching changes...

See the Lakers.

 

You missed the first 6 games of the season? At that point, many posters here thought we played reminiscent of Gundyball. And Fire McHale threads littered the front page before his troubles led to Sampson taking over.

Be careful what you wish for. Because if McHale's return means we go back to the old ways, you'll get exactly the JVG style of basketball.

 

it doesn't take a genius to tell you where we need roster improvements.

2 and 3 is taken care of with Harden/Parsons.

5...Asik is most solid, esp compared to other centers in the league.

PG and PF are our two weakest spots.

no stats..no nothing.

 

It's funny that we had a logjam of PF and yet the PF position is the weakest.

Anyway, crossover, good post.

The caveat is that defensive stats for individuals aren't very reliable. Also, I agree with those who pointed out that Douglas is not a PG. Harden is more of a PG than Douglas is.

 

The coaching is awful

 

I am (and have been) one of Patterson's biggest critics. But I think folks called it accurately when they predicted he would be a "Landry replacement." And as much as I was a Landry fan, that truthfully is such a low bar for a starting power forward in the NBA that we shouldn't be surprised that now that he has become that (Carl Landry).

I don't know if Jones is the answer (seems like a slightly taller Landry/Patterson, or if we should try Greg Smith there until we have more confidence in Donuts. But at some point they need to break out of the "6-8 to 6-9 jump shooting "power" forward" and get someone that is truly "powerful" (rebounds, set's picks, makes other teams pay for going into the paint) or at least creates matchup problems for the other team (ala Dirk, which I hope Donuts develops into).

 

Yes, that is a good comparison. The Lakers went from an innovative coach with fantastic player management skills that was among the most successful coaches of all time to an unsuccessful coach that made questionable game decisions and failed horribly with the players he was given.

Replace the word Lakers with Rockets. Pretty much works.

 

We've had injuries, to both Delfino and then Parsons (which happened to be against Durant....) so the small sample size for SF has been blown up by the best 3 players in the position being in the second half of the small sample size.

 

This has nothing to do with Moreyball, all rookies has this tendency. They do not like to take forced shots and prefer to take shots from their familiar region. IMO, the only thing that could be testament to Moreyball as a strategy for this organization is the W/L. If Morey is an assistant, sure, that would work to judge his work.

Bad defense was certained before the start of the season when Morey made the statement that this would be the roaster for this season.

I think PP is doing fine considering the popularity of decent PFs in this league than any other positions.

Lin is playing terrible. There is no excuse for this though it is partly because of coach. I think he needs to play his own style. His best role is offense game that he played in Knicks, not as a pure ball distributor. The current style put too much offensive pressure on Harden and would hurt the team overall. If nothing is going to change at the end of this season, Lin needs to go to get a pure PG here.

No doubt about this.

 

Agreed. McFail is an utter joke that has no business being around any NBA team, just like Isiah.

 

my favorite rockets coach ever. Should never have been fired in the first place. still lives in Houston and hasn't closed the door on coaching again. what are we waiting for?

 

Pat is horrible! Can't board or set a pick to save his life.

If you're a guard who depends on pick & roll situations for your bread and butter (Lin & Harden), the first thing you need is someone to set a good pick. Pat just simply refuses and instead just rolls off leaving the ball handler stuck with a double team. I can't believe the coaches haven't said something to him about this.

SMH