From above, we see that the relationship between height and team defensive efficiency is not so clear, indicating that there are many other variables (team philosophy, experience) to be considered. The Houston Rockets, as poorly as they’ve defended of late, are actually not too far below league average (though note that the data for the chart above was from before the trade). Still, they have clearly slipped, and the absence of an interior shot-blocking presence has to be the considered the primary reason. But there are others:  many new players added to the team is a factor, as well as the Rockets playing more younger players.

Here is another look at this relationship, showing individual front-court players (who’ve played at least 500 minutes ).

“DEF on” is simply the team’s defensive efficiency with each player on the floor, and “DEF On/Off” is the team’s defensive efficiency with the player on the floor minus the efficiency with the player off the floor. Note that the lower either of these metrics are, the “better” the team defense is with that player (presumably, of course this is a bit of a simplification).

Conclusions? There is a very minimal correlation between increased height in the front-court and better defensive +/- stats, but really the only conclusion one can draw is that  (again) height is not a strong indicator of defensive impact. We see that there is actually a somewhat stronger relationship between increased age and defensive impact, which isn’t a surprise.

The problem here is that height is not necessarily a useful indicator of how much “interior presence” a player provides due to his length. What we can do is combine height and shot-blocking prowess into a single metric and see if there is a stronger relationship between that and defensive impact. We will explore this issue further in coming weeks.

For those readers that are unfamiliar with what adjusted +/- is measuring, think of it this way. Suppose you have a particular 5 on 5 matchup, and you want to know what the expected point differential would be after some number of possessions. The idea is to come up with a number which describes each player’s +/- impact, such that if you add up all these numbers for one side and subtract all the numbers from the other side you’ll arrive at this expected point differential. Now in reality basketball isn’t so simple, and sometimes certain player combinations works better together or certain matchup are more favorable than others. And there are many other variables to consider. Still, if we knew for each player what those numbers were on average, that could be valuable input in deciding which players to pursue in an acquisition or which players to give more playing time to.

Alright, so what is offensive adjusted +/- and defensive adjusted +/-? Think of offensive adjusted +/- as the number of points a player is expected to add to his team’s score (per 100 possessions) by being on the floor, relative to a “league average player.” Similarly, defensive adjusted +/- would correspond to the number of points a player is expected to add to the opponent’s score (per 100 possessions) by being on the floor, relative to a “league average player”.  So, the higher the offensive APM the better, and the lower the defensive APM the better. Note that when I say “points a player is expected to add”, I’m not simply referring to the number of points he scores individually or the number of points the opponent he matches up against scores individually.

I’ll give the usual “stathead” disclaimer: these are interesting to look at, but are not by any means the end-all, be-all player rating. For most players, the ratings makes some sense and can serve as a sort of validation of our subjective opinions or what other methods (e.g. PER) might say. For some players, especially when only looking at data over just 1 year, there are some strange results, but that is to be expected. The model is not perfect, and there is noise in the data. It is interesting to look at nevertheless.

Below, we’ll take a look at some of the results. Again, refer to HoopNumbers for all the raw data.

League-Wide Trends

I want to start by looking at some league-wide trends. HoopNumbers provides RAPM, Offensive RAPM, and Defensive RAPM for each of the last 4 seasons, as well as combined numbers over all the 4 seasons (“equal weighting”), and combined numbers that are weighted more heavily to recent seasons (“time-weighted”). Consider the following scatter charts, relating RAPM results to PER for the last 4 seasons:

Average over last 4 seasons, minimum of 1000 total minutes played

PER is referring to Hollinger’s Player Efficiency Rating. For those that are unfamiliar with it,  it is essentially a per-possession summary metric that adds the good things a player does and subtracts the bad things a player does, according to the boxscore. We see that PER has almost no relationship whatsoever to Defensive RAPM, but there is a clear relationship to Offensive RAPM.  This makes sense. The boxscore, as you probably know, is heavily offense-centric. That isn’t to say you can’t do any better than PER at explaining defense with the boxscore, but there are obvious limitations.

Now, let’s look at the same charts, focusing on the current season:

Minimum 1000 played this season.

And here is the chart focusing again on current season, but the RAPM is stabilized by taking into account data from previous 3 seasons (these are the “time-weighted” RAPM stats):

Minimum of 1000 minutes played; RAPM stabilized using data from prior 3 seasons.

We see a similar pattern, as expected. Out of curiosity, I wanted to take a closer look at the “top 30 players” according to PER. What sort of RAPM do they have?

RAPM (4 year time-weighted) vs PER for top 30 PER players (yes, top 30, chart has typo).

I labeled the points with each player’s name. This type of presentation is interesting to me, because I see RAPM and PER as mutually validating eachother. RAPM is based purely on how a team’s performance changes with the player on the floor, while PER is based purely on the numbers the player puts up individually in the boxscore. If a player rates highly in both, and many of the above players do, that gives me more confidence in affirming their status as an elite performer. Here is another chart, again for the top 30 PER players, showing Offensive and Defensive RAPM (4 year, time-weighted):

Offense-Defense RAPM Map (4 year, time-weighted) for top 30 PER players; minumum 1000 minutes played (yes, top 30, chart has typo).

Again, note that the lower the Defensive APM the better. So players falling in the lower right quadrant have an “above-average” offensive and defensive RAPM. Players falling the upper-right quadrant are above average in offense, below average in defense according to RAPM analysis.

RAPM Stats for 09/10 Houston Rockets (the important ones)

Below are some charts for the Houston Rockets players that have played significant minutes this year (I include Kevin Martin):

The ratings for just this season are not so reliable, so I wouldn’t make so much out of Andersen’s good defensive rating, or Martin’s poor offensive rating. I’d consider these anomalies. By HoopNumber’s +/- analysis, Chuck Hayes rates as one of the top defensive player over the last 4 years in the league, and even for this season, in a starting role, he has the best defensive rating on the team, closely followed by Kyle Lowry. Some may doubt that a 6’6 center can rate so well defensively. I would ask those people to just take a look at Chuck’s performance against Toronto last night.

Interestingly, this analysis finds Battier to be a below average defender this season, and only slightly above average over the last 4 years. In fact, it gives more credit to Battier’s offense than defense, which is hard for me to rationalize so I don’t think I’ll bother.

Scorer-Assistant Table. Shown for 9 Rockets rotation players before All-star break.

Perhaps a more telling way of representing this information would be to show assist-rates when two players are on the floor together. For example, instead of showing that Brooks assisted 51 Ariza baskets so far this season, I may want to know how many assists to Ariza he gets for every 36 minutes they share on the floor.  That’s what the following table shows:

Scorer-Assistant Per 36 minutes Table. Shown for 9 Rockets rotation players before All-star break.

A few comments:

• Kyle Lowry does a nice job distributing to all his teammates when he shares the floor with them.
• Carl Landry is strictly a finisher.
• It is interesting that, per-minute, Chuck Hayes assists more Brooks baskets than any player other than Lowry. Brooks has the ball in his hands quite a bit and has to create most of his shots, but he does get some easy looks from Hayes backdoor passes or hand offs from the high post
• It is interesting that Brooks has a higher assist rate to Landry, while Lowry has a higher assist rate to Scola. Brooks does a pretty good of penetrating in half-court sets and finding Landry on the baseline. I think Lowry isn’t asked to create as much offense in half-court situations. Lowry does excel in transition, and Scola is a great running partner. Perhaps those pairs are better suited together?

Data in this post was tabulated from the play-by-play CSV files provided by Ryan Parker at BasketballGeek.com.

Below, I plot a summary of this data for the average NBA player and the average Rocket player.

Average NBA Player vs Average Rocket (per 48 minutes)

From this, we see that the Houston Rockets get more shots at the rim than the average NBA player, though they do not make proportionately more of them. The Rockets are only 27th in the league in FG% at the rim (56.7% versus 60.6% for league average). Another feature that can be observed from these curves is the Rockets, as a team, like shooting threes, and they dislike shooting long twos. From an efficiency stand point, this looks to be a smart strategy as the Rocket effectively shoot 51% on their 3-point attempts but only 38% on long twos.  The league as a whole shoots better on 3s (eFG% is 53%) and only 39% on long twos, but they attempt more long twos.

Point Guards

Now, I’ll break it down by position, starting with point guards:

NBA PGs versus Rockets PGs (per 48 minutes)

We see here that the scoring-oriented Brooks excels as a 3-point shooter, making effectively 59% of them. No other player on the Rockets is as adept at getting his own shot from the perimeter, which is reflected by the red color showing up more prominently in his chart. Brooks gets a lot of shots at the rim, which in itself is impressive for a player of his small stature and slight build, but his conversion rate of 50% is below average at his position (the average NBA PG hits 56% of his attempts at the rim). His assist rate is less than the average PG for all ranges shows here. The midrange game is where I see Brooks improving over the next few years. His size will always limit him at the basket, but as he improves his sense of pace with the dribble and gets more comfortable taking the pull-up jump shot, we should expect to see more shots coming in the 10-23 feet range. Even though there’s no chance he would be able to match his 3-point efficiency with these midrange jumpers overall, the ability to generate such a shot on a consistent basis will make it more difficult for opposing teams to defend the Rockets.

We can see that Lowry shoots less than the average PG at all ranges except at the rim. I expected Lowry to be a better midrange shooter for the Rockets, but he’s actually taken more 3s and shot them more effectively (eFG% of 44%) than he has midrage shots.  He is by no means a good 3-point shooter, but he’s hit enough of them to be a threat.  Lowry’s assist-rate on 3-pointers leads the team by far, and is well ahead of the league average for PGs. In fact, Lowry is 5th in the league in assists for 3-pointers per minute among players who’ve gotten significant minutes, trailing only Chris Paul, Steve Nash, Jameer Nelson, and Jason Williams. Not shown here is rate of drawing fouls and getting to the line, something Lowry is particularly good at for his position.

Swing Men

Next, we take a look at swing men (SG/SF):

NBA Swing Men versus Rockets Swing Men (per 48 minutes)

The much maligned Trevor Ariza beats out the average NBA swing man in scoring at the rim, though the conversion rate is only 56% versus 60%.  Ariza attempts a lot of threes, and shoots 45% on them (effectively) compared to 53% for the average swing man. Ariza will attempt long-twos, but he shoots only 31% on them compared to average 39% for his position.  The Rockets rarely run plays for him shooting off of screens which means that a vast majority of the shots he gets from the long-two range are unassisted, so the results should not surprise.  One positive I’ll make note of is Ariza’s assist-rate on shots inside of 10 feet eclipses the other two Rockets swing men and is above average for his position.

Battier’s chart is interesting to look at.  He takes almost no long twos in the Rockets offense, acutely aware I’m sure that these are typically low efficiency shots. The majority of the points he generates from the field come on assisted 3-pointers. He shoots the three at a slightly-above average clip for his position, though considering virtually every one of his attempts is assisted I expect him to shoot it better than he has. The Rockets will occasionally run a post up play for him when he’s being defended by a smaller guard, and at times he takes it upon himself to drive to the basket and get a shot up. He’s not especially effective at either, but it does help for him to be a more active participant on offense.

Chase Budinger is a gun-slinger who is 13th in the league in 3-point shots attempted per minute, though his conversion rate is just average for his position.  Unlike Ariza or Battier, Budinger will frequently use screens to free himself for long twos, and he hits a very healthy 50% of his shots from that range. At the rim, we know that Budinger has the size and athleticism to finish well in traffic, but he’s only converted on 51% of them compared to 60% for players at his position. Budinger seems to have a good feel for the offense and handles the ball fairly well, but he has not distinguished himself as a passer. Part of that might be due to his quick release. If he gets the ball and he has any daylight he’s shooting it, which limits the number of “system assists” he may otherwise get.

Bigs

Finally, charts for the “bigs” (PFs and Cs):

NBA Bigs versus Rockets Bigs (per 48 minues)

What distinguishes Scola is he takes almost twice as many shots (per minute) from the 16-23 foot range than any other Rocket, with almost all of them being assisted. Scola hits 41% from this range, which is actually not much better than what the league average PF/C will shoot (40%).  Scola has a relatively show release, and consequentially when defenders rotate to him he’ll try to pump fake them and then either draw the foul, go around them, or pass back out to a 3-point shooter. Scola lacks the ability to put the ball on the floor and then pull up from midrange. Within 10 feet, Scola has an effective post game with his primary weapon being a right-handed hook shot, and his secondary weapon being a right-handed scoop shot.  He’ll turn left now and then, but will invariably turn back the other way to get a shot off with his right hand. He has no turnaround jumper in his repertoire, which may be for the best because he lacks the length and leaping ability to shoot that shot over an outstretched defender.

Chuck Hayes shoots almost exclusively around the basket, and I can only recall him attempting one jump shot in his entire NBA career (suffice it to say, it wasn’t pretty).  I was actually surprised to see that Hayes shoots more often at the rim than Scola. Per-minute, his rate of shot attempts at the rim has actually gone up over twice as much since last year. Chuck hits on only 48% of these shots, however, well below the 63% that the average PF/C will convert.  Hayes does have a higher assist rate for shots at the rim than any other Rocket big and is fairly above average amongst NBA bigs (which isn’t bad, considering none of the players he typically plays with are great finishers). He slips bounce passes to cutters out of the high post better than any other player on the team.  Hayes also has an above-average assist rate for 3-pointers. He’s especially good at tracking down offensive rebounds and quickly passing out to an open 3-point shooter before the defense recovers.

This season, Carl Landry has emerged as not only the Rockets best inside scorer, but one of the best interior scorers in the league. Per-minute, Landry is third in the league in field goals made inside of 10 feet.  He is the only player on the team whose FG% at the rim significantly surpasses the average at his position (69% for Landry, 63% for the league average big).  Landry’s assist-rate is easily the lowest on the team and is bottom 20 league wide, but that has as much to do with his potency as a scorer as it does with any lacking court sense on his part. One area where Landry has actually regressed compared to his first few seasons is shooting from the 16-23 feet range. He’s shot only 30% on them this year (average for bigs, again, is 40%), compared to previous years where he was close to 50%.

David Andersen was brought in to be a floor-spacing big, and he appears to have satisfied that role well enough. Rick Adelman likes to pair him with Carl Landry, which probably contributes to his lack of scoring at the basket (and, likewise, contributes in part to Landry’s inside scoring). Andersen shoots a good percentage (over 45%) from 10-15 feet and 16-23 feet, but he has struggled on his 3-point shot (he shoots them for an effective 41% conversion rate, which is very poor). I expected Andersen to be used more in the high post, in the role Chuck Hayes often finds himself in, but the Rockets don’t seem to run that set with him very often. This is strange to me, because he seems to have a versatile offensive game from midrange compared to our other bigs. I’m hoping that in time he becomes a more potent shooter/passer from this position.

Histogram for blocks per 100 possessions.

Histogram for drawn charges per 100 poss (from HoopData.com)

Kyle Lowry leads the team in drawing offensive fouls (and is second in the league), but most of the fouls he draws are from illegal screens away from the basket. Last year, the Rockets drew about 2 offensive fouls per 100 possessions and had a block rate of 4.6, which would put them in the middle of the pack in both categories this year.

The red line shows fast break points per game over the last 3 games, the blue line is fastbreak points per game since the beginning of the season, and the green line is our fast break points last year (at 8.2, we were dead last by a significant amount). Currently, we are ranked 8th in the league in fastbreak points, which is great though during the first month and a half of the season we were top 5. As you can see from the chart above, there was a lull in a fastbreaking for much of the first half of January. But the Houston Rockets have picked it up during the 6-game homestand.

Here is a list of the top players so far this season based on this statistical +/- model:

Player              Team     PS     GP     mpg      SPM
james,lebron        cle      SF     48     38.6     13.6
wade,dwyane         mia      SG     47     36.4     9.7
paul,chris          nor      PG     38     38.8     9.2
duncan,tim          san      C      42     32.5     7.8
howard,dwight       orl      C      47     35.1     6.8
ginobili,manu       san      SG     40     26.7     6.7
bryant,kobe         lal      SG     47     38.4     6.7
smith,josh          atl      PF     46     33.7     6.5
durant,kevin        okl      SF     46     39.9     6.2
roy,brandon         por      SG     40     38.2     5.8
love,kevin          min      PF     28     30.2     5.6
billups,chauncey    den      PG     38     32.7     5.6
anthony,carmelo     den      SF     38     37.9     5.5
kidd,jason          dal      PG     46     36.1     5.5
camby,marcus        lac      C      45     31.1     5.5
gasol,marc          mem      C      46     35.6     5.2
wallace,gerald      cha      SF     45     42.1     5
oden,greg           por      C      21     23.9     4.8
gasol,pau           lal      PF     30     36.4     4.8
davis,baron         lac      PG     46     34.2     4.7
rondo,rajon         bos      PG     43     36.9     4.7
granger,danny       ind      SF     30     36.3     4.6
nowitzki,dirk       dal      PF     46     38.1     4.6
bosh,chris          tor      PF     47     35.9     4.6
hilario,nene        den      C      46     34.3     4.2
arenas,gilbert      was      PG     32     36.5     4.2
pierce,paul         bos      SF     39     36       4
wallace,ben         det      C      45     30       3.9
frye,channing       pho      C      48     29.8     3.7
williams,deron      uta      PG     41     37.1     3.7
wallace,rasheed     bos      C      41     24       3.6
murphy,troy         ind      PF     37     31.1     3.6
kirilenko,andrei    uta      SF     40     29       3.6
boozer,carlos       uta      PF     45     35       3.6
lowry,kyle          hou      PG     46     25       3.5
iguodala,andre      phi      SG     46     40.1     3.5
martin,kevin        sac      SG     14     36.2     3.4
odom,lamar          lal      PF     47     31.3     3.3
harden,james        okl      SG     46     22.9     3.3
noah,joakim         chi      C      44     32.9     3
anderson,ryan       orl      PF     39     16.3     2.8
fernandez,rudy      por      SG     30     22.9     2.7

Here is the list for the Rockets:

Player              Team     PS     GP     mpg      SPM
lowry,kyle          hou      PG     46     25.0     3.5
battier,shane       hou      SF     46     33.1     0.8
ariza,trevor        hou      SF     45     38.2     0.6
hayes,chuck         hou      C      46     21.9     0.5
landry,carl         hou      PF     45     27.2    -0.4
brooks,aaron        hou      PG     46     35.4    -0.5
scola,luis          hou      PF     46     30.8    -0.8
budinger,chase      hou      SF     38     18.6    -1.9
andersen,david      hou      PF     46     14.2    -2.3

There are some surprising results for the Houston Rockets players. Lowry leads the team by far. The two rookies (Budinger, Andersen) are at the bottom, which actually shouldn’t be a surprise. Everyone else is between -1 and 1.  While having a rating less than 0 is technically “below average”, keep in mind that “average” in this case is much higher than the statistical median for all the NBA players. That is, there are far more below average players in the league than above average players. An average player corresponds to a player who is basically starter-quality. If you are familiar with John Hollinger’s PER system, this would be a player with a 15 PER.  Compare these results, based on boxscore indicators, to the results Wayne Winston had for the Houston Rockets as of his January 20th blog on the team.  It may not be obvious from the lineup data he shows in that link, but it is possible from the information he provided to arrive at his adusted +/- ratings for each of our 9 main rotation players — both at the offensive end and defensive end. I show this graphically below:

This depicts offensive and defensive adjusted +/- for 9 Rocket players.

Left to right, players go from bad to good offensively. North to south, players go from bad to good defensively (the more negative, the better).  Winston’s results differ substantially from what the statistical +/- would predict for some of the players, in particular rookie David Andersen.  I tend to trust the results from the statistical +/- formula more in this case. Winston has Andersen as rating well above average both on offense and defense. But Andersen, in my humble opinion, is the worst defensive player the Houston Rockets have had since Steve Novak; I can only assume that Winston’s method is allocating defensive credit to Andersen at the expense of other Rocket players (especially Landry and Budinger, who rate as awful on defense according to Winston). Winston’s results on the offensive side make a little more sense, subjectively.

Returning to the statistical +/- results, it is a surprise to see both Battier and Ariza rate so highly. The SPM model likes players who take a lot of threes (which they do), probably because such players tend to spread the floor, leading to more efficient team offense. I had thought that Ariza’s poor scoring efficiency would sink him for this rating system, but it turned out that he does enough other things well according to the model (particularly his high 3-point attempts, his steals, and his versatility index) to end up rating above average. Remember before when I said this is not a fool-proof system? I believe Ariza is a case where his offensive ineptitude is not being well-captured by this model; this is a case where Winston’s assessment rings more true.

To close, below I show a histogram of SPM ratings for all NBA players who have played at least 500 minutes:

SPM Distribution for all NBA players (Rocket players highlighted)

Again, 6 Rockets fall between -1 and 1 in the histogram above. The two rookies have less than a -2, which may after all be appropriate (though I think Budinger has shown the potential to become a starter-quality player some time soon).  According to statistical +/-, Kyle Lowry has separated himself from the pack. He does not score a lot of points, and he does not shoot (or make) a lot of threes. But, by the box score, he does everything else really well. He gets to the line, he gets a ton of offensive rebounds (leads the league amongst PGs), he gets a good number of steals, and he has a high versatility index. And this doesn’t even account for the fact that Lowry leads the league in drawing offensive fouls. Has he been the Houston Rockets’ best player this season? I am not sure I would say that, but I do think he’s been the most important Rocket in regaining lost momentum from the starters’ frequent lulls to start each half. And I think many people, especially the national media at large, have under-appreciated his impact.

The red line represents league average true shooting% (about 54%).  It is rare indeed when Ariza is able to match this number; he has done it only 9 times this season.

And here we show the distribution over these games:

The “norm” curve represents the expected distribution of TS% for 45 games, the number of games Ariza has played thus far, based on season average TS% (about 46%) and the standard deviation (about 13%).  The green curve represents the corresponding expected distribution for the following 6 comparable players: Richard Hamilton, Rudy Gay, Luol Deng, O.J. Mayo, Andre Iguodala, and Caron Butler. They were chosen because they are the 6 wing players who’ve played at least 1000 minutes this season that get the closest number of scoring attempts per game to Ariza (about 17).  Over the games for these 6 comparable players, the mean TS% for them was 52% and the standard deviation was 13%. You can think of the green curve as representing the distribution of TS% we should expect from Ariza considering how often he shoots.

For our “Ariza Watch” in the 1st quarter, I observed some good and bad. At 10:30, Ariza got himself wide open on a cut to the basket, but Shane Battier missed him and elected to dump it down to Scola instead (who also missed Ariza open at the basket). Scola ended up scoring in the post, but those are the type of opportunities the Rockets should take advantage of. Shane could have given Ariza an easy scoring opportunity with a lob. At 9:50, Ariza faked the 3-point shot and drove past his defender, hitting Battier in the corner. The shot was missed, but it was a well-executed play.

At 7:10, Ariza received the ball against Charlie Bell from Luis Scola. Having 5 inches on his opponent, this is a matchup Ariza should (in theory) take advantage of. But I sensed right away that it was a poor play, knowing full well that Ariza is a weak finisher when contested and has absolutely no touch around the basket unless its a point-blank dunk/layup. As it turned out, he never even got the shot off — he turned and fumbled the ball away. At 4:20, Ariza was tested defensively for the first time. Bell executed a V-cut on the baseline, cutting back to the corner to get the ball from Jennings. In these situations, Ariza likes to cheat over the screener and intercept the pass instead of following his man back to the 3-point line. In this instance, Ariza just missed getting a hand on the ball, leaving Bell with a wide open look in the corner which he missed. Is this the type of defense the Rockets want him to play? At 2:55, Ariza made a sloppy pass to a cutting Hayes that was intercepted. He fouled Warrick hard in transition, in apparent frustration.

Lowry and Landry checked into the game with about two and a half minutes remaining in the first. Landry has gotten the publicity as a 6th-man of the year candidate (a topic I wrote about last month), but Lowry may rightly be considered the true Rockets energizer off the bench. He entered the game with the Rockets trailing by 6 points, but the Rockets outscored the Bucks by 13 through the rest of the first half. To end the first quarter, Lowry freed up Brooks with a great screen in the corner, found Hayes cutting underneath for the layup, got a defensive rebound and passed ahead to Ariza for a transition layup, and he drew an offensive foul by strategically careening off Bogut. He continued his all-out activity into the 2nd quarter. First possession, he found Budinger in the corner for a three, and then swooped in for the offensive rebound and found Landry under the basket. At 9:00, he drove past Meeks for an and-1. At 8:28, after Thomas blocked Landry’s shot and recovered the ball, Lowry snuck in and stole the ball away, and later in the possession got a put-back score off a Brooks miss. At 7:55, Lowry again stole away a rebound from Thomas and Warrick. At 6:20, Kyle pushed the ball up the floor and found Andersen for a 3-point make in transition. At 4:30, Lowry hit a trailing Budinger for a 3-point make in transition. At 4:05, Lowry drew a foul in transition, getting to the line.

Landry’s first half was decidedly less effective, with the Bucks putting Kurt Thomas, a superior post defender, on him. Thomas, much like Hayes, takes full advantage of his strong base and quick hands when defending the post. At 10:30, Thomas used a fronting defense to deny the ball to Landry (the Rockets score anyway off a well-executed basket cut from Budinger and well-timed pass from Andersen). At 9:40, Landry turned to face Thomas in the post but had to force up a bad shot after Thomas got his hand on the ball on the way up. And at 8:28, Landry got his shot blocked against Thomas in the post. Defensively, he made a few mistakes as well. At 10:20, he failed to box out Kurt Thomas, who scored on a put back. At 9:05, he was late on a rotation as Jennings blew past lead-footed David Andersen for the layup. Landry contests shots and moves his feet well enough. He has little presence as a defensive rebounder, and is too often unalert in help defense. The encouraging thing is that he has fully acknowledged this and clearly wants to improve. Its a work in progress.

As was recently discussed, the Rockets have struggled at the start of second halves, particularly since the beginning of December. In this third quarter, Scola again scored consistently while guarded by Bogut or Mbah a Moute. Ariza again provided a mixture of good and frustrating plays. This time, I’ll start with the bad. At 9:50, he turned it over while running a pick and roll with Scola. At 4:03, he made a hustle player chasing down a loose ball at half-court, but instead of picking it up and finding his teammates in transition, he dived onto the ball and called a timeout. Unnecessary. At 2:15, Ariza fell asleep against Delfino, allowing him to score off a baseline cut while he watched the ball. At 1:45, he drove into the paint and air-balled a 6-foot shot taken over Bogut. At 1:10, Ariza missed a 3-pointer and failed to promptly get back in defense, allowing Delfino to take and make a wide open transition 3-pointer. The good: he drained a passively contested 3-pointer over Delfino at 9:01 from the corner. At 6:08, he triggered a steal on one end and found Scola in transition. At 3:55, he drove into the defense against their zone and found Scola in the lane. Good pass, but Scola wasn’t able to finish.

Lowry and Landry entered the game with three and a half minutes to go in the third quarter, but this time the Rockets finished the quarter getting outscored 2-5. On back to back possessions, Lowry’s pass to Scola was contested by Bogut, and Lowry’s shot off a drive was contested by Bogut. He had a big defensive impact on this game. The Rockets finally scored on their last possession of the quarter, as Lowry drove and found Scola for an 18-foot jump shot which he nailed.

At the start of the 4th quarter, the Rockets were up 6 points — 79-73. They started the quarter with a lineup of Lowry, Ariza, Budinger, Landry, and Scola, while the Bucks used Ridnour, Bell, Delfino, Mbah a Moute, and Bogut. The Bucks used a zone defense throughout this quarter, forcing the Rockets into stand-still offensive sets and numerous 3-point misses. Through the first 6 minutes, the Rockets were only 1 for 9 from the field. The Rockets have been most successful in the fourth quarter with Landry becoming the offensive focus. Landry was quiet through most of the 4th quarter, but at 5:00 he had one of the more spectacular finishes of the season, doing a 180 in the air, flipping the ball over his head, and scoring while drawing the foul. At 3:24, he got himself to the foul line off a quick spin move against Bogut. In the end, the Rockets were outscored in the quarter 19-12, and the game went to overtime.

In the overtime session, Landry went to work. At 4:30 in OT, he scored while facing up against Bogut on the right block. At 4:00, he scored and drew a foul while posting up on the left block (against Bogut). At 3:44, he blocked a shot by Jennings. These 3 plays in the first minute and a half of OT helped put the Rockets up 5 points. Scola scored the next two baskets for the Rockets — an 18-footer off a pass from Brooks, and a paint score from a deflected Ariza pass he picked up . The Rockets retained a comfortable lead until their closing offensive possession, before which a putback by Bogut closed the gap to two points. Brooks, in a familiar position for him, hit only 1 out of 2 free throws. Brooks is a good free throw shooter, but like many younger players does not yet seem comfortable in converting from the line in crunch time situations. On the Bucks final possession, down three, they executed their play perfectly. Jennings drove and found Ridnour in the opposite corner for the wide open 3-point look (his defender, Lowry, got screened out of the play). It was the best shot the Bucks could have hoped for in that situation, but they missed it. The Rockets won, barely avoiding a second overtime.

Is this a playoff team in the Western Conference? The Rockets GM tweeted after this game that the Rockets are not a playoff team without change. That a sense of urgency is needed to build up and sustain bigger leads. The Rockets have lost 6 of 7 of their previous road games (the lone win occuring against the hapless New Jersey Nets). It is true that they have won 9 over their previous 10 home games, with many more to come this month, but needing overtime to defeat the Minnesota T-Wolves and Milwaukee Bucks does not inspire confidence.

We are just not playing well.Not getting good shots¬ making shots we get.TOs.Start of 1st&3rd needs new approach.AB/Carl kept us in it.

– Daryl Morey (twitter)

Charting the Houston Rockets Starts Over the Course of the Season

How bad have the starters been performing of late? Below, I’ve charted their offensive and defensive performance in the 1st quarters for each game, looking only at possessions that came before the first Rockets substitution. The bars represent efficiency for each game, while the lines are a “last 4 games” average. Red color represents points scored, while blue color represents points allowed.  Note that there is a blank for the game played at Oklahoma City on 11/29, because that is the game in which starter Luis Scola was injured on the first possession.

There has indeed been a recent downturn in the 1st quarter starts, but of late the Rockets have only trailed the last two games  — a back-to-back on the road against the Lakers and Suns. It is troubling that the Rockets would have had such a difficult time scoring against the poor Phoenix Suns defense (they managed only 7 points in 12 possessions, while allowing 21 points). But should we be concerned with the larger trend? Before the Lakers game, the Rockets starters ran out to a favorable start in 4 of 5 games, and they played the Hornets evenly on the road. And for the first month of the season, the Rockets generally played well at the start of games.

But 1st quarters are only a part of the story. This is what the starters have done to begin 3rd quarters:

Now we see a much more troubling trend. In 7 of the last 9 games, Rockets starters have been outplayed to start the 3rd quarter (before their first second-half substitution). The offense has ranged from mediocre to anemic over that stretch, and the defense has struggled also on average (with particularly poor defensive efforts coming against Orlando, Cleveland, New Orleans, and Phoenix).

Below, I summarize the net results (that is, points scored minus points allowed) for the 1st and 3rd quarters. Here, the red color refers to the 1st quarter start, the blue color refers to the 3rd quarter start, and the green color refers to the combined net efficiency to start both halves. From this, the starters did their job pretty well through the first month of the season. At the beginning of December, the team had a string of poor first quarter starts, and over the last few weeks the team has struggled in their second half starts.

The following summary stats indicate how poorly the Rockets starters have sputtered since the first month:

Entire season ...
      start_1st      start_3rd  start_combined
OFF      107.0          98.1          102.7
DEF      105.1          99.3          102.4
NET      +2.1           -1.2          +0.3

Since beginning of December ...
      start_1st      start_3rd  start_combined
OFF      100.0          90.0           94.8
DEF      106.9         102.4          104.6
NET      -6.9          -12.4          -9.8

The defensive efficiency has been respectable, but we expect at least that with a unit consisting of Shane Battier, Chuck Hayes, and Trevor Ariza. It has not been close to good enough to overcome the awful offense.

Patterns in Starters’ Performance and Final Margin

Strangely, of late there has been a large divergence in the performance at the start of each half. An interesting pattern emerges when I plot net efficiency in the 1st quarter versus net efficiency in the 3rd quarter (below). The 14 times the Rockets starters have “lost” in the 1st quarter (the left quadrants), they’ve “lost” the 3rd quarter 9 times (the lower-left quadrant). But in the 20 times the Rockets have “won” the 1st quarter (the right quadrants), they’ve “won” the 3rd quarter only 9 times (the upper-right quadrant). This would seem to suggest that when the Rockets starters play poorly to start the game, its because they’re typically outmatched. But when the Rockets starters play well to start the game, it is much less an indicator of superiority — opposing starters will adjust to the Rockets more frequently than the Rockets can adjust to the opposition.

How are these slow starts impacting the Rockets ability to win games? Ultimately, this is the concern. Below, I have plotted the final margin against the combined performance to start the 1st and 3rd quarters (which will usually represent about 1/4th of the game).

Interestingly enough, these result show that the starters are not losing their portion of the game any more often than the team loses overall. The team’s overall win-loss record is 20-16, but if we only consider the start of the 1st and 2nd halves, they have a “win-loss” record of 21-15. If we only consider the start of the 1st half the record is 21-14 (very good), but considering only the start of the 2nd half the record is 14-20 (opponents are adjusting, as I said earlier).

Closing Remarks

Being a team of role players, there is much less of a gap (perhaps no gap?) between the players that are 1-5 in the rotation versus 6-10. Consequentially, it should come as no surprise that the starters would be outmatched on many nights by opposing, star-laden units. But because of the depth and talent on the bench, the Rockets have been well-equipped to overcome this. The numbers show that the starters have played pretty well when looking at the season as a whole, but they having been trending seriously southward since the beginning of December.

It should be fairly obvious that major improvement on the offensive end is necessary. The competition should ease up over the coming weeks. This could be an opportunity for the current starting group to regain some of the cohesiveness and chemistry they seemed to have in the month of November. I could also see a change in the starting lineup, with Carl Landry replacing Chuck Hayes seeming most likely.  I am hesitant to endorse that particular change, as the defensive performance with Scola and Landry on the floor has been consistently awful for years, but it is possible that Landry’s insertion would jump start the offense enough to offset this. Other suggestions I’ve read from other fans include replacing Chuck Hayes with David Andersen or replacing Trevor Ariza with Chase Budinger or Kyle Lowry. A combination I would like to see more of is Carl Landry and Chuck Hayes, but I understand the concern in using such a tandem. With Hayes having effectively no shooting range, the defense is free to load up on Carl Landry and (likely) reduce his effectiveness as an interior scorer.

• OFF: offensive efficiency or points scored per 100 possessions
• DEF: defensive efficiency or points allowed per 100 possessions
• PACE: possessions per 48 minutes
• EFF: point differential (per 100 possessions)
• OFF_std, DEF_std, PACE_std, EFF_std: these are standardized statistics, where a value of 0 would be league average, the more negative the worse the team is, and the more positive the better the team is.
• OFF_DEF_std: this indicates how balanced the team is, more positive means the team is offensively-inclined, more negative means the team is defensively-inclined
• And for each of the above, you can choose between 2 weeks (stats over prior 2 weeks) or season (stats from the beginning of the season)

The gadget is fairly customizable. By default, it shows offensive efficiency vs defensive efficiency, though using the standardized versions you might get a better gauge of how “above average” or “below average” a team is along a given metric. Another view I would suggest taking a look at is EFF_std vs OFF_DEF_std.  Along one dimension you’ll see how strong or weak the team is overall, and along the other dimension you’ll see their offensive/defensive orientation. You can also set the color and/or size of each point to depend on a specified metric. The Trails check box will control whether a path is drawn for the teams you’ve selected. This is useful if you want to get a stronger visual sense of how a team is trending over time. Here’s a  snapshot to illustrate what it can show you:

Tracking Rockets Offense/Defense over time

In the above example, the changing color from one point to the next reflects changes in pace. The size of the points change to reflect overall efficiency (a “-” in the middle indicates periods where the Houston Rockets were getting outscored). You can also zoom in by dragging a rectangle over an area of interest. Opacity of non-selected teams can be adjusted by clicking the wrench in the lower-right corner as well.

Try it out for yourself:

In addition to the motion scatter plot, you may also find the charts in the other two tabs interesting. The bar chart in the middle tab will show you how the ranking of selected teams along specified metrics change over time. The third chart summarizes the fluctuations of specified metrics over time with a line graph. In the near future, I will look to add “four factor” stats as well (i.e., field goal efficiency, rebounding, turnovers, free throws). If you have any suggestions on other numbers that might be interesting to track, please let me know in the comments section.

During the Lakers-Cavs Christmas Day game, Mike Breen asked his broadcasting partners to comment on how underrated Andersen Varaejo has been as a 6th-man of the year candidate. Jeff Van Gundy, as usual in the Houston Rockets’ corner, used the opportunity to speak up about the amazing year Carl Landry has been having off the Rockets bench. According to him, Landry has been the league’s best reserve so far, and it isn’t close.  Such a claim may have come as a surprise to a large portion of the audience who haven’t seen the Rockets play due to their absence on the national television stage. Rockets fans know Landry has been really good, but even they may not fully appreciate how effective he’s been.

Allow me to put it more concretely. So far this season (up through 12/24, that is), 136 players have played at least 10 minutes in at least 10 games coming off the bench.  Amongst this pool of players, Landry ranks 1st in points scored per minute and 1st in TS%  (points scored per shot attempt, including free throws). Moreover, he ranks 1st in PER (a per-minute, pace-adjusted player rating based on the boxscore) and 1st in Win Shares (total contribution towards a team’s wins, based on the boxscore).  The chart below illustrates his dominance amongst reserves by these metrics, and I’ve highlighted the other 6th-man of the year candidates mentioned in the most recent ESPN.com Awards Watch (and Kyle Lowry also, who’s worthy of being in the discussion).

Win Shares, PER Scatter Plot. Landry dominates both per-minute and in total contribution, by the boxscore.

That, I think, makes for a pretty good case in support of Jeff Van Gundy’s contention. But I would like to make still one more point. Would you believe that, for games through 12/24, Carl Landry leads the league in 4th quarter scoring? Here are the top 20 players so far this season in points scored per game in the 4th quarter:

rnk     Player              TS%        pts/4th
1       Carl Landry         69.7%       6.8
2       Carmelo Anthony     60.8%       6.6
3       J.R. Smith          58.7%       6.4
4       Dwyane Wade         56.9%       6.3
5       Brandon Roy         57.8%       6.1
6       Dirk Nowitzki       54.8%       6.0
7       LeBron James        55.2%       6.0
8       Jason Terry         61.5%       6.0
9       Danny Granger       55.1%       5.8
10      Kevin Durant        52.8%       5.8
11      Brandon Jennings    52.1%       5.7
12      Kobe Bryant         55.7%       5.5
13      Monta Ellis         58.1%       5.5
14      Will Bynum          53.9%       5.3
15      Earl Boykins        69.3%       5.3
16      Russell Westbrook   56.7%       5.0
17      Tyreke Evans        50.5%       4.9
18      Al Harrington       59.9%       4.8
19      Chris Paul          60.0%       4.8
20      Mo Williams         54.7%       4.8

Notice also that Landry not only tops this list in 4th quarter scoring, he’s also been the most efficient scorer amongst the top 20. And its not as if the Rockets have played a ton of blowouts and Landry has just been padding his stats in garbage time. In the second half, when the Rockets need buckets, Carl Landry has been their goto player.  Below, I show another scatter plot, indicating both points scored and efficiency, for the top 100 players in  4th quarter scoring. I’ve grayed out players who have started at least 40% of their games so as to highlight the reserves. I’ve also adjusted the label size based on total minutes played to focus on the “important” players for our consideration.

4th Quarter Scoring -- Landry stands out as a 4th quarter scorer, whether amongst reserves or the entire league.

What’s interesting is that not only does Landry’s fourth quarter proficiency set him apart from other players, he’s the only “big” who operates primarily in the post who is even close to the top as a top 4th quarter scorer. What makes him so good? Landry possesses elite quickness when operating in the post or facing up, and he makes up for his relatively short stature by having incredible strength and coordination in the air.  He has also expanded his post moves to include an accurate jump hook or fallaway in the lane. Perhaps most importantly, Landry has become a dead-eye free throw shooter. With Yao, the Rockets had a big man who they could depend on to shoot 85% from the line. Its quite a luxury to have an option in the post who’s both adept at drawing fouls and will convert at a high percentage, and Landry has continued the tradition in Yao’s absence.

Update

Upon reading the 4th quarter scoring stats above, readers may have been surprised to see players such as Kobe Bryant falling outside the top 10.  While I do believe, for this discussion, minutes played in the 4th quarter is relevant, a more accurate portrayal of the best 4th quarter scorers would be to look at it per-minute. I’ve updated the table above to show 4th quarter minutes played per game and also scoring per 36-minutes.

rnk     Player             4th_min    pts/36min    TS%      pts/4th
1       Carl Landry         9.9         24.7      69.7%      6.8
2       Carmelo Anthony     7.5         31.6      60.8%      6.6
3       J.R. Smith          9.5         24.0      58.7%      6.4
4       Dwyane Wade         7.7         29.6      56.9%      6.3
5       Brandon Roy         8.4         26.1      57.8%      6.1
6       Dirk Nowitzki       8.4         26.0      54.8%      6.0
7       LeBron James        7.3         29.7      55.2%      6.0
8       Jason Terry         10.4        20.8      61.5%      6.0
9       Danny Granger       8.8         23.7      55.1%      5.8
10      Kevin Durant        8.8         23.4      52.8%      5.8
11      Brandon Jennings    8.4         24.6      52.1%      5.7
12      Kobe Bryant         6.2         31.8      55.7%      5.5
13      Monta Ellis         8.1         24.2      58.1%      5.5
14      Will Bynum          9.5         20.0      53.9%      5.3
15      Earl Boykins        8.6         21.9      69.3%      5.3
16      Russell Westbrook   7.5         23.8      56.7%      5.0
17      Tyreke Evans        8.2         21.4      50.5%      4.9
18      Al Harrington       8.5         20.4      59.9%      4.8
19      Chris Paul          6.8         25.5      60.0%      4.8
20      Mo Williams         9.2         18.7      54.7%      4.8

And below I’ve included another scatter plot, this time showing top 100 players in 4th quarter scoring per 36 minutes.

Fourth Quarter Scoring Per 36 minute; Landry still excels.

Just to recap, in my first post I looked at individual box score stats for players within the first 7 seconds of possessions. This gave me an idea of which players are most involved offensively during fast breaks, though I didn’t distinguish plays which may have evolved early in a possession but were not technically fast breaks (e.g., an out-of-bounds play out of a timeout). In my second post, I tried to show how pace for the Houston Rockets fluctuates over the course of the game, and how this relates to various players being on the floor (in particular, the PGs).  I used two metrics to assess “pace’ — time per possession and time to first shot, drawn foul, or turnover per possession.

To get a better understanding of how the up-tempo attack is being used for the Rockets benefit, I wanted to take a closer look at how frequently the Rockets get the shorter possessions vs the longer possessions, and also how their efficiency relates to this. Again using the play-by-play data from Basketball Geek, I recorded both the duration of each possession and also possession triggers or events that started a possession. This allowed me, for instance, to distinguish possessions triggered by a steal, by a defensive rebound, or by a made field goal (amongst other possible events). I also recorded how many points were scored on each possession (ignoring technicals foul shots) to determine offensive efficiency.  As I did this for every possession played this season across all games, I can compare the way in which the Rockets use their possessions to other teams or the league average.

In the graph below, I’ve shown the percentage of possessions lasting less than t seconds given various possession triggers. I thought this graphical representation would be interesting because, before, I looked only at the average duration of possessions.  Here, we can visualize it more as a distribution. I’ve also shown the data for the league as a whole, for comparison. The solid lines are percentages for all offensive possessions, while the other lines represent possession after particular triggers. Note that overall, the Rockets are almost exactly at league average (the solid red and blue lines overlap), though the Rockets are one of the best offensive rebounding teams (which will have the effect of extending possessions). I haven’t shown it here, but my guess is that if I plot a similar graph for time to first shot, drawn foul, or turnover, it will show the Rockets to be getting into their offense faster than league average.

In the next graph, I show offensive efficiency for possessions lasting less than t seconds. Overall, the Rockets do appear to be close to league average, though they tend not to be as efficient in possessions lasting under 15 seconds. By looking at the lines for the particular possession triggers, I find that what’s contributing to this overall subpar efficiency on short possessions is the performance after defensive rebounds (the dashed lines). It is possible that the Rockets lack of a great defensive rebounding presence (with all due to respect to Luis Scola) limits their ability to get easy scores in transition off FG misses. It is evident here that the Rockets have done an excellent job in capitalizing off of their steals, with an offensive efficiency nearly 10 points higher than league average overall.  Trevor Ariza, for all his faults on the offensive side, has contributed significantly in this respect.

In the following graphs I carry out the same comparison, but with the Rockets from a season ago (actually, for 72 of the 82 games last year). We see that both overall, and for the various listed possession triggers, the Rockets are getting more short possessions than last year. Again, this graph is actually understating how much more quickly the Rockets are getting into their offense because it is based on possession time instead of time to first shot/foul/turnover.

For possessions after steals, we see a large improvement for the Rockets this year over last, a consequence of them pushing the tempo more effectively. Overall efficiency is about the same rating over all possessions, though the Rockets are more efficient in possessions under 10 seconds (meaning they are probably less efficient in possessions longer than 10 seconds). Overall efficiency on possessions after made field goals is about 5 points worse for the Rockets this year, unsurprising because its hard for them to run on those possessions.

There are a couple of directions we could go from here on this investigation of Rockets up-tempo play:

• A good next step would be to plot these graphs using time to first shot/foul/turnover rather than possession time. The Rockets are trying to run more than these graphs probably give them credit, but their offensive rebounding extends a lot of their possessions.
• How do different 5-man lineups impact the shape of these curves?
• It’s evident here that there is offensive value in certain triggers created by the defense. For corresponding reasons, there is also defensive value in running an effective offense. Carrying out a league-wide study of which teams get the most offensive value from their defense, and vice-versa, could be interesting.

Update

I took a look and it turns out that discarding the extra possession time due to offensive rebounds does not show the Rockets to be appreciably faster than the league as I initially had guessed. On average, a Rockets possession lasts 15.7 seconds compared to 15.6 seconds for the league (virtually the same).  Only considering the time to first shot/shooting foul/turnover each possession, for the Rockets and the league I found it to be 14.1 seconds.

Below, I’ve summarized both the average time to first shot/shooting foul/turnover (or poss_first_time) and offensive efficiency (or ORTG) for possessions following various triggers, both for the Rockets and for the league as a whole.

Some comments:

• The “other” trigger refers to possessions that did not start from a steal, defensive rebound, or opponent’s made field goal. Other than the occasional jump ball, these constitute possessions taken from out of bounds where the clock had stopped. The Rockets look to be slightly slower and with less efficiency compared to league average on these possessions.
• As noted above, the Rockets excel efficiency-wise when capitalizing off of steals.
• The Rockets play notably faster than league average after made field goals, indicating an effort to quickly strike back when scored on.
• Off of defensive rebounds, relative to the league they appear to be slow in the ensuing possession. This isn’t a surprise; since the Rockets are small there is more of an emphasis on “gang rebounding” on the defensive end, limiting opportunities for perimeter players to leak out in hopes of a miss and long outlet pass.

The first thing I want to look at is how the Rockets pace changes over the course of the game.  What do I mean by “pace”? A conventional definition is “possessions per 48 minutes”. This would work fine, but here I will use something slightly different — average time (in seconds) per offensive possession. For the Houston Rockets, an average possession lasts about 15 seconds. But because the Rockets are so aggressive at going after offensive rebounds, looking only at time per possession could give a skewed reading on the  extent to which the Rockets are looking to run.  So, I also consider another metric which is the average number of seconds per possession before the Rockets take their first shot, draw their first foul, or commit a turnover.

The graph below depicts how both of these metrics change over the course of the game. I’ve segmented each quarter into 3 minute chunks, though the last “chunk” actually represents the final 3 minutes of regulation and any overtime minutes (the Rockets have played only 5 minutes of overtime so far this season). The bars represent total number of Rockets possessions in each segment this season. The red line (denoted “poss_time”) is average time per possession, and the green line (denoted “poss_first”) is average time to first shot, drawn foul, or turnover per possession.

A couple things stood out to me when I saw this. There is a considerable gap between the red and green lines, which I imagine corresponds to the Rockets’ superior offensive rebounding. The Houston Rockets tend to slow down, overall, in the second halves, and particularly in the middle of the fourth quarters. And, finally, it was interesting to me how the number of possessions within each 3-minute chunk appears to go up over the course of every quarter. In particular, there is a prominent spike in number of possessions at the end of each quarter.  Curious, I decided to look closer at the time per possession for the Rockets versus their opponent.

From this, an interesting pattern emerges.  In the last 3 minutes of each quarter, Rockets opponents spend on average about 1 second less with the ball per possession. Is this a sign that the Rockets transition defense slips to close out quarters? Is it that the Rockets pick up their defensive rebounding? Are they better at forcing turnovers?  I’d like to look into this further in a future post. But for now, I will proceed by considering how the pace changes based on who is on the floor.  To illustrate this, I took a page from 82games.com and their “player pair” +/- chart.

The table below shows change in average time-per-possession with the player who’s name is on the very top row, versus without that player. Consider the cell for the ‘Brooks’ row, and the ‘Lowry’ column. This means that Rockets possessions are 1.6 seconds shorter when Brooks is paired with Lowry versus when he’s on the floor without Lowry. The cells are color coded so that the more red they are, the more the player at the top of the column is “picking up the pace”. Conversely, the more blue they are, the more the player at the top of the column is slowing the team down (pace-wise  — not necessarily saying that’s a bad thing).

Houston Rockets Player Pair Time-per-Possession Impact

I also created the same table for the poss_first metric described above. I expect the results to be similar, and they are:

Houston Rockets Player Pair Impact on time to first shot/drawn foul/turnover per possession

The bottom row serves as a summary of each players On/Off impact on the the pace.  As we’ve been hearing, Lowry does indeed appear to have a significant impact on increasing his team’s pace. Battier, on the other hand, is at the opposite end of this spectrum. Pretty much across the board, with whichever players he’s sharing the floor with, the pace will slow down with him. Earlier results indicated that Battier does not produce a lot, in terms of individual “box score” stats, in the first 7 seconds of possessions. Couple that with these findings and it should be fairly clear — Battier is not an ideal fit for a fast-paced offense.

Returning to Lowry, observe that in his row, Brooks has the biggest impact in picking up the pace when joining him on the floor, just as its true going the other way (Lowry picks up the pace more than any other player when joining Brooks on the floor). This, of course, makes a lot of sense. Both PGs are quick, and are potent in their own way in transition. Lowry is great at pushing the ball down the floor, drawing fouls, and finding players at the basket or on the 3-point line. Brooks is a potent shooter in transition and has the greater ability in penetrating defenses that are more or less set. Put them both on the floor, and the pace should go up dramatically. Consider the following graphs:

With Brooks and no Lowry, possessions take a little longer. Adding Lowry, especially in the second halves, the possessions seemingly shorten by 2-3 seconds on average — a dramatic pace increase. But its perhaps worthwhile to look still deeper,  considering also how many possessions each combination gets together over the course of the game:

The above line graphs show the pace fluctuations with Brooks, Lowry, and both together over the course of the game on average, where gaps in the lines correspond to data points with 0 possessions. The bars indicate the number of possessions within each time segment (relevant, because not all the data points for the lines are based on a significant number of possessions). From this, we see that Lowry gets the bulk of his minutes and possessions in the second and fourth quarters. In those time periods, generally speaking, it doesn’t appear to be so much Lowry by himself but rather the combination of Brooks and Lowry that contributes significantly to the pace increase.  So, if the Houston Rockets really want to increase their pace, I think I can safely suggest one of two things: (1) sub in a more transition-oriented player (Budinger!) for Battier, and (2) pair up their PGs.

One of the points of emphasis for the Houston Rockets this season has been to increase the pace and get more of their offense early. The reasons are straightforward.  Without the dominant low post option in Yao Ming that they’ve relied on for these many years, and with the availability of their premier playmaker, Tracy McGrady, in question, the Rockets could no longer depend on their half-court offense for 48 minutes.  Last season, the Houston Rockets ranked 19th in pace and 30th (that would be dead last) in fast break points. That needed to change, and it has. They are currently ranked 9th in pace and as recently as a week ago were 5th in fast break points.

But let’s dig a little deeper. We know that the gold standard for up-tempo play in the modern era was Mike D’Antoni’s “7 seconds or less” offense under Steve Nash. With that in mind, and using the fantastic play-by-play spreadsheets provided by Ryan Parker over at Basketball Geek, I’ve gone through all the Rockets possessions this season and recorded what they’ve done within the first 7 seconds. I didn’t filter out half-court possessions after a timeout, but for the most part this should give a good representation of what the Rockets have done in their “early offense”. Here are the results for the 9 primary rotation players:

                 MIN    PTS  FGM  FGA  3PM  3PA fouled FTM  FTA  AST  3PAST  TOV OFOUL
Aaron Brooks     696    73   21   42   8    17   17    23   27   22     1    16   3
Trevor Ariza     822    96   37   55   8    19   13    14   17   13     3    8    1
Shane Battier    681    10   2    6    1    4    7     5    6    6      3    3    0
Luis Scola       605    45   17   20   0    0    18    11   12   6      2    6    0
Chuck Hayes      504    3    1    2    0    0    3     1    2    8      1    1    1
Kyle Lowry       502    41   8    17   0    0    24    25   29   31     7    9    2
Carl Landry      547    41   16   19   0    0    12    9    10   0      0    6    3
Chase Budinger   375    42   19   36   3    9    5     1    6    4      0    2    0
David Andersen   287    7    2    3    0    0    9     3    4    1      0    1    1
total            1013   363  124  202  20   49   111   95   118  91     17   54   12

The MIN column is just the total number of minutes each of the players have played this season. All the other columns show what each player has done in the first 7 seconds of Rockets possessions.  The fouled column refers to fouls the player has drawn, 3PAST is number of assists for 3-pointers, and OFOUL is number of offensive fouls the player has committed. The final row shows totals for the entire team.

Next, let’s look at that same data per 48 minutes. For this table, I’ve added a TS% column to the end indicating each player’s scoring efficiency:

                 MIN    PTS  FGM  FGA  3PM  3PA fouled FTM  FTA  AST  3PAST  TOV OFOUL  TS%
Aaron Brooks     696    5.0  1.4  2.9  0.6  1.2  1.2   1.6  1.9  1.5   0.1   1.1  0.2  67.7%
Trevor Ariza     822    5.6  2.2  3.2  0.5  1.1  0.8   0.8  1.0  0.8   0.2   0.5  0.1  76.8%
Shane Battier    681    0.7  0.1  0.4  0.1  0.3  0.5   0.4  0.4  0.4   0.2   0.2  0.0  57.9%
Luis Scola       605    3.6  1.3  1.6  0.0  0.0  1.4   0.9  1.0  0.5   0.2   0.5  0.0  89.0%
Chuck Hayes      504    0.3  0.1  0.2  0.0  0.0  0.3   0.1  0.2  0.8   0.1   0.1  0.1  52.1%
Kyle Lowry       502    3.9  0.8  1.6  0.0  0.0  2.3   2.4  2.8  3.0   0.7   0.9  0.2  68.9%
Carl Landry      547    3.6  1.4  1.7  0.0  0.0  1.1   0.8  0.9  0.0   0.0   0.5  0.3  87.6%
Chase Budinger   375    5.4  2.4  4.6  0.4  1.2  0.6   0.1  0.8  0.5   0.0   0.3  0.0  54.3%
David Andersen   287    1.2  0.3  0.5  0.0  0.0  1.5   0.5  0.7  0.2   0.0   0.2  0.2  73.5%
total            1013   17.2 5.9  9.6  0.9  2.3  5.3   4.5  5.6  4.3   0.8   2.6  0.6  71.5%

A few comments:

• This confirms how good Lowry is on the break — 31 “early” assists, including 7 to 3-pointers, lead the team by far.  He also draws a ton of fouls.
• Brooks, in contrast, shows great promise as a scorer in transition, but his playmaking for others leaves a lot to be desired. He’s very turnover-prone and has a terrible assist to turnover ratio for a PG.
• Ariza’s got nearly twice as many field goals in early offense as anyone else. He leads the team in points scored per minute in the first 7 seconds.
• How about Chase Budinger? He’s definitely been active running out in transition. He’s struggled efficiency-wise pretty much across the board, but you have to like the energy he’s brought.
• Earlier in the year, there was an article in the Houston Chronicle on how Carl Landry needed to improve in running the floor in the transition. Looks like he’s got it. He and Luis have both been great on the break.
• Shane and Chuck … not a big part of the early offense attack.
• Only 20 for 60 on early offense 3-pointers. That’s 33%, lower than I would have expected. Keep in mind that the percentages shown here could be tainted to a degree by quarter ending possessions where a contested shot against a set defense is required. Update: After filtering out end of quarter possessions, the team is now 20-49 on 3s for a healthy 41%.

So that’s interesting, but how does this compare to what the Houston Rockets did a year ago? Basketball Geek has provided 72 of the 82 Rockets games for the 08/09 season in its data section. Following the same procedure as above, here are the stats for last year:

                 MIN    PTS  FGM  FGA  3PM  3PA fouled FTM  FTA  AST  3PAST  TOV OFOUL
Yao Ming         2217   49   11   16   0    0    31    27   37   0      0    16   3
Ron Artest       2090   99   33   59   5    18   29    28   38   22     2    14   1
Luis Scola       2182   115  40   59   0    0    43    35   51   5      1    8    3
Aaron Brooks     1696   132  39   71   7    13   41    47   51   26     4    18   4
Carl Landry      1328   46   15   19   0    0    23    16   20   0      0    13   4
Von Wafer        1154   123  45   79   8    17   27    25   32   11     0    8    3
Rafer Alston     1308   61   20   48   1    7    20    20   24   35     7    17   4
Tracy McGrady    958    39   13   26   5    14   12    8    9    5      0    4    0
Shane Battier    1816   30   6    10   3    6    17    15   15   2      0    8    2
Kyle Lowry       538    52   17   28   0    0    27    18   26   25     7    11   2
Brent Barry      706    27   9    22   2    10   8     7    7    6      1    4    0
Luther Head      243    9    3    6    0    1    2     3    3    2      0    3    0
Chuck Hayes      705    0    0    1    0    0    1     0    0    3      0    4    0
total            3456   787  253  448  31   86   281   250  315  142    22   129  26

And, again, here are the stats per 48 minutes:

                 MIN    PTS  FGM  FGA  3PM  3PA fouled FTM  FTA  AST  3PAST  TOV OFOUL  TS%
Yao Ming         2217   1.1  0.2  0.3  0.0  0.0  0.7   0.6  0.8  0.0   0.0   0.3  0.1  75.9%
Ron Artest       2090   2.3  0.8  1.4  0.1  0.4  0.7   0.6  0.9  0.5   0.0   0.3  0.0  65.4%
Luis Scola       2182   2.5  0.9  1.3  0.0  0.0  0.9   0.8  1.1  0.1   0.0   0.2  0.1  70.6%
Aaron Brooks     1696   3.7  1.1  2.0  0.2  0.4  1.2   1.3  1.4  0.7   0.1   0.5  0.1  70.6%
Carl Landry      1328   1.7  0.5  0.7  0.0  0.0  0.8   0.6  0.7  0.0   0.0   0.5  0.1  82.7%
Von Wafer        1154   5.1  1.9  3.3  0.3  0.7  1.1   1.0  1.3  0.5   0.0   0.3  0.1  66.1%
Rafer Alston     1308   2.2  0.7  1.8  0.0  0.3  0.7   0.7  0.9  1.3   0.3   0.6  0.1  52.1%
Tracy McGrady    958    2.0  0.7  1.3  0.3  0.7  0.6   0.4  0.5  0.3   0.0   0.2  0.0  65.1%
Shane Battier    1816   0.8  0.2  0.3  0.1  0.2  0.4   0.4  0.4  0.1   0.0   0.2  0.1  90.4%
Kyle Lowry       538    4.6  1.5  2.5  0.0  0.0  2.4   1.6  2.3  2.2   0.6   1.0  0.2  65.9%
Brent Barry      706    1.8  0.6  1.5  0.1  0.7  0.5   0.5  0.5  0.4   0.1   0.3  0.0  53.8%
Luther Head      243    1.8  0.6  1.2  0.0  0.2  0.4   0.6  0.6  0.4   0.0   0.6  0.0  61.5%
Chuck Hayes      705    0.0  0.0  0.1  0.0  0.0  0.1   0.0  0.0  0.2   0.0   0.3  0.0  0.0%
total            3456   10.9 3.5  6.2  0.4  1.2  3.9   3.5  4.4  2.0   0.3   1.8  0.4  67.1%

For last season, we see that Von Wafer and Kyle Lowry were by far the most active scoring in transition.  Tracy McGrady, as we know all too well, wasn’t particularly interested in taking part in an early offense attack and over half his shots in this situation were 3-pointers (you know, the ‘trot down the floor and shoot a 3 on the break’ variety). We also can observe the marked improvement from Carl Landry in running the floor hard in transition this year compared to last, scoring over twice as many points per minute. Its also interesting to compare Ron Artest last year to Trevor Ariza this year. Both like to gamble on defense and are active as a disruptive force on that end. But its evident here how much more effective Ariza has been at turning those forced turnovers into offense.

                 MIN    PTS  FGM  FGA  3PM  3PA fouled FTM  FTA  AST  3PAST  TOV OFOUL  TS%
Ron Artest       2090   2.3  0.8  1.4  0.1  0.4  0.7   0.6  0.9  0.5   0.0   0.3  0.0  65.4%
Trevor Ariza     822    5.6  2.2  3.2  0.5  1.1  0.8   0.8  1.0  0.8   0.2   0.5  0.1  76.8%

And finally, let’s take a closer look at the team totals for last year and this year:

                 MIN    PTS  FGM  FGA  3PM  3PA fouled FTM  FTA  AST  3PAST  TOV OFOUL  TS%
total (08-09)    3456   10.9 3.5  6.2  0.4  1.2  3.9   3.5  4.4  2.0   0.3   1.8  0.4  67.1%
total (09-10)    1013   17.2 5.9  9.6  0.9  2.3  5.3   4.5  5.6  4.3   0.8   2.6  0.6  71.5%

The results speak for themselves. Not only are the Houston Rockets getting significantly more in transition this year, the efficiency has also gone up. If people are wondering how the Rockets have managed to be in the top half in offensive efficiency facing the hardest schedule thus far, this should tell part of the story. The other part, offensive rebounding, is a topic for another post.

Trevor Ariza     777     2.3    3.8    0.5   1.3     0.8      0.9    1.1     0.8    0.2      0.6    0.1