|
|
| CARVIEW |
Kyrie Irving is going to win Rookie of the Year, and he would get my vote, even though as you’ll see it’s not quite that clear cut from an advanced stats perspective. Here, we’ll look at how this year’s freshman class performed in three of my homegrown statistical metrics: ezPM, A4PM, and PSAMS.
ezPM
(Note these data are also always available in the “ROY” page.)
From the table one can see that Kenneth Faried and Kawhi Leonard did most of their damage in the rebounding department (REB100). Rebounding will likely always make these two box score studs. Kyrie Irving and Isaiah Thomas, perhaps, predictably, have very unbalanced ratings between the offensive and defensive side of the ball, with both being offensive standouts and defensive liabilities. It doesn’t help playing on really bad teams, since the defensive rating is largely shared by teammates. Iman Shumpert, conversely, appears to be a future defensive star, assuming his recent ACL injury doesn’t significantly affect his physical gifts. Ricky Rubio, despite being mostly known for the offensive flare in his game, also had a very high defensive rating according to ezPM. Note that Klay Thompson, who by all accounts will be among the first or second All-Rookie teams, comes in at #16 and had a fairly negative ezPM rating. The box score doesn’t seem to like him at this early point in his career.
PSAMS
Go to Google Spreadsheet
PSAMS is all about scoring. Three players really shined here: Kyrie Irving, Klay Thompson, and Isaiah Thomas. Kyrie appears to be the rare player who will be able to do damage in every area of the court. He can get inside, hit the mid-range jumper, and shoot beyond the arc. He also has the ability to get to the line. In fact, as I wrote about recently, Kyrie arguably had one of the greatest rookie scoring campaigns in the 3-pt era. Klay Thompson, while not (yet, anyway) having the inside scoring ability that Irving does, is undoubtedly going to be a 3-pt assassin and mid-range threat in this league. I think to say teams will have a tough assignment defending Klay and Stephen Curry is putting it mildly. Can you think of a backcourt combo that had this much pure shooting potential? I’ve thought about it, and honestly, have not been able to come up with a good comp yet. After Klay, we have Jimmer Fredette, oh, wait, Isaiah Thomas, who turned out to be a steal as the last player drafted. He needs to work on his mid-range game, but other than that, Thomas looks like he will be a very productive scorer. Other observations…Alec Burks really needs to improve his jump shot or teams will inevitably play well off him to force him to shoot it, as opposed to driving to the basket. Same with Kenneth Faried. He can score inside, no doubt. But if he can develop some kind of mid-range game, he can go to the next level offensively. Oh. Why doesn’t Jon Leuer play more on a team that needs better shooters? That’s a bit of a head scratcher.
A4PM
Go to Google Spreadsheet
In terms of the adjusted four factors, Irving is not among the top rookies, due to his poor defensive rating (the worst of the 20 rookies with >1500 possessions played this season). Hopefully, that is something he can improve, because he is such a special offensive player. His offensive rating is the highest on the list. I don’t think it’s a completely untold story that Irving’s defense is a liability, but it’s pretty clear that ROY voters tend to overlook such things. “All rookies suck at defense.” Not quite true according to these data. Chris Singleton, as advertised, put up a good defensive rating on a very bad team. That’s impressive. It will be interesting to see what happens with him if the Wizards get the #2 pick and do the sensible thing and draft Kidd-Gilchrist. I know at least one team out West (located currently in Oakland) that could use a guy like Singleton. The other defensive standouts were mentioned earlier, Shumpert and Rubio (thus giving some more confirmation). Unlike the box score stats, A4PM does like Klay Thompson. Phew.
In terms of individual factors, I’ll point out some notables. Faried and Burks are monsters on the offensive glass. Klay Thompson, touted for his size compared to Monta Ellis, appeared to somehow be an even worse rebounder his rookie season. I can’t imagine that not improving, but somebody on the Warriors staff needs to give him that message strongly. Going back again to defense, you can see the specific negative effect of both Irving and Thomas on defensive eFG% (remember a positive rating here means the opponent shot better!).
Now, for completeness, here are the ratings for players with less than 1500 possessions. You should probably take these with a grain of salt given the small sample size.
Go to Google Spreadsheet
The real DPOY?
Tyson Chandler was awarded the 2012 DPOY yesterday. Nobody was surprised by this, including myself. People did seem to be quite shocked and dismayed that Serge Ibaka got second place. If DPOY is stat-based, it’s likely only to the extent that players get above a certain threshold of blocks or steals. Of course, around these parts, we like to dig deeper and try to measure the true impact of a player on all parts of the game — those both seen and unseen. With that said, let’s see what the defensive half of A4PM (adjusted four factor +/-) has to say about DPOY. I’ve split the data into two sets, one for players who had >3000 possessions, and the other for players between 1500 and 3000 possessions. There’s not really much to say, except Andre Iguodala and Luol Deng probably should have got more votes. And, oh, Tyson Chandlerdoesn’t come anywhere near the top 5. Maybe those Ibaka nay-sayers are getting it wrong?
>3000 Possessions Played
1500 – 3000 Possessions Played
]]>So, once I get an interesting new idea in my head, I tend to obsess about it (perhaps, too much). Yesterday, I wrote about a way to compare players to each other using a “distance” measure of statistical similarity. Some time after I wrote that, I had a Eureka! moment and thought, hey, I should just put current NBA players into the model, and see who the current draft compares to. This is my first stab at it, using college stats from the last six draft classes (going back to 2006). I only used the basic pace-adjusted stats this time around, so I think there’s a lot of room for improvement. But I wanted to put something up, because I think the results are neat. There are definitely some head scratchers (Anthony Davis compared to Demar DeRozan?!). Oh, and in case you’re wondering, Jae Crowder is the next Jeremy Lin.
Click to enlarge.
Update
A couple of things. 1) It’s better to normalize each stat, so that scales are similar. 2) A polar dendrogram save a lot of space.
Click to enlarge.
]]>
The first thing I want to do − before assigning any value to players — is get some rough idea of statistical similarities. This involves performing some type of cluster analysis. As a first step, I took a bunch of (pace-adjusted) stats from DraftExpress and calculated a correlation matrix, which is show here graphically:
Click to enlarge.
To read off the correlations, simply start with one of the categories on the left side and move across. Bluer shades are more correlated, and redder shades are less correlated. For the ellipses, in addition to color, you can also look at how skinny they are. If the ellipse is very skinny, then the absolute value of correlation is closer to one, and if it is more circular, then the value is closer to zero.
Just to go through one example, let’s look at MIN (minutes), which is easy to dissect. You can see going across to the right, MIN is highly negatively (inversely) correlated with PF (personal fouls) and highly positively correlated with PCTTMPOSS (% of Team Possessions). It is not extremely well correlated with any other stat (except maybe blocked shots, which has an inverse correlation). One potential use of the correlation matrix is to pick out stats which you want to put in a model. If you don’t want to have a million variables in your model, for example, you could choose one or two from groups that appear to be highly correlated. This will help create a more parsimonious model, as they say.
Ok, now the really neat part. I took the 52 players that are in DX’s “Qualified Players” database (I think these are the players eligible and likely to be drafted) and calculated a distance between them using a function in R called — wait for it — dist():
This function computes and returns the distance matrix computed by using the specified distance measure to compute the distances between the rows of a data matrix.
To translate that a bit, basically we take all the stats shown in that correlation matrix and figure out how far apart players are from each other based on those stats. Players that have very similar stats will have a distance close to zero, while players with very different stats will be much farther away from each other.
Click here to go to Google Spreadsheet
One nice way of visualizing the distance (“player similarity”) is by creating a dendrogram. (You’ve probably seen a dendrogram before in high school biology.) Think of it as a tree with branches. The closer the branches are to each other, the more similar the players.
Player similarity dendrogram.
This is a pretty cool result for having put so little thought into it. You can see, for example, that a bunch of the big centers all cluster together near the bottom. I should point out that actual positions were not part of the dataset, so what you’re seeing is simply due to the statistical values in that spreadsheet. Going forward, I’d like to fiddle with the dataset and see which stats have more or less impact. Once a smallish set of factors is decided upon, I can then move on to the more critical task of player valuation.
]]>To calculate R^2 values (a measure of correlation), I used a lower threshold of 1000 possessions played in 2011 and 2012, which turns out to be 246 players. I ran a simple linear regression weighted by the total possessions played in the two seasons. The function call in R looks like:
Call: lm(formula = A4PM12 ~ A4PM11 + 0, weights = POSS11 + POSS12)
For A4PM, I found that 
With that said, here are some plots, so you can see for yourself that A4PM appears to give a tighter fit:
Here are the same plots with players, instead of dots:
Click to enlarge.
Click to enlarge.
In these plots, being closer to the upper-right corner means a player has been consistently better in both seasons, whereas being closer to the opposite (lower-left) corner means the player has been consistently worse. To find players who “improved” or “became increasingly positive”, look for players closer to the upper-left who had low ratings last season, but more positive ratings this season. A good example here is Larry Sanders or Beno Udrih (according to A4PM).
While I’m on the topic, I might as well hand out the award for Most Improved Player — whoops, I mean the player most increasingly positive according to A4PM in a way that is factually accurate and not at all intended to be inflammatory. Here is a table of Δ-values (A4PM12-A4PM11) for all players who played at least 1000 possessions in 2011 and at least 2500 possessions in 2012. (I’ve given the possession data in the last two columns, so if you want to use your own cutoffs, have at it).
You should keep in mind three things when looking at that table: 1) Regression to the mean; 2) Aging curves; and 3) Noise. Don’t be surprised to see young players at the top and older players at the bottom for both reasons. There are some really interesting names on the list at both ends (Tyson Chandler?!). Don’t be surprised to see outliers (they’re everywhere!). Anyway, it’s still an interesting list, and somewhat correlates with the impression of the MIP discussion in the media at-large (e.g. Dragic, Hayward, Rush, Gallo). Looks like Blake Griffin and Russell Westbrook also took significant steps forward this season. Just throwing this out there, but my bet is that next season, one of those two or maybe both will be in serious discussion for league MVP.
Ok, now some more fun plots to wrap up the post. Here are year-to-year plots (with player names) for each of the 8 components of A4PM. But first, a table of 
| RANK | FACTOR | R^2 |
| 1 | OEFG | 0.20 |
| 2 | OREB | 0.16 |
| 3 | OFTR | 0.16 |
| 4 | DTOR | 0.14 |
| 5 | OTOR | 0.13 |
| 6 | DFTR | 0.12 |
| 7 | DEFG | 0.08 |
| 8 | DREB | 0.07 |
Going forward, it will be important to keep in mind these values when looking at which players had surprisingly good or bad seasons according to +/-. For example, if a player has a very high DREB rating, it may be more likely due to noise. Conversely, if a player has a high OEFG rating, that is more likely to be “real”. Of course, as I said earlier, some form of multi-season averaging is clearly needed to improve predictive capability. Ok, now here are the plots like I promised.
Keep in mind that negative ratings are better for this component (i.e. own team has fewer turnovers when player is on).
Negative rating here means opponent went to the line less often.
Negative rating means opponent had lower eFG%.
DREB is somewhat of a misnomer. Really, this is the change in the opponent’s OREB% when the player is on. Therefore, negative ratings are better here.
Let me give an example to illustrate. The first series below is CHI vs. PHI. Let’s look at isolation plays (ISO) for each team. On offense, PHI ran ISO 12.5% of the time this season with a 0.84 PPP. Teams ran ISO against CHI 11.2% of the time with an efficiency of 0.74 PPP. Averaging those together, we can expect PHI to run ISO 11.9% of the time against CHI with an efficiency of 0.79. We can do the same calculation the other way. CHI ran ISO (only) 7.2% of its plays with an efficiency of 0.74 PPP. Teams ran ISO against PHI 10.2% of the time with an efficiency of 0.82 PPP. Therefore, we can expect CHI to run ISO 8.7% of its plays with an efficiency of 0.78 PPP. The rest of the calculations for each play type are similar.
The treemaps show the expected scoring in a visual and interactive format. The size of each tile is proportional to the % of total plays (i.e. each side sums to 100%) and are sorted from the upper left to lower right in descending order. The color of each tile represents the efficiency (PPP). If you click on a tile, it will pop up a tooltip with the data from the table.
In addition to Synergy, for each series I give expected TOV% and ORB%. Recall that the four factors are eFG%, TOV%, ORB%, and FTA/FGA. Because Synergy wraps together eFG% and FTA/FGA (since PPP includes free throws), we don’t need to consider these two separately.
Synergy play categories:
- ISO (isolation)
- BALL (pick and roll, ball handler shooting)
- POST (post-up)
- MAN (pick and roll, roller shoots)
- SPOT (spot-up shot, normally assisted)
- SCREEN (shooter coming off screen)
- HAND (hand-off to shooter)
- CUT (shot coming off a cut to the basket)
- ORB (shot comes immediately after rebound)
- TRANS (transition scoring)
- OTHER (i.e. desperation plays, etc.)
Eastern Conference
Chicago Bulls (1) vs. Philadelphia 76ers (8)
Expected Scoring
| PHI | CHI | |||
| PLAY | % | PPP | % | PPP |
| ALL | 100 | 0.88 | 100 | 0.89 |
| ISO | 11.9% | 0.79 | 8.7% | 0.78 |
| BALL | 13.4% | 0.78 | 13.2% | 0.76 |
| POST | 9.0% | 0.85 | 9.2% | 0.85 |
| MAN | 5.9% | 0.92 | 5.7% | 0.87 |
| SPOT | 16.6% | 0.93 | 18.6% | 0.95 |
| SCREEN | 4.3% | 0.88 | 6.3% | 0.80 |
| HAND | 3.7% | 0.81 | 2.5% | 0.86 |
| CUT | 8.4% | 1.12 | 9.3% | 1.15 |
| ORB | 5.6% | 0.97 | 6.9% | 1.02 |
| TRANS | 13.4% | 1.11 | 11.2% | 1.11 |
| OTHER | 5.8% | 0.38 | 6.5% | 0.41 |
Expected TOV%/ORB%
| TEAM | TOV% | ORB% |
| CHI | 13.35% | 28.70% |
| PHI | 11.85% | 25.05% |
Summary
Chicago narrowly tops Philadelphia in expected scoring, although they have been without Derrick Rose for a large portion of the season. Chicago also has a decided advantage in rebounding. Philadelphia’s main advantage, and it’s an important one, is turnover margin. Prediction: Chicago in 6.
(2) Miami Heat vs. (7) New York Knicks
Expected Scoring
| NYK | MIA | |||
| PLAY | % | PPP | % | PPP |
| ALL | 100 | 0.89 | 100 | 0.92 |
| ISO | 14.0% | 0.75 | 11.9% | 0.81 |
| BALL | 10.9% | 0.71 | 11.7% | 0.83 |
| POST | 6.1% | 0.85 | 10.2% | 0.81 |
| MAN | 5.6% | 1.03 | 4.7% | 0.95 |
| SPOT | 22.6% | 0.96 | 19.0% | 0.99 |
| SCREEN | 4.1% | 0.84 | 5.0% | 0.84 |
| HAND | 2.2% | 0.72 | 1.6% | 0.78 |
| CUT | 7.5% | 1.19 | 8.5% | 1.19 |
| ORB | 5.6% | 1.07 | 5.0% | 1.08 |
| TRANS | 12.8% | 1.09 | 13.5% | 1.14 |
| OTHER | 6.5% | 0.41 | 6.9% | 0.39 |
Expected TOV%/ORB%
| TEAM | TOV% | ORB% |
| MIA | 15.20% | 26.45% |
| NYK | 15.35% | 26.35% |
Summary
Miami has a large advantage in expected scoring. The turnover and rebounding factors are even. Given the ups and downs of each team during the season, and the number of stars who could get hot, this one could surprise people. Prediction: Miami in 6.
(3) Indiana Pacers vs. (6) Orlando Magic
Expected Scoring
| ORL | IND | |||
| PLAY | % | PPP | % | PPP |
| ALL | 100 | 0.89 | 100 | 0.92 |
| ISO | 7.6% | 0.77 | 9.9% | 0.76 |
| BALL | 13.5% | 0.76 | 10.4% | 0.76 |
| POST | 11.6% | 0.88 | 12.9% | 0.89 |
| MAN | 6.5% | 1.01 | 6.5% | 1.00 |
| SPOT | 20.4% | 0.96 | 19.4% | 0.96 |
| SCREEN | 5.0% | 0.92 | 4.7% | 0.90 |
| HAND | 2.1% | 0.88 | 1.9% | 0.86 |
| CUT | 6.7% | 1.22 | 7.5% | 1.20 |
| ORB | 6.0% | 1.04 | 6.0% | 1.05 |
| TRANS | 5.4% | 1.18 | 12.1% | 1.12 |
| OTHER | 7.5% | 0.34 | 6.8% | 0.49 |
Expected tov%/orb%
| TEAM | TOV% | ORB% |
| IND | 13.00% | 26.85% |
| ORL | 14.30% | 27.10% |
Summary
Indiana has the advantage in scoring and turnovers, with neither team having a clear advantage in rebounding. And obviously, that’s not taking into account that Dwight Howard is out. In theory, this seems like the most likely series to be a straight sweep, but the Magic have 3-pt shooters that can get hot in a game. Prediction: Indiana in 5.
(4) Boston Celtics vs. (5) Atlanta Hawks
Expected Scoring
| BOS | ATL | |||
| PLAY | % | PPP | % | PPP |
| ALL | 100 | 0.90 | 100 | 0.89 |
| ISO | 9.5% | 0.79 | 10.8% | 0.77 |
| BALL | 11.9% | 0.73 | 10.2% | 0.76 |
| POST | 10.3% | 0.83 | 9.5% | 0.80 |
| MAN | 6.4% | 0.96 | 4.6% | 0.96 |
| SPOT | 19.3% | 0.95 | 22.3% | 0.93 |
| SCREEN | 6.4% | 0.96 | 4.8% | 0.84 |
| HAND | 2.2% | 0.82 | 1.9% | 0.65 |
| CUT | 8.2% | 1.22 | 8.1% | 1.20 |
| ORB | 4.4% | 1.12 | 5.4% | 1.03 |
| TRANS | 12.5% | 1.13 | 13.2% | 1.15 |
| OTHER | 6.8% | 0.37 | 6.9% | 0.40 |
Expected tov%/orb%
| TEAM | TOV% | ORB% |
| BOS | 14.80% | 23.65% |
| ATL | 14.15% | 25.75% |
Summary
Boston has a slight edge in scoring, while Atlanta has the advantage in rebounding and turnovers. Given that scoring is more important than the other two combined, I’ll go with Boston in a close series. Prediction: Boston in 7.
Western Conference
(1) San Antonio Spurs vs. (8) Utah Jazz
Expected Scoring
| SAS | UTA | |||
| PLAY | % | PPP | % | PPP |
| ALL | 100 | 0.96 | 100 | 0.91 |
| ISO | 8.2% | 0.85 | 8.8% | 0.78 |
| BALL | 14.2% | 0.86 | 11.4% | 0.78 |
| POST | 9.7% | 0.79 | 13.6% | 0.84 |
| MAN | 7.0% | 1.08 | 5.2% | 0.95 |
| SPOT | 20.3% | 0.99 | 16.9% | 0.91 |
| SCREEN | 4.7% | 0.95 | 5.8% | 0.91 |
| HAND | 2.4% | 0.87 | 2.1% | 0.87 |
| CUT | 7.7% | 1.25 | 8.2% | 1.16 |
| ORB | 5.6% | 1.06 | 6.4% | 1.07 |
| TRANS | 12.6% | 1.19 | 13.5% | 1.15 |
| OTHER | 5.9% | 0.45 | 6.3% | 0.43 |
Expected tov%/ORb%
| TEAM | TOV% | ORB% |
| SAS | 13.30% | 25.65% |
| UTA | 12.95% | 27.10% |
San Antonio has a huge scoring margin, and Utah’s advantages in the other two factors are relatively modest. Prediction: San Antonio in 5.
(2) Oklahoma City Thunder vs. (7) Dallas Mavericks
Expected Scoring
| DAL | OKC | |||
| PLAY | % | PPP | % | PPP |
| ALL | 100 | 0.90 | 100 | 0.92 |
| ISO | 9.7% | 0.78 | 12.0% | 0.86 |
| BALL | 12.4% | 0.81 | 14.1% | 0.81 |
| POST | 10.9% | 0.81 | 8.5% | 0.81 |
| MAN | 6.9% | 0.96 | 4.9% | 1.01 |
| SPOT | 19.4% | 0.95 | 18.7% | 0.94 |
| SCREEN | 4.5% | 0.88 | 5.2% | 0.92 |
| HAND | 2.0% | 0.93 | 1.9% | 0.87 |
| CUT | 8.1% | 1.11 | 8.1% | 1.18 |
| ORB | 5.8% | 1.01 | 5.6% | 1.05 |
| TRANS | 12.0% | 1.10 | 13.3% | 1.16 |
| OTHER | 6.4% | 0.46 | 5.8% | 0.43 |
Expected tov%/orb%
| TEAM | TOV% | ORB% |
| OKC | 14.60% | 26.50% |
| DAL | 13.20% | 25.65% |
Summary
OKC has the edge in scoring and rebounding, although not a huge one. Dallas has a considerable advantage in the turnover margin. This should be a barn burner. Prediction: Oklahoma City in 7.
(3) Los Angeles Lakers vs. (6) Denver Nuggets
Expected Scoring
| DEN | LAL | |||
| PLAY | % | PPP | % | PPP |
| ALL | 100 | 0.93 | 100 | 0.92 |
| ISO | 11.3% | 0.79 | 11.5% | 0.80 |
| BALL | 11.6% | 0.80 | 9.4% | 0.83 |
| POST | 9.1% | 0.80 | 15.6% | 0.84 |
| MAN | 5.4% | 1.02 | 4.4% | 0.93 |
| SPOT | 19.7% | 0.94 | 18.6% | 0.97 |
| SCREEN | 4.7% | 0.85 | 4.8% | 0.96 |
| HAND | 2.0% | 0.86 | 1.8% | 0.84 |
| CUT | 7.6% | 1.19 | 8.3% | 1.24 |
| ORB | 6.0% | 1.02 | 6.2% | 1.06 |
| TRANS | 15.5% | 1.19 | 10.8% | 1.17 |
| OTHER | 5.2% | 0.44 | 6.7% | 0.41 |
expected tov%/orb%
| TEAM | TOV% | ORB% |
| LAL | 14.25% | 27.40% |
| DEN | 12.40% | 26.45% |
Denver has the edge in scoring and turnover margin, which are more important than rebounding. But I didn’t really like the trade of Nene, and I think these stats should be discounted somewhat. Prediction: Los Angeles in 7.
(4) Memphis Grizzlies vs. (5) Los Angeles Clippers
Expected Scoring
| LAC | MEM | |||
| PLAY | % | PPP | % | PPP |
| ALL | 100 | 0.91 | 100 | 0.91 |
| ISO | 11.8% | 0.82 | 11.8% | 0.77 |
| BALL | 13.9% | 0.83 | 11.2% | 0.75 |
| POST | 8.6% | 0.81 | 11.7% | 0.84 |
| MAN | 5.3% | 0.97 | 5.1% | 0.97 |
| SPOT | 20.9% | 0.97 | 17.3% | 0.95 |
| SCREEN | 3.9% | 0.98 | 4.1% | 0.81 |
| HAND | 1.9% | 0.75 | 1.9% | 0.83 |
| CUT | 7.1% | 1.24 | 9.1% | 1.19 |
| ORB | 6.0% | 1.03 | 6.5% | 1.08 |
| TRANS | 12.0% | 1.09 | 12.7% | 1.15 |
| OTHER | 6.8% | 0.41 | 6.6% | 0.47 |
Expected TOV%/ORB%
| TEAM | TOV% | ORB% |
| MEM | 13.90% | 28.30% |
| LAC | 14.50% | 28.40% |
Summary
This one is really even in pretty much every way. Memphis has an edge in turnovers and homecourt. Should be fun. Prediction: Memphis in 7.
]]>Jayson Williams, freshly released from prison, was quite a prolific offensive rebounder. If you don’t believe me, look at this table of the top offensive rebounding seasons (by ORB%) in the 3-pt era:
You don’t see Kevin Love on there, but you do see Jayson Williams occupying 3 of the top 5 slots, with Rodman nabbing the other two. At this point, you might be thinking, well, maybe Williams put up those numbers but they were “empty” and that New Jersey wasn’t all that great a rebounding team. And that he was “stealing” those rebounds from other players. You’d be wrong. New Jersey was right up there at the top of the league in ORB% those seasons.
This post is nominally about Jayson Williams, but it was inspired by a thread over at RealGM that was debating Dennis Rodman’s HOF credentials. When I compiled this list of great offensive rebounding seasons, what I realized is that: 1) Jayson Williams was also a really good offensive rebounder during the same period as Rodman; and 2) It really puts into perspective the change in philosophy on offensive rebounding that seems to have taken place over the last decade or so.
Whereas teams like New Jersey and those Bulls teams were rebounding around 35% or higher on offense, now the top team (Bulls again) are right around 33% and the next highest team (Utah) is at 30%. And the league average has dropped from 31.4% in 1998 down to 27% today.
This all raises a question of how do we value the historical significance of offensive rebounding? Was it more important back then than it is now (perhaps, due to rule changes)? Or have teams decided that it was not as important as they thought it was back then (perhaps, due to changes in analytics, etc.)? Or was it equally important then and now and we should just take the averages for what they are? And depending on your answers, how does that affect how we view the specific accomplishment or records of players like Dennis Rodman (and Jayson Williams) during that time? When we see a “great” offensive rebounder today like Kevin Love — who, btw, is at 11.6% (his career low) — how do we value that number? Is it even good anymore to be a good offensive rebounder? I say that mostly in jest, of course. But I think it points out more than ever, that we need to think about rebounding (and any stat really) in the context of the entire game.
We shouldn’t just throw out numbers and make lists and not think about the implications. Today more than ever, we have the tools to dig deeper than that. And we should whenever possible.
So no, I don’t think Jayson Williams belongs in the HOF, but I do think it’s an interesting question to consider for all the myriad reasons mentioned here and others I surely left out.
]]>Here are the rookies in 2012 with greater than 300 FGA attempted. Recall that 1.0 is the greatest DIST a player can have, 0 is what an average player would have, and -1 would be very bad. I’ve also standardized the rating according to how rookies peform. That’s given in the STD column. You can see that Kyrie Irving has been a very, very good scorer. He is 2.5 standard deviations above an average rookie. Klay Thompson (yes!) and Isaiah Thomas have also been quite good.
2012 Rookies
Here is the list of rookies since 1980.
Complete Rookie List since 1980
Go to Google Spreadsheet
Michael Jordan was an incredible scorer even as a rookie. And Kyrie Irving is among the top 10. One thing I tweeted about yesterday is that it’s fun to see names of players that I know nothing about pop on lists like this. Jeff Ruland (drafted by the Warriors, but never played for us) was a big (6’11” 280 pound) center who put together a few great seasons, but his career ended early due to injury. And Steve Johnson was a PF that played for us in 1991 before retiring. He was actually traded to Portland for Mychal Thompson (father of current Warrior Klay Thompson) in 1986. From what I can tell, Steve Johnson was basically David Lee before there was David Lee, because his MO was all about scoring and no defense. (Sorry, DLee, but it’s true.)
Actually, that last part is surprisingly not completely accurate. There actually was another David Lee (David G. Lee) who was drafted by the San Francisco Warriors back in 1968.
]]>