Fair warning: this post contains lengthy discussion of advanced statistical analysis and I know that might not "seem like fun" to certain people. Stick with it, there's a lot to learn and discuss.
Some of you might remember back to this summer when an invisible ninja passed some Synergy Sports reports about the Blazers my way. If you're not familiar with Synergy, it's a subscription service used by NBA teams that provides detailed statistical analysis and comparative ratings of every player and team in the NBA on both offense and defense.
How does it work? Briefly, Synergy charts every play of every NBA game and uses that raw data to calculate advanced metrics and gauge efficiency on a points per possession basis in various situations. As an example, through 20 games this season, Synergy reports that 21% of Portland's offensive possessions end with spot-up jump shots. Portland shoots 40.2% on these spot up jumpers and scores 1.03 points per possession on these shots. As you might guess, Synergy says Steve Blake takes the most spot up jumpers of any Blazer, followed by Martell Webster, Brandon Roy and Rudy Fernandez.
Synergy then takes these raw numbers and efficiency numbers and compares them to the rest of the league. Overall, the Blazers rank in the 72nd percentile of NBA teams in terms of efficiency from spot up jumpers. This earns the Blazers a rank of "Very Good" (the scale runs from "Poor" to "Excellent") compared to the rest of the NBA. Similarly, Synergy tells us that Blake ranks in the 67th percentile of NBA players for spot-up shooting, also earning a "Very Good" rating. It probably doesn't surprise you to learn that Andre Miller ranks in the 42nd percentile for a rating of "Average."
It is well-documented that the Blazers management staff is interested in statistical analysis of players. But what about the coaching staff?
I took up that question with Nate McMillan after Wednesday's practice. McMillan, a noted videotape breakdown fiend, surprised me with his honesty about how his coaching staff uses player breakdowns. "Oh, we show our players their numbers," he said, explaining that his primary focus in using advanced statistics is as a benchmarking tool for charting player development and consistency. "Whether we compare them after 20 games last year where you were, and 20 games where you are now. We also rank our team overall [compared to other teams] and share that information."
McMillan then pointed to a large whiteboard, probably four feet tall by six feet wide, that you've likely seen in the background of pictures taken at the Practice Facility. "That's what that board is over there. It's numbers for the players to see where they are compared to the league standings." The board, filled with writing in various colors, recalls an elementary school teacher's "gold star" positive reinforcement teaching tool. Each player's name is listed and, after each name, a list of numbers follows, charting their progress or lack thereof.
While the board is meant as a general overview and motivational tool, Synergy Sports' profiles (and other proprietary play-charting methods and software) drill down to the most specific of specifics, answering questions like: Does Brandon Roy do better on pick and rolls from the left side, right side or the top? How does he fare from each spot when he uses the pick, avoids the pick, gets doubled, or deals with a switching defense? If he uses the pick, does he do better taking a dribble jumper, a runner or going to the hoop?
The depth is absurd, jaw-dropping the first time you see it. To McMillan, though, these statistics and breakdowns have been a part of his coaching arsenal long enough that they barely warrant mention. "We've got all of that," he says without any reaction as I list off the pick and roll breakdowns to him that I just checked off above. "We have all of that information." Even the runners compared to the dribble jumpers off the left-side single-covered pick and rolls? "All of that is available," he says, almost bored. "All of that is available."
Which brought to mind the old adage: data is only meaningful when acted upon. How exactly does the Blazers coaching staff give meaning to all of these numbers on a day-to-day basis?
Before games, players and coaches review any pertinent personal or team adjustments. Playing against a running team? Adjustments might include: ease off attacking the offensive glass or pick up the primary ball-handler earlier in the shot clock. Players also receive a condensed summary of their opponent's strengths, weaknesses and tendencies prepared by scouting personnel. These reports boil down the advanced numbers to easily-digestable tidbits but also include more subjective analysis drawn from video. Last year's tendency report for Kobe Bryant, for example, warned of his ability to sell pump-fakes and get under defenders to draw fouls. This became a point of emphasis for every Blazers wing.
That's the first level of statistical usage that virtually every NBA player is exposed to. Whether players take to it or not is, according to McMillan, a personal preference. Whether McMillan throws more numbers at his players, in the form of personal instruction or team-based adjustments, generally depends on the player's individual level of interest. "Some guys are stat guys," McMillan notes. "Some guys don't look at that as much." Come on Coach, name some names. "I won't get in to who [really digs into the numbers] but they know. When you're playing well, you know you're playing well. When you're not. You're not." Memorably, Nicolas Batum spent more than five minutes standing outside the Memorial Coliseum's cramped locker room this preseason, carefully studying the scouting report for Grant Hill. (Hill ended up a gametime scratch due to a minor injury.)
McMillan went on to say that Blazers management uses advanced statistics to help target free agents and even as a contract negotiation tool. While he is more open to the use of statistics than you might expect from an old-school NBA player, numbers won't be completely replacing video analysis for McMillan any time soon. "[As a coaching tool], the visual is good. Video is good as well as numbers and stats. We use both. But what's more available to us [and accessible to the players] is the video."
So what happens if the video tells you one thing and the numbers tell you the opposite? "I like video," McMillan concluded. "It is my preference."
What's useful for an NBA coaching staff can be informative for fans too. Over the summer, I used last year's year-end Synergy reports to write this post about Brandon Roy and this post about LaMarcus Aldridge. This year, Synergy has put out even more detailed reports that are updated during the season and are of great help to the media and fans, for two reasons...
- It helps confirm your understanding of player strengths and weaknesses
- It can reveal things about players and player tendencies that you might not have realized or been able to quantify
The rest of this post will focus on that second point.
Click through to read five interesting revelations from Synergy's profiles of the Portland Trail Blazers through the first 20 games of the season.
-- Ben Golliver | (benjamin.golliver@gmail.com) | Twitter
Remember, these conclusions and data apply to the first 20 games of the 2009-2010 NBA season only. The larger the sample size, the stronger the conclusions. Also, remember that these numbers are as objective and complete as you're going to find anywhere. I've presented both positives and negatives below. If your perception differs with Synergy's findings (which probably will happen, it did for me) don't go ranting for 15 minutes on a podcast panic. As I've said before, data is meant as a complementary tool to further discussion.
1. Greg Oden Has Been the #1 Most Efficient Offensive Weapon... and #2 Might Surprise You
Let me guess: you're coming around to the idea that Greg Oden is (or could be) a better second offensive option that LaMarcus Aldridge. You've seen some improvement in Oden's smoothness on offense, you wish Aldridge was a bit more assertive and you can't get enough of Oden's backboard-slapping dunks. Sure, it can be tough to watch him struggle when he's posted up and he's known to throw up an errant hook shot, but Oden has come a long way.
What if I told you, through 20 games, that Synergy's numbers say, unequivocally, that Greg Oden is a more efficient, more consistent offensive weapon than anyone else on his team, Brandon Roy included?
Stat heads that are familiar with John Hollinger's PER rankings might not be surprised by this, as Oden is the only Blazer ranked in the league's top 20. But what if I told you Roy wasn't even in second place?
According to Synergy, Oden ranks in the 83rd percentile in overall offensive effectiveness, earning a ranking of "Excellent." By comparison, Roy is currently in the 61st percentile for a rating of "Good." Aldridge? He falls in the 79th percentile for a rating of "Very Good."
How is this possible?
Primarily Roy struggles here because of his field goal percentage, which is down nearly 4 points from last season. Synergy points out that his numbers when shooting on a low shot clock (4 seconds or less) are particularly bad: he rates in the 19th percentile for such shots. Hollinger joked about flaming bag passes earlier this year. Well, I think we know who the recipient of those has been.
The numbers also show that Roy is only "Average" generating points for himself in the pick and roll (that probably doesn't surprise you), a situation he is often in. Interestingly, Roy has really struggled when set up on the left side pick and rolls when he is coming back towards the middle, confirming a long-held observation that he prefers going left rather than right.
As for Oden's success? It can be attributed to his absurdly high field goal percentage, ability to draw fouls, ability to convert put-backs and his very good shot selection. In the categories where he really struggles (getting doubled in the post, hook shots from the left block), Oden takes very few shot attempts, limiting the damage done to his efficiency. Whether Oden does this because he doesn't have confidence in those situations or because he has recognized that he's not very effective in those situations is up for debate. Either way, his ability to recognize these limitations makes him more valuable as an offensive player than you might think.
The only thing truly holding Oden back offensively is, as you might have guessed, his turnovers. According to Synergy his turnover percentage is almost triple Aldridge's and nearly double Roy's. If those turnovers start to decrease over the course of the season, you're potentially looking at a player who could finish in the 90th percentile or above for offensive effectiveness in just his second season as a pro.
Look out.
2. LaMarcus Aldridge is Underrated on Defense
4. There is a Hole in Greg Oden's Defense
If you listened to this week's Dontonio Wingcast, Kevin Pelton dropped a bomb on me: the Blazers have been more effective defensively when Joel Przybilla is playing than when Greg Oden is playing. This might seem couterintuitive to what we see while watching games. Some players seem to fear Oden, avoiding the lane. When he is challenged, Oden often comes up with blocked shots or at the very least alters the ball-handler's path. His footwork is getting better and he seems to cover a lot of ground. Not to mention, he cleans the glass better than all but a few players in the league.
So how could the Blazers possibly be playing worse defense with Oden on the court?
The answer isn't all that complicated: opposing teams have found other ways to score against him, and score efficiently. The key? Avoid pounding the ball into the paint, attacking Oden directly. By far Oden's best defense, according to the numbers, comes when he's defending one-on-one on the block. He rates out "Very Good" overall in the post, "Excellent" in some situations. Those taking an indirect or face-up approach, however, are enjoying success against him in a number of ways.
When Oden is isolated on his man away from the block, his defense dips sharply to a "Poor" rating and opponents shoot nearly 53% from the field against him. In practice, this includes jumpers, pull-up jumpers and other shots over the top that Oden might hesitate to contest for fear of fouling. When more agile bigs drift even further out -- up to 17 feet -- their field goal percentage still remains above 47%, rating Oden "Below Average" in defending these shots. Even when he defends players in face-the-basket situations in the key and at close range, his opponents shoot nearly 60% from the field and his rating is "Poor." Again, this likely can be attributed, in part, to an unwillingness to commit fouls by fully contesting shots.
The good news is that Oden eliminates many second-chance points by clearing the defensive boards so effectively. But to truly succeed defensively, he will need to improve the aspects listed above. If teams can continue to score so easily by avoiding post-up situations and extending their post player's offensive range, even slightly, his highlight blocked shots and amazing rebounding percentages won't impact wins and losses as much as they should.
This isn't a crisis for Oden by any means. You'd be much more concerned if he was leaking easy buckets right at the basket. As is, he's holding down the blocks quite well already in his second year. More effective defense away from the basket should develop along with increased experience and comfort.
5. LaMarcus Aldridge: Bad Passer?
As mentioned above, LaMarcus Aldridge currently stands as the second most effective Blazer offensively. His shooting percentage is solid and his versatility ensures that he's "Above Average" or better in efficiency in almost every offensive situation that Synergy tracks: spot-up, pick and roll, cutting, transition, put-backs, etc. To no one's great surprise his post-up game needs some work (he rates "Average" and finds himself posting up on more than 40 percent of his offensive touches) but overall he is a solid offensive weapon.
However, according to the numbers, there is something seriously up with his passing. Simply put, the Blazers really, really struggle to score when LaMarcus Aldridge passes the ball out of the post.
For most big men, criticism of their passing focuses on one thing: turnovers. In Aldridge's case, though, he avoids turnovers on 93% of his passes. For a near seven footer, even a fluid freak like Aldridge, that's impressive. But even though his passes hit the mark, they don't lead to buckets.
Look at some of these numbers. Spot up shooters hit just 30.3% overall off of passes from Aldridge when the defense commits a second man to guarding him in the post. It doesn't really matter where he passes from: when he's on the left block they shoot 29.2%, the right block 33.3%. When teams hard double Aldridge -- that is, send a second defender to really harass him with the ball, forcing a pass -- he's even worse. The extra attention doesn't lead to more turnovers but it does lead to even worse shooting once Aldridge decides to pass. Overall, spot-up shooters hit less than 17% of their shots after Aldridge has been hard doubled. Again, it doesn't much matter which block Aldridge is passing from.
For comparison purposes, Aldridge trails Oden in virtually every pass out category that Synergy tracks. Overall, the Blazers shoot nearly 7 percentage points better when Oden passes out than when Aldridge passes out when the defense commits to the post player. Both players struggle when getting hard-doubled, which doesn't happen a lot, but shooters off Oden passes shoot more than 8 percentage points better than when Aldridge is the passer. Oden is also substantially more effective at hitting cutters when the defense commits, rating "Good" and in the 60% percentile. Aldridge cannot be ranked in this category because, according to Synergy, he has only hit 2 cutters for baskets from the post during the first 20 games combined.
One possible explanation for his passing struggles is that Aldridge often uses a lot of the shot clock while methodically dribbling in the post, trying to sense when the defense will come at him. If the defense comes late and he decides to pass rather than shoot, his spot-up shooter is then forced to shoot against the shot clock, which tends to cause more misses.
Another possible explanation is that Aldridge isn't delivering the passes in a matter that is conducive to robotic spot up shooting. When players work on spot up shooting they generally receive passes from the coaching staff that resemble a swing pass from a guard around the perimeter. Aldridge generally isn't throwing that type of pass. Many times he is passing the ball from above his head (7 or 8 feet off the ground, rather than waist height). The downward angle of the pass may reach the shooter below where he might expect to receive a chest pass and thus affect his shooting motion. A pass from Aldridge might also have some extra mustard on it as he tries to beat a rotating defense.
A final explanation could be that, contrary to a Nate McMillan claim last week, Aldridge simply isn't making the right reads and his passes are finding shooters that aren't as open as they should be.
It should also be noted that the Blazers' shooters haven't been great to start this season. But, for whatever reason, Synergy's numbers through twenty games suggest that these shooters have enjoyed noticeably better results when receiving passes out of the post that come from Oden rather than Aldridge.