Author’s Note: This is an important post for me. It’s the type of analysis I would like to bring on a regular basis, combining statistical information with game tape analysis. Any feedback would be appreciated as I try to find my voice in writing and improve with each post. You can send me comments and criticisms on twitter @bradford_NBA or through e-mail at bradfordcoombsNBA@gmail.com. No criticism is too harsh. Thanks for reading and please spread the word.
Fresh off a rare Playoff appearance and armed with cap space, picks and motivation to improve, Hornets GM Rich Cho started the offseason with an emphatic THUD as starting power forward Josh McRoberts took his unique talents to South Beach. The Heat offered Josh the full mid-level (4 years, $23m) and the Hornets chose not to match the offer.
This wasn’t a Lance Stephenson/Indiana situation. Charlotte had nearly $20m in cap space to play with and were nowhere near the tax line. Rich Cho’s hands weren’t tied. He simply decided that: A.) McRoberts wasn’t worth that much money for the Hornets mainly because B.) he believed Josh’s eventual replacement was already on the roster: Cody Zeller, the team’s lottery selection in the 2013 draft.
Cho, Coach Clifford, and owner Michael Jordan showered McRoberts with praise over his eighteen month stretch as a Bobcat – with good reason. Josh was a vital piece of Charlotte’s success last season and was often the glue that held an iffy offense together. Will Cho’s gamble pay off? Can Cody replace Josh’s contributions or will this seemingly minor exchange of role players backfire into chaos?
Dissecting Josh’s Game
It’s worth looking at exactly what McRoberts did to help the offense go and to see how Zeller’s skill-set fits into a similar role. When Zeller replaced McRoberts on the floor he played a similar, though reduced, role in the offense. Clifford had both McRoberts and Zeller play mostly from the outside, involving them in a series of pick-and-pops and dribble hand-offs to get the ball moving from side to side and into the hands of perimeter ball-handlers on the move.
McRoberts thrived as a secondary ball-handler on the perimeter. He often initiated the offense at the top of the key and looked to make plays from the outside. His surprising three point shooting was a major plus, both in the points added and space created, but it was his playmaking ability coupled with a low turnover rate1 that made him such an effective role player.
Looking beyond the numbers gives us a better understanding of how and why he was so valuable. I watched every2 assist and turnover McRoberts recorded last season. With assists, I tracked what type of action led to the assist and whether it resulted in a three pointer, a mid-range shot, or a lay-up/dunk.
The idea was to identify how McRoberts operated within the offense and what types of results his actions were producing beyond just a made shot. I broke the various actions up into dribble hand-offs (including give-and-go’s), drive and kicks, drive and dishes, hitting cutters, direct passes (post entry, swing, stationary teammates), kick-outs from the post, dump-offs in the post, and fast break passes (outlet or on the break). In all, I charted 321 out of 348 assists, including the playoffs.
The first thing to look at is the offensive system and what type of actions led to McRoberts’s assists. Looking at this specific set of results strips out some of the stagnation of less involved plays like post-ups and isolations. Clifford’s system clearly called for plenty of ball and player movement with multiple series of actions taking place in a single possession.
There were very few plays that didn’t have some kind of dribble hand-off involving McRoberts and they accounted for 18% of his recorded assists. He hit cutters for 73 baskets or 22.7% of his assists. Gerald Henderson in particular stood out as a strong cutter and finisher, very aggressively attacking the rim off the catch, as did Michael Kidd-Gilchrist. Assists on what I defined as direct passes added up to 27.4% of the total. This is an important part of the offense because the majority of these passes were into the post for Jefferson to quickly go to work.
McRoberts was a very good post entry passer, having a good feel for where to put the ball and where other defenders were on the court. He wasn’t asked to create out of straight post-ups much, registering only 26 assists on dumps and kick-outs, and those numbers might be a little generous based on my classifications.
Finally, Josh assisted on 62 baskets off drives, dumping the ball off 28 times and kicking it out 34 times. He almost always looked to pass on the drive and was a very patient ball handler, never overcommitting or getting out of control. At times his drives were so patient there was hardly a difference between a dribble drive and a post-up. He committed 7 offensive fouls all season and only committed one charge on a relatively questionable call by my eyes. This tendency bears itself out in his shot distribution chart where only 33% of his shots came around the rim. For better or for worse, Josh was always looking to get his teammates involved.
One thing that stood out was how simple a lot of these assists were. McRoberts certainly wasn’t short on flair at times, but his greatest attribute in my mind is his court vision and awareness. When paying attention it’s easy to see how active his eyes were when he had the ball, never zeroing in on an individual player, the basket, or the ball. He was much like a quarterback in that regard, going through progressions and reads and looking defenders off. Much like a good quarterback, McRoberts was also able to put the ball in a location that allowed the recipient to immediately make a play. I credited him with 41 turnovers in 82 games due to bad passes and the majority of those were off deflections or miscommunications on cuts. Very few were the result of a poorly placed pass. The key to McRoberts’s effectiveness as a facilitator was his efficiency. His ability to make the right pass to the right player in the right spot made life easier for everyone on the offensive end.
The Efficient Point Forward
As the game and the analysis of the game have evolved efficiency has become a defining attribute of successful basketball teams. People aren’t just interested in the number of points scored, but how those points were scored. Part of that evolution has been determining which types of scoring opportunities lead to a higher success rate.
The once bemoaned death of the mid-range game is now recognized as a natural progression towards more efficient basketball. Three-pointers, lay-ups, and free throws are the priority for offenses. The Bobcats weren’t a terribly efficient offensive team, ranking 24th in the league with an offensive rating (points per 100 possessions) of 103.6 per basketball-reference.com. With McRoberts on the floor the team had an offensive rating of 106.2, which would have put them just below the league average of 106.7. When he was off the court, that number dropped to 101.1.
Only the 76ers and their abomination of a team had a worse offensive rating than the McRoberts-less Bobcats of last season. While those numbers are dramatic, they likely overstate his value. Turning to ESPN’s real plus-minus metric, which accounts for the value of the other players on the court, McRoberts added .03 points per 100 offensive possessions. For some added perspective, Paul Millsap posted a .04 ORPM (offensive real plus-minus) while David Lee had a -.12 ORPM. I think it’s fair to assume Josh’s true value was somewhere in the middle. He was an important cog in the offense, but not a foundational cornerstone like Al Jefferson (1.31 ORPM, 2nd among centers).
Having looked at how McRoberts helped generate points and understanding his overall impact on offensive efficiency, let’s take a look at the actual results:
McRoberts assists generated 148 layups, 108 mid-range shots, and 65 three pointers. I personally don’t frown on mid-range shots quite as much as some statistically inclined people, but even for me that’s not a great distribution. While the lay-ups are great, ideally some of those mid-range shots would move a couple steps towards or away from the basket. This is mostly a function of how the roster was constructed: Kemba Walker, Gerald Henderson, and Al Jefferson are all mid-range type players. Henderson in particular often takes a dribble in from the three point line when he could just let it fly from deep.
Thirty-five of McRoberts’s assists that lead to mid-range shots were of the direct variety. While some of those were post passes to Jefferson, others were the result of players setting up inside the arc. The team would benefit from players extending themselves out to the perimeter as it would improve spacing, something that was almost painful to watch on tape, and create more three point opportunities.
The action with the second highest number of mid-range results was dribble hand-offs. These came in the flow of the offense, usually at the elbows. Ball-handlers receiving the hand-off could certainly look to attack more often, rather than settling for so many jumpers. The team could also look to extend those plays out to the three point line, creating more space to drive or to pull up from deep, though it’s possible Steve Clifford wants that action taking place where it does.
There is one caveat to all this passing information. Only completed assists have been charted. Looking at SportVU data, teammates converted 54.4% of all assist opportunities generated by McRoberts. That’s actually slightly better than Chris Paul, whose team converted 54% of their opportunities. This provides some confirmation for what the tape showed, that not only did McRoberts find open men but he got them the ball where they had an opportunity to make a play. Additionally, SportVU reported .7 secondary (hockey) assists per game and .5 free-throw assists (passes that led to a shooting foul where the shot was missed and at least 1 free throw was made). These numbers are very similar to Joakim Noah’s, who played a very similar role as a facilitating big for the Bulls, though he did produce 1.7 more points per 48 minutes than McRoberts did.
Can Cody Keep Up?
Which brings us to the big question: What will things look like with Zeller manning Josh’s spot and what will he need to improve upon for the team to avoid a regression?
First and foremost, Cody should not try to “be” McRoberts. He is a different player, despite having a similar pigmentation and hailing from the same home state. Having watched a lot of Cody Zeller tape, specifically how the offense was run when he replaced McRoberts on the court, it’s clear that Coach Clifford is going to put him in similar spots and expects him to make plays from those spots.
Starting with the raw statistics, Zeller averaged 13.7 assists and 13 turnovers per 100 possessions. As a reference point, McRoberts averaged 32.7 assists and 8.1 turnovers per 100 possessions. Both players’ numbers come via stats.nba.com. SportVU data credited Zeller with .1 free-throw assists per game, .3 secondary assists, 2.3 assist opportunities per game, and 7.1 points created by assist per 48 minutes. The disparity in numbers between these two power forwards makes sense when you watch the games: While Zeller played a similar role to McRoberts, receiving the ball in the high post with a charge to pass, attack, or set a screen as a part of a dribble hand-off, it was in a smaller role.
Something that stood out in the McRoberts tape was the confidence his teammates had in him. They saw him as a safety valve, looking for him frequently to keep things going. They clearly did not have the same confidence in Zeller, not looking for him as quickly. Zeller was also quicker to move the ball, spending less time surveying the defense to find open cutters. By letting McRoberts walk, the front office is putting their trust in Zeller to have the confidence to look to make plays and be more than just a ball mover on offense.
Looking at the same assist categories and distributions as I did with McRoberts reveals a very different type of player. Where 19.3% of McRoberts’s assists were the result of drives to the basket, drives made up 23.5% of Zeller’s assist. Zeller wasn’t nearly as adept at hitting cutters as McRoberts, registering only 7 assists off cuts out of 81 total, or 8.6% to McRoberts’s 22.7%. Part of this is a product of being quicker to move the ball as well as being more aggressive attacking the basket. Additionally, Zeller was more comfortable operating out of the post: 19.8% of his assists came out of the post while McRoberts created 8.1% of his assists out of the post.
Zeller’s kick-outs were a valuable way of generating three point attempts. McRoberts aversion to contact as a ball-handler made him steer clear of the post, resulting in only 11 kick-outs for threes overall. In far more limited minutes and opportunities, Cody kicked the ball out of the post for 7 threes. While that’s a tiny number for an entire season, as Zeller’s opportunities and minutes increase those types of plays can be a productive way of generating points. Overall, Zeller had a much healthier distribution of shots on which he assisted. Of 81 assists, 34 led to lay-ups, 28 led to mid-range shots, and 19 led to three point shots.
Cody Zeller: STRENGTHS
There was a lot to like in Zeller’s playmaking: He was both quicker and more aggressive off the bounce than McRoberts. He was able to get a shoulder past his man and draw help defenders, then find teammates at the basket or on the perimeter. He doesn’t have the flair that McRoberts does, but he often makes the simple, correct pass.
His passes also have a zip that McRoberts’s can’t match. In the March 3rd game against Miami, McRoberts threw a nifty cross-court pass along the baseline for a Luke Ridnour three. He used a lot of spin on the ball to get it around the defenders and to his man. In a similar play while facing the Warriors on February 4th, Zeller took a dribble along the baseline then fired a one handed pass over the top to the opposite corner for an Anthony Tolliver three that I had to rewind and watch several times. Thase two plays are a prime example of how McRoberts and Zeller differ in their passing styles.
Several times, Cody showed great patience while pivoting. There was no panic as he kept the ball out of reach of defenders and was able to find a teammate for a bucket. For all his turnovers, he was very strong with the ball, not allowing it to get knocked out of his hands by feisty defenders. He used his size to keep the ball out of reach while keeping his eyes up and surveying the floor.
Zeller also used his size to make direct passes over the top rather than some of the more crafty ways McRoberts used to get the ball to guys. Cody made good use of ball fakes when passing into the post, keeping both his defender and the post defender guessing. He was particularly effective in dribble hand-offs, using his back-side to create space and prevent his man from getting a hand in to muck up the hand-off. Overall, he set much better screens than McRoberts, who often didn’t make contact with the defender (he seriously hates contact). The foundational elements are there for Zeller to build on and, in time, he should become a solid passer and play-maker for the Hornets.
Cody Zeller: WEAKNESSES
If you’ve ever watched a Draft Express scouting video (you should; they’re superb) you know that feeling of getting really excited about a player while watching their strengths, only to be crushed with the sadness of their weaknesses. Zeller’s short but sweet tape of 81 assists was really encouraging. Immediately following that with 82 turnovers3 was a gut punch and was more in line with my memory of his rookie season. November and December were particularly brutal, as he tallied almost half of his season’s turnovers, 38 out of 82 to be exact. Over those two months, he had 19 bad passes4, 7 mishandles with the ball, 7 offensive fouls, and 5 travels. The deer in the headlights description was completely accurate. Overall, he got whistled for 16 offensive fouls, threw 36 bad passes, some of which were complete head scratchers, including one to a lady on the first row and another on an inbounds play after a made basket, traveled 11 times, mishandled the ball 17 times, and even got called for 3 seconds twice.
He was often indecisive and out of control, driving to the basket and jumping in the air with no plan to shoot or pass. There were times it was clear he didn’t know where his teammates were going to be. He lacked the necessary court awareness, failing to see secondary defenders or throwing the ball into crowds. On drives he had poor balance, getting his shoulders well ahead of his legs and trying to complete plays around defenders rather than going through them.
While there are valid concerns about how the Hornets will fill the void left by McRoberts’s departure and in particular how Zeller fits that role, Cody has the pieces to do a lot of what McRoberts did so well while using his own unique skill-set. He will need to improve his upper-body and core strength, something he has talked about before.
A stronger core will help him maintain his balance and leverage his speed with the ball while better upper-body strength will enable him to go through defenders at the basket and finish for himself or find an open teammate. A stronger approach to attacking the basket will also result in more fouls drawn and less travels and desperate heaves. He could benefit from eliminating spin moves off the dribble from his repertoire altogether. His aggressiveness with the ball is a valuable trait, as he averaged 4.15 drives per 48 minutes while McRoberts averaged 3.16 drives per 48, as calculated using SportVU data. That aggressiveness also shows itself in Zeller’s higher usage rate of 18.2% as compared to 13.8%.
One thing Cody could learn from McRoberts is to have patience on his drives. If he learns to slow down at times, rather than barreling into the teeth of the defense, he will be able to create more opportunities for others as defenders collapse. His most important path to improvement is gaining experience. It took a while for the game to slow down for him, and when the playoffs came around it seemed to be back to square one. He will need to learn how to read defenses, account for the speed and length of NBA athletes, operate and know where other players are within the offense, and finish without getting blocked. That seems like a lot to improve but it should come with experience. Cody has a chance to be every bit as good an overall offensive player as McRoberts, and a better scorer and defender.
When McRoberts agreed to his deal with Miami, it was assumed Zeller would just take over his role, including the starting position and starter’s minutes. Not long into the official start of free agency, the Hornets went out and signed another power forward in Marvin Williams. He had similar shooting numbers to McRoberts, a low turnover rate, was a better rebounder, but isn’t nearly the passer McRoberts is (or Zeller for that matter). Expect Williams to be the starter and Zeller to play essentially the same role as last season, at least to start. While the idea of contending vs rebuilding/tanking is an oversimplification, it is true that Coach Clifford and the Hornets are looking to build on the success of last season. Guys like Zeller, Noah Vonleh, and PJ Hairston will have their opportunities. Clifford knows the importance of player development, but not at the cost of winning.
While everyone expects Zeller to be improved from his rookie season, he’s going to have to earn his opportunities and minutes. I predict he starts the season coming off the bench for around 20 minutes per game, just as he was at the end of last season. I would guess that the organization is hoping he will take over the starting spot by the All-Star break. While Vonleh is several years away from being ready to really contribute, the clock is ticking for Cody Zeller. If he doesn’t make strides, both as a shooter and a playmaker, the team may start looking at Vonleh sooner, rather than later.
1. 8.1 turnovers per 100 possessions and an assist-to-turnover ratio of 4.01 that ranked 2nd in the league behind only Chris Paul.
2. Videos are available at nba.com/stats. Some of the videos don’t really match up to what they’re supposed to so really I watched a video that was supposed to be every assist. A couple games were marked wrong, but the vast majority were right over the season.
3. The real numbers are 94 assists to 89 turnovers including playoffs but footnote 2 still applies.
4. Playing with the bench unit for most of the season didn’t help. I counted 5 interior passes to Bismack Biyombo that should have led to dunks but resulted in turnovers.
VERY SPECIAL THANKS to Timmy Hoskins for “The Dude” artwork. It is FANtastic.