Perhaps you remember these two tables (with stats via the wonderful hoop-math database) that I created for the piece about self-generated offense. First the PG:
Then our table of all positions, with some overlap in the guys who are especially notable:
You’ll notice that here we have many of the more successful players from Part 1 and Part 2. (In terms either of improvement or still being very good after a diminishment of success.) Damian Lillard, Khris Middleton, Jimmy Butler, Kemba Walker, Isaiah Thomas, CJ McCollum, Rodney Hood, Zach LaVine, Jordan Clarkson. Indeed, of all these players, the only guy that completely failed as a passer was Jimmer Fredette, dropping from a best of 6 assists per 40 in college to a best of 4.4 assists per 36 in the pros, and having the unfortunate burden of being a PG, a position in which the very best players are often above 7 assists per 36 and sometimes approach or even exceed 10.
Not coincidentally, Jimmer Fredette lacks athleticism. We could see this on the court, but we could also see it with his 33″ max vert, which merely confirms what our eyes saw on the court. As we’ve discussed in our two previous pieces, it turns out that the ability to pass the ball often starts with the ability to threaten the defense with one’s scoring. And that’s an ability Fredette lost when making the transition to pro ball. At least, not in any way besides shooting three pointers off the catch, which he was moderately good at.
So now let’s look at these players in more detail, looking not only at their ability to score from multiple levels of the offense with efficiency (at least in their best college seasons) but also their ability to pass the ball in college.
That’s Damian Lillard, Kawhi Leonard, Reggie Jackson, Klay Thompson, Draymond Green, Khris Middleton, Jimmy Butler, Kemba Walker, Isaiah Thomas, CJ McCollum, Jae Crowder, Rodney Hood, Matthew Dellavedova, Jordan Clarkson, Brandon Knight, Zach LaVine, Kyrie Irving and Marcus Smart.
I’d also like look at Chandler Parsons, Kyle Anderson and Cory Joseph, since we have hoop-math data on him, as well as several players who were less successful. Iman Shumpert, Bradley Beal (who is still young), Avery Bradley, Isaiah Canaan, Nick Johnson, Darius Johnson-Odom.
First Things First
The first thing we have to do is to look at the height, vert scores and passing of players who aren’t listed in Part 1 and Part 2. That is, Kawhi Leonard, Reggie Jackson, Klay Thompson, Draymond Green, Brandon Knight, and Marcus Smart.
1) The first thing we notice is the fishiness of Kawhi’s vert score. 32.5″ seems entirely too low.
2) Another thing I notice is that Kawhi’s Assist totals seems curiously low as well. You can tell he’s a very good passer that not only makes passes that progress the offense, but actually makes passes that create offense. I’ve seen him make cross-court passes that are somewhat reminiscent of Lebron (though not nearly as frequently) and those are plays few guys in the league can make.
And I think the answer might come if we look at Potential Assists per pass. As per Synergy stats, Kawhi Leonard has almost as many potential assists per game as a player like Rodney Hood (5 for Kawhi, 5.3 for Hood), but Kawhi’s come in far fewer passes (about 32 or 33 for Kawhi, about 46 for Hood), which means each pass that Kawhi makes has a far greater chance of being the kind that generates value than a player like Hood, even though Hood’s passing numbers on the surface seem better.
When Kawhi passes though, there’s actually a far greater chance that it puts pressure on the defense. Then you put Kawhi in the context of a team that can pass with players like Tony Parker, Manu Ginobili and Tim Duncan and we start to see the way to an answer. Though this is just a guess.
3) Next we tend to notice that the same trends seem to apply here as elsewhere. Taller or more athletic players have the capacity to improve, while having less athleticism is often a hindrance. (Though Klay Thompson is also probably a better athlete than the vert score would have us believe.)
4) Draymond Green throws a couple of new wrinkles into the equation in that I believe he was somewhat out of shape prior to the draft and definitely not in the condition he’s in now. That’s a factor that can skew athletic testing downward. Also, he’s the first player who profiles with a Point Guards passing ability but who plays a Power Forward position. Something we really only see in Ben Simmons this year. And something we almost never see in most years.
5) With Draymond Green, we are also witnessing, to some extent, the limits of his athleticism. I believe part of the reason he’s been a less effective player from the field in Games 3, 4 and 6, at least on offense, is because he’s often being guarded by smaller defenders or Lebron and getting many fewer 4-on-3 opportunities due to the adjustments Cavs have made on their defense, choosing to trap much less frequently and for the most part abandoning the strategy on screens, the plays that were in large part responsible for unleashing the Warriors juggernaut in Games 1 and 2. (Well, that and Cleveland’s decision to double players off-ball. Either by accident or on purpose.)
6) One could make the same argument for Kyle Anderson, a player who’s still a good passer for his position in the NBA, but whose effectiveness is much diminished from college. Anderson has 7.6 Assists per 40 as a sophomore in college, but only 3.6 Assists per 36 last year, his second in the NBA.
One reason for this is perhaps the Spurs system, but the major reason is almost assuredly that Kyle Anderson went from primary initiator in college to off-ball Wing at the NBA level. At 6’9″, he has great size for the position, but Anderson lacks explosion. Still, his ability to pass at the Power Forward position, which he played 34% of the time this year (according to basketball-reference) would be a perpetual mismatch.
7) If we look at Draymond Green’s career, we see the major difference from Year 2 to Year 3 is that Green went from playing Power Forward and Center around 1/3 of the time in his rookie and sophomore years, seasons in which he was even utilized at SG, 5-10% of the time, to playing PF and C 100% of the time. And thus Green jumped from 7% and 11% Assist rates in Year 1 and 2 to 16% and 29% in Years 3 and 4. That possibility was created because of the advantages Green holds at the 4 and 5 that simply weren’t present at the 2 or 3.
Second Things Second: Scoring Data for Shumpert, Beal, Johnson-Odom, Isaiah Canaan, Nick Johnson and Kyle Anderson.
I would include Avery Bradley as well, but the data’s not available. However, there’s several predictive indicators in his profile beyond his inability to easily get his own shot at the rim.
1) Looking at the first five players, the first thing we notice and that all of these players have in common is that none of them were all that great at getting to and making shots at the rim. Shumpert and Beal are the best at 1 unassisted make at the rim (including putbacks) every 7.5 attempts from the field. But when you realize that they also have elite scoring percentages at the rim, both at or near 70%, you realize that these players really weren’t challenging the defense off penetration all that often.
A fact that’s particularly concerning for Shumpert when you realize he was an upper classman and his team’s primary ball handler. Beal was at least young, only a Freshman and sharing the scoring load with Kenny Boynton and Erving Walker.
2) None of these players have a particularly impressive FTr, though Kyle Anderson’s is decent. Which is another indicator that they didn’t greatly put the defense into trouble off of penetration. And players who can’t do that are often going to have a difficult time creating for others at the NBA level.
3) As a contrast to Kyle Anderson, look at Andre Miller who had a .579 and .490 FTr, as compared to a .399 and .490 FTr. It doesn’t seem to be that great of a difference, and Andre Miller did finish with a career rate similar to that of Anderson’s sophomore season, but it’s probably a lower bound number for success in this area, at least for relatively unathletic players who can’t shoot off the dribble for distance.
4) I’m pretty sure the hoop-math data for older players is a little wonky. For instance, Bradley Beal’s numbers on hoop-math don’t match up with those on sports-reference. And this is true for Parson and Joseph as well.
5) It should be stated again. Kyle Anderson and Bradley Beal are both very young and either showed great success in college or possess decent athleticism and scoring ability so we shouldn’t close the book on either getting good just yet. Regardless, we can see that not scoring very frequently at the rim, as compared to one’s overall FGA, is potentially a negative indicator going forward.
One of the only players to overcome this is Khris Middleton, and he was a good passer in college and is a threat to score in every area off the catch or off the dribble.
6) Looking at the last at Chandler Parsons, we see a player with some real strengths in his profile and some real question marks. On relatively low usage, which is actually a sign of his intelligence as a player and not a detriment to his profile, as it is so often interpreted with respect to many players. (One has to know the player and the system to know what the number of FGA a player takes truly means.)
One strength is scoring at the rim. He does that as well as many of the most athletic PG, and most of them are PG. We can tell this because over 1/6 of his FGA end in an unassisted make at the rim.
7) Besides that, Parsons had a questionable shot from mid-range and from the free throw stripe, but excellent three-point numbers in semi-low attempts. (A high percentage of unassisted makes.) We can also guess by the ridiculously high FG% from two for the rest of Parsons career (57% is the lowest, 62% the highest) that this was a one year blip in which Parsons mid-range jump shot learned how to impersonate a barf bag.
8) We know Parson’s gone on to reasonable, but not great, success passing the ball in the pros and also very good success in terms of shooting efficiency and scoring from all areas. And it’s all there in his profile, when you look at his size, his scoring success from multiple areas of the floor, his dribbling ability, and his hugely positive passing indicators. (Hugely positive ratios in terms of Assist to Turnover and Assist % to Usage. Also, he’s a 20% Assist Percentage guy from SF/PF in college.
9) I’m sorry I didn’t talk about Joseph earlier. I initially forgot to put he and Parsons in the table, which is the reason why. Joseph is a guy who greatly resembles the others in terms of getting to the rim. In fact, he’s even worse. But Joseph is stronger than virtually every other prospect in another area. In terms of creating and making unassisted jumpers.
Despite only shooting 296 shots in the sample, 51 unassisted makes on jumpers. And he shot 38% overall on two-point jumpers and 39% overall on threes. These are hugely positive numbers, especially for a 19-year-old.
10) I’d also like to point that Joseph is yet another prospect for whom a crappy field goal percentage at the rim in college meant nothing towards his pro success. He even shoots a far greater percentage of his shots there now as compared to when he was in college.
30-40% in recent seasons, and he’s been above 60% at the rim in every year since his second. 61.5% for his career. With seasons as high as 65% and 66% in there.
If a player has a hole in their game, missing at the rim is a good place for it to be. The NBA really is a more open game and that’s the easiest place for a player ultimately to improve, at least if he has any athleticism.
Other Indicators That Might Lead to Future Passing Success: Ancillary Passing Statistics
Assist Percentage, Usage and Their Relationship, Part 1
1) This table is arranged by assist percentage. What we’re going to see as we move through these statistics is that TJ McConnell was an incredible statistical prospect by many measures, and that extends beyond these tables to his actual defensive effect on the court, to his scoring numbers where he scored from multiple levels efficiently and on unassisted baskets.
Around one sixth of his field goal attempts were unassisted makes at the rim. 63 unassisted makes from mid-range where he shot 48.6%. These are elite college numbers and we saw the effect of that this year.
It’ll be good to keep in mind when we examine the prospects that most resemble McConnell, not just in terms of the way he plays (for which Fred Van Vleet is as close a match as can be found) but in terms of being a PG who doesn’t have the requisite athleticism/frame combo to make him an ideal match for the NBA. And in that sense, we’re also talking about Kay Felder (my favorite of this group due to there being real upside), Tyler Ulis and Nic Moore. These are all top-notch college players who for the reasons outlined above might not translate.
2) Assist Percentage is a decent indicator at the extremes, especially if you read it through the lens of what position the player played in college and what position the player might play in the NBA. But there are massive problems to viewing the numbers out of context.
3) For instance, at the bottom of the list we find relative non-passers for their height or position, Avery Bradley (an excellent ball mover but not a creator), Bradley Beal, Zach LaVine (I would not be surprised if his rookie year numbers are his best in terms of volume of assists, since he’s since been moved off-ball and for good reason), and Jae Crowder (again, a good ball mover), but we also find Jimmy Butler. And we know that Jimmy Butler has since become a very good passer, and as we’ve discussed before, there are a number of indicators (Scoring, free throw rate, turnover numbers, ancillary passing statistics, athleticism for size, rebounding) that should have allowed us to see that such an outcome was on the table from the start.
Or at the top of the list, we get McConnell, Turner, Payton, Thomas, Green and Jackson, but we also get a number of players whose passing is much less effective (if not volume, in terms of effect on the game) like Dellavedova (still okay, but there’s a reason he’s not getting minutes in the finals and Shumpert is and it has to do more than with just defense), Kyle Anderson (moved off-ball and not to PF), and Marcus Smart (moved off-ball, definitely until he gets a jump shot.)
These players who moved off ball are still all great passers for their role, but their passing no longer affects the game in the same way, since they aren’t as often initiating plays. And that’s important to remember with regards to players like Denzel Valentine, Kyle Collinsworth (who might play point), Caris LeVert and Thomas Walkup, for which this draft has a fair few.
4) One thing that I’d like to point out is that tall Wings with over 20% Assist Rates have decent track records of becoming good players, even when they don’t become great passers. And even looking at the ones that show up on this list: George, Hayward, Middleton, Parson, Thompson, Dr. Green (even better when they are a PF), Turner, Anderson, all of them are at least useful as ball movers and some of them add significant value with their passing.
Keep it in mind this year, because this year has a lot of tall Wings with good passing numbers. It’s something you rarely see in abundance, though it should be pointed out that even this point has caveats which we can understand by looking at a couple of players from last year, JP Tokoto (no scoring ability from anywhere but the rim and he couldn’t get there himself) and Royce O’Neale (sub-standard athleticism with ultimately questionable shooting numbers due to them being mostly assisted).
5) It doesn’t really relate wholly to this category, but I’m going to try to mention every player I included with some level of detail, and I wanted to say something about Otto Porter. I almost didn’t write about Otto Porter. I came this close to not including him, and then I remembered his scoring line in his two seasons at Georgetown.
Which is to say, I suddenly realized that Porter Jr. is perhaps the key prospect, or at least one of them, to understanding how scoring and the expectation of passing in the future are related. What’s more, understanding Otto Porter’s scoring line at Georgetown in context with where he’s had success and failed is perhaps critical to understanding just who Taurean Prince is as an offensive prospect (a non-passer now, a non-passer going forward) and perhaps also, to a slightly lesser extent, Patrick McCaw.
For in spite of the fact that Otto Porter Jr. looked like a good passer at Georgetown (and we can say that his 18.4% Assist Percentage is very good for a Wing, even when stacked against a group of the best guys), there are reasons we should have been highly dubious about his passing ability translating. To see this better, we should look at one of Otto Porter’s scoring lines from Georgetown. (Thank you hoop-math.com.)
What we see here is a player who can’t generate offense. Only 29 unassisted makes at the rim despite 350 overall field goal attempts. That’s less than a 1:11 ratio, when 1:6 or 1:7 is good. And it gets worse when you realize 22 of those unassisted makes came on putbacks. Which is to say, he never got to the rim on his own dime.
That’s one bad indicator. Now for the second.
Otto Porter Jr. had excellent success from mid-range and three, shooting 41% on mid-range jump shots and 42% on threes. The problem, almost none of them were self-generated. Only 21 two-point jumpers came on his own dime, and only 6 three pointers. And it looks even worse when looking at the percentages which, especially from 2-point range, are exceedingly poor for any prospect, in terms of the level of offense he generated himself.
This didn’t necessarily make Porter Jr. a bad prospect into the future, and we’ve seen Porter Jr. become a good player already, but it does mean that we should have been viewing his passing numbers mostly in the context of what it said about his potential intelligence as a player (even there he looked like a lost turkey his rookie year) and not as an aspect of the player’s game where we should expect production. Which means that Otto Porter Jr. was a player who was always going to have to rely on his shooting, his rebounding and his defense to provide value. Luckily for him, there are a decent amount of indicators to suggest that he could be a fairly elite jump shooter in the near future, and he’s a solid enough defender to provide some value.
6) So why does this relate to Taurean Prince? Let’s forget that Taurean Prince has all the signs of a bad passer and low acumen offensive player (very poor assist to turnover rates, very poor assist percentage to usage, semi-high turnovers in general) and let’s skip straight to the hoop-math data, which should allow us to see through the fact Taurean Prince had 3 assists per 40 this year, which is decent enough for a college Wing, at least on the surface:
Here we see the same kind of profile. Only 30 unassisted makes at the rim out of 420 shots. Which is to say only one in fourteen field goal attempts ended in an unassisted make at the rim. What is worse, 28 of those 30 unassisted makes came on putbacks. Meaning Prince never got to the rim.
We also see somewhat better numbers for the mid-range and three (37 unassisted two-point jumpers, 9 unassisted three-point makes) to go along with correspondingly lower percentages. (36.5% from two and 34.3% from three.) Not even Free-throw rate helps Prince Jr. out, as he’s only at 34%. Which is to say, it’s probably not going to matter than Prince Jr. is sudden and athletic, definitely more so than Porter Jr., because his decision-making and ability to get to the rim were poor in college. That’s likely to remain the case. Meaning that when we evaluate Prince into the future, we should be banking most of our offensive evaluation on how we view him as a three-point shooter and as a cutter, because those are the areas of the offensive game he’s likely to bring value.
Does that make him a bad prospect? Not necessarily, but it certainly limits his potential value, especially if you don’t view him as a high volume guy (I don’t know) or an exceptional three-point shooter. (This is impossible to tell based on the numbers, but if I were to conjecture a guess, I’d guess he ends up between 32% and 42%, with that 34% to 38% in non-career years being the sweet spot, meaning he’ll bring offensive value in some years and not in others. Though he’ll still be dangerous enough from distance that defenses have to pay attention to him. Just a guess. Could be a 3&D player. But could just as easily be a D without the 3 player, and maybe without the D, which would just leave rebounding.)
Assist to Usage Ratio Part 1
1) Assist Percentage to Usage Rate is one of my favorite easy ways to put Assist Numbers into context. What we quickly see is that there are far different standards for Wings than there are for PG.
PG who do bad here, anything around or beneath 1:1 being bad, are mostly players you want to avoid, at least as your team’s PG. (Avery Bradley is a fine player, he’s not a PG, and if he couldn’t guard 2G, there might be some problems.)
2) That being said, there is an exception, known as the Damian Lillard corollary now and into the future. Which is to say, if the player has scoring data like Damian Lillard and athleticism like Damian Lillard and dribbling ability like Damian Lillard, you might be a in line to see some meaningful improvement. And if you’ll note in the scoring data up top, Damian Lillard’s 88 unassisted makes at the rim, 40 unassisted three pointers, excellent percentages and high FTr of nearly 50% (it’s amazing considering how often he shot from three) puts him into rare territory.
In respect to this year, this is the pathway for players like Demetrius Jackson and Josh Adams, who are each like a poor man’s Damian Lillard in terms of college profile, athleticism and frame.)
3) Jordan Clarkson is also a player who improved his surface passing numbers, at least as a rookie. And he got to the rim almost as well as anyone not named Elfrid Payton. 108 times. (I do think he improved his surface passing numbers without improving his passing ability or effectiveness. But that’s partially a discussion for another time.) He’s also proved to be a somewhat capable three-point shooter in the league as well.
Which is to say, getting to the rim is probably the most important part of this equation.
4) I wouldn’t be surprised if youth were ultimately an exception as well, given that there’s athleticism and scoring ability. (If there were numbers for Jeff Teague, it would perhaps bear this out. Anthony Barber is a poor man’s Jeff Teague.)
5) So what does a good PG look like? TJ McConnell. 2:1 is what a great passer looks like. These are almost always guys with some level of vision and definite willingness to pass. And we’ve seen McConnell, Dellavedova and Evan Turner all had regular season value with their passing despite being below average athletes for their position.
6) It should also be noted that every player above 1.2:1 on this list ended up a real-deal NBA guy. That’s not to say all such players are NBA guys. These are all players who had many other tools as well, whether it was effort, intelligence, scoring ability, defensive ability, etc . . . The other parts of these players games were also NBA worthy, but 1.2:1 is a pretty good indicator for any player, even a PG.
7) For Wings, it’s much different, which is to say, we shouldn’t expect the same passing numbers from them. And we should point out that even the poor player’s on this list have decent numbers for Wings. Players like Buddy Hield or Jamal Murray, who are basically .45:1 guys are worse. Which is to say .5:1 or .55:1 should be seen as a kind of lower bound. Below that, I doubt we’re going to see that much meaningful passing from the player.
I believe Brandon Roy, even with his turnover problems and his exceedingly low minutes totals, was better than that as a Freshman, and if he wasn’t as Freshman, he was definitely much better by the time he was a sophomore.
8) We see a reason here CJ McCollum probably projects better as a small Wing than an average sized PG. He had the passing numbers of an above average Wing in college and he’s become an above average passer as Wing in the pros, but it’s dubious as to if he could handle the passing requirements of the PG position.
Though he’s another player who had excellent unassisted scoring numbers in his best season. 87 unassisted makes at the rim, 90 unassisted jumpers, and solid to excellent percentages across the board considering burden.
9) Nick Johnson is an interesting player as a small guy with good numbers for an off-ball player but bad numbers for a guy who would have to stay on-the-ball. As he played next to McConnell, it’s always going to be difficult to know exactly what to make of this kind of player. And I’d bet a lot of the burden falls on how well they score. With McCollum, he came up roses. With Nick Johnson, in limited minutes, not so much.
10) Middleton, Hayward and George’s numbers are exceptional for Wings. Klay Thompson’s is very good, but he’s also noticeably less sudden than George or even Hayward.
Middleton and Thompson represent two possibilities in terms of passing development for a less athletic player. (Thompson is fine on defense, a lot because he tries like a motherfucker, but he rarely wins off the dribble and when he does it’s entirely because opponents have to overplay his jump shot. Having an exceedingly dangerous jump shot is the first step of a dribble-drive attack. Players with quickness with the ball and a jump shot often win their match-ups before even setting the ball down on the floor, so it’s lucky for the league there are exceedingly few of them. We are getting treated to a couple in this Finals though in Kyrie and Curry.)
11) It should be noted that this Middleton and Thompson corollary doesn’t wholly fit with a player like Denzel Valentine. He passes like Point Guard, not like a Wing with a plus skill. There’s a difference.
12) Jimmy Butler. We have to talk about Jimmy Butler again. This 0.6 range is kind of a threshold. It represents in many cases a willingness to pass the basketball, which is the first step to actually being a good passer. If a player isn’t willing to pass the basketball, he’ll never be good.
This, unfortunately, is something we often see with Kyrie. If Kyrie were a more willing passer, it’s highly possible he’d annually be in top 5 to top 10 offensive players in the league. Effort on D and willingness to pass are really the two big holes in his game. And in the games we see Cleveland winning, Kyrie definitely has done better in both. (He’s also displayed brilliant shot making ability, even when he takes some bad shots. Which is to say, he’s in many ways been closer to the player Curry has been over the past two years than Curry. One reason perhaps, he doesn’t premeditate what he’s going to do and thus is a more willing driver of the basketball. Whereas Curry’s so amazing he can somewhat premeditate what he’s going to do and still do it. But this occasionally leads to questionable decision-making in terms of shooting the basketball or trying to drive it.)
13) Of course, there’s no hard and fast rule about 0.6:1 being great. This is just my experience from doing lots of division and being kind of a nerd about this. There’s just a point below which improvement is very unlikely, and that point is probably somewhat below 0.5:1 and actually closer to 0.6:1.
Though I’d be lying if I told you I knew exactly. I look at the numbers and watch the game, but a lot of the work happens in my head beneath my conscious knowing and how I came up with the solution is sometimes mysterious even to myself. At times, that makes it much more difficult to explain why the answer seems solid to me (or at least approaching solidity.) And it makes it harder to back-trace as well in order to understand just where certain thresholds might be.
14) We’ll probably talk more about Jimmy Butler as we go on, but this is Butler’s poorest indicator, and it’s still more than solid. If you notice the scoring data above, Butler is close to a 1:6 ratio of unassisted makes at rim as compared to FG attempts. He hit 44 unassisted jumpers from two and three combined, despite low usage. He has excellent percentages on two-point jumpers. (38%). Half of his threes were unassisted makes, which explains the somewhat poor percentages to a degree. (Which aren’t even that poor, considering the whole season or the whole sample, the hoop-math data for that season is flawed.) Excellent free throw shooting which presages shooting success. (78%.) And also a crazy FTr or near .7, and it’s even better when just considering his two-point shots.
15) Kawhi. We need to talk about Kawhi. He’s a player whose passing numbers haven’t improved from college, but whom I feel has made considerable strides as a passer. The data would say he has less likelihood to project as a passer than many players, mostly because his shooting success was somewhat unpredictable. Though everything was unassisted, he hit a lot from the rim (53 times) and there are a ton of unassisted makes on jumpers (63) on somewhat low attempts. (363.)
Kawhi’s passing and assist to turnover numbers, which we’ll spotlight later, are all notably worse than Butler’s. We shouldn’t forget how young Kawhi was (an Age 19 sophomore) or how athletic. And we’ll see the rebounding and defense numbers as well, which bare this out, even if his vertical test didn’t. He was athletic and big and young, plus a willing and able passer in both of his seasons. These are pretty good signs.
When you compare Kawhi to a player like Andre Roberson (6.8% Assist Percentage, 14.9% Usage, .456:1 ratio, which is similar to his .44:1 career ratio) we can begin to see a noticeable difference between potentially like prospects. (Not like in actuality, like based on their surface statistical appearances as hyper athletic rebounders and defenders who don’t score efficiently or well from distance.) Roberson is also perpetually below 1:1 assists to turnovers and considerably worse than Kawhi.
16) I’m going to say this now. Sub 10% Assist Percentage, probably a flag, no matter what the Usage. And especially for a non-Center I can’t remember too many players like that becoming passers, if any. Centers may be an exception.
Assist Percentage, Usage Percentage and Relationship (Career Numbers)
Assist Percentage (Career)
1) I won’t say that much here, since we’re covering the same ground as we’ve previously covered. It did seem important to put the numbers down to add extra context.
2) One thing that becomes very obvious that being a PG leads to a better Assist Percentage. Every one at the top of the list is a PG. Everyone except Draymond Green, which should tell you how special Draymond Green was as a passer. (Green did occasionally or even often initiate Michigan State’s offense, but it wasn’t every time since they did have smaller players who could dribble and perform traditional PG functions as well.)
3) Point Guards with sub 25% career Assist Percentage numbers, we should perhaps question. I don’t know how well this would bear out if we had a larger group of players. But here the only guy even close to being below 25% to become a good passer was Damian Lillard and we’ve already gone over some reasons as to why he’s special.
4) I’m guessing that being above 18% for one’s career is a good sign for a future Wing, with the obvious caveat that these players need other parts to their game to play in the NBA.
Assist Percentage to Usage Ratio (Career)
1) The same basic rules for ratios apply. Of course career numbers should be expected to be slightly worse than seasonal bests. So adjust accordingly.
2) One other point is that for most part, I’d bet the seasonal bests are more important, but the story of the numbers is also important. For instance, Isaiah Thomas’s seasonal best occurred when he was a junior, since he was becoming a better and more able passer throughout his entire career.
Whereas for a player like Buddy Hield, his seasonal best came as a Freshman, probably mostly by the happenstance of not shooting all that often. He’s consistently gotten worse as a passer throughout his career (especially considering these types of peripheral ratios) and so we should consider his seasonal best as an outlier that doesn’t have that much to tell us about who Buddy Hield as a player.
This is important to keep in mind. And ultimately, trend analysis should be as important as that of finding critical thresholds. Sometimes though, you don’t have to assign exact numbers to understand the phenomena. The story the numbers tell is that obvious. Such is the case with the career of Buddy Hield. And it’s largely what is born out on the court, were we to watch how he plays.
He really is a one or two skill guy on offense. (Shooting for accuracy and shooting for volume.) And you can largely find guys like this on the open market, especially if they don’t play D to great effect. Danny Green and Wesley Matthews pre-injury are the outlier guys as 3&D players. (Shane Battier and pre-superstar Kawhi being others.) The fact that so few 3&D guys produce as much value should tell us something.
3) There are other kinds of player whose progression send mixed signals. Players whose passing success vacillates up and down. Or perhaps their passing success trends down but still remains decent overall. Kemba Walker and Iman Shumpert are two examples on this list. Jameer Nelson and Andre Miller are two examples from off of it.
These are Point Guards who pass less as their scoring responsibilities become greater. the difference between Kemba Walker, Jameer Nelson, Andre Miller and Iman Shumpert is scoring effectiveness on jumpers, at least out to mid-range, and the ability to get to the rim. (Andre Miller shot 60% from two as a junior and 56% as a senior, he could really do both things.)
We can perhaps see this best by reviewing the shot-data above. Kemba Walker had 95 unassisted makes at the rim. The ratio, of slightly less than 1:7 of such makes to field goal attempts is slightly less impressive, except then you remember, he managed to shoot the ball 700 TIMES (!!!!!!!!!!!)
That’s a mouth with a lot of exclamation points, as 700 is a lot of attempts, and almost no college player ever could shoot the ball that many times and retain any sense of efficiency. Much less actually have somewhat reasonable Assist:Usage numbers. It doesn’t say a lot about his ability to be a PG, and from a passing perspective, Walker does leave a lot to be desired, but passing is enabled by scoring success and Walker has enough of it to be a significant player in the league.
Walker’s success in college of course didn’t stop at the rim. He also had 136 unassisted makes on two and three-point jumpers. That’s like 3 or 4 seasons worth of unassisted makes for even a fairly good shooting profile. That’s how insane that junior year was from scoring perspective.
Assists Per 40 and Turnovers
Assists Per 40
1) One of the best questions to ask about prospects is this: Does the prospect have an elite skill? Almost no prospect becomes a great player without doing something exceedingly well, Jimmy Butler and Khris Middleton included.
TJ McConnell and Draymond Green had elite skills at least in terms of passing, if not elsewhere, and a lot of other game to back up those skills.
2) 8+ assists per 40 guys are rare, especially if they have other skills to back up that passing. Felder, Dunn, Valentine, Collinsworth have hit that number in one of the last two years and all have some game to back up their passing. Elite skills have a lot of room for attrition.
3) Just an aside on Felder because I’m guessing I talk about him a lot when I address 2016 players who might pass in the future. His profile reads like Isaiah Thomas mixed with Pierre Jackson mixed with Ty Lawson mixed with Kemba Walker. Three of those players became highly successful offensive players and the other was a sub 6’0″ guy with an Achilles injury, which is the kind of player that can least afford to lose athleticism.
I’m probably going to bet pretty big on Felder in my final rankings, one because I’m just going remove any player, besides those who project at PF/C or C, who doesn’t project to possibly do at least one of these two things: Pass or Play defense. I’ll take the miss on one and two skill shooters. One of the things a Big Board should do is not just to pick out the players one thinks has a chance to be most successful, but also to pick out the ways one wants to miss on prospects. The cost of selecting one prospect is always at the risk of not selecting another.
4) Clearly the numbers mean different things for different players. Taller players, usually Wings, have much less of a burden to be exceptional passers.
5) If Assist Percentage is available, it’s generally the better number than Assists per 40. Though looking at both can’t hurt.
Turnovers per 40
1) In the middle of the pack, I’m not sure if turnovers per 40 means all that much. But at the extremes, I do think it tells us something potentially useful. Evan Turner, Iman Shumpert, Elfrid Payton, Paul George and Klay Thompson were all fairly high turnover players in their best seasons as creators. (Shumpert without even creating all that well.) Evan Turner has had consistent turnover problems in the NBA, as had Paul George. Klay Thompson has become a total non-creator. Elfrid Payton has a high turnover percentage but a somewhat acceptable level of actual turnovers.
As for Iman Shumpert, I don’t know what to say.
I will say that during the last game my wife proposed a rule change where Cleveland should get to call a timeout every time Iman Shumpert touches the ball, just so Lebron wouldn’t get that look in his face. (You know the one, somewhere between resignation and FYM. Fuck Y’all Motherfuckers. And by “Y’all” I mean Chalmers and Shumpert, for whom Lebron’s on court contempt seems mostly reserved. )
I don’t think his was always this spastic. Not even last year. Though he was never much of a creator. This year, Shumpert seems much more frazzled on the court, and it’s not just the change in hair style (which is awesome and for which Shumpert gets my gratitude), and it’s not just on offense. It’s like he’s not even all there on defense. And when Shumpert came into the league that was his bread-and-butter.
2) At the extreme low-end we get Jimmy Butler, Rodney Hood, Otto Porter, and Jae Crowder. They haven’t all become passers but they’ve all become good players. Of them, Jimmy Butler became a good passer. And Rodney Hood, while still not great, has become good at passing in a way where he’s genuinely surprised me. Especially for a guy whose not super sudden.
Harrison Barnes was another super low turnover guy in college. If these guys have other skills, it’s definitely a good signifier. Though obviously the higher a player’s usage and creation, the more they should and will turn the ball over. Turn overs generally happen out of dribbling or passing the basketball and if player’s only finish, they really shouldn’t have all that much opportunity to turn it over.
Assist to Turnover Rate
1) I used to underrate the importance of this statistic. I always thought it had some importance, but before this year, I didn’t see how tied in it was to a player’s success in any number of categories that extend beyond passing.
Just look at this list basically all of the players with elite assist to turnover seasons are good except Nick Johnson.
2) Now this list is obviously selected to be the best-of-the-best, so we should expect these players to be good, but at the very least, you can see that they do all have solid or much better than solid assist-to-turnover seasons.
3) Some of that is because the player’s are PG. Basically all player’s with a decent amount of assists have solid assist to turnover seasons. At a certain point, it’s just impossible to turn the ball over more than you pass it, and we can perhaps see that a little as we move down the list. But that’s not the case with the players at the top.
To be a non-PG with a better than 1.5:1 Assist to Turnover Ratio is really good. To be a PG with a better than 1.7:1 Assist to Turnover Ratio looks to be a sign of something as well.
4) It’s not an end all be all, but it is indicative of something deeper than just assists and turnovers, deeper than passing and decision-making, something which perhaps correlates to intelligence, but also to dribbling skill and perhaps in some ways even to functional athleticism. (For how a player uses their athletic gifts is not exactly the same thing as possessing athletic gifts in the first place.)
5) Once again, McConnell is elite at another thing related to handling the ball and decision-making.
6) Pay attention the non-PG. The ones who show up well here are guys to mark down.
7) Pay attention to guys with low assist totals who are still approaching or better than 1:1 in Assist to Turnover Ratio. Guys like Butler and Hood and Hayward (his freshman season.) It’s very difficult to have a better than 1:1 Assist to Turnover Ratio and not to generate huge amounts of assists. If the player does this and has other parts of their game that work well, they are ones that I would note.
Assist to Turnover Ratio (Career)
1) Again, we’re covering the same ground as before, so I won’t say that much. I do want to mention Paul George. He’s not necessarily at the top of any individual stat (besides turnovers), but he does well in everything. He scored at all levels of the offense with efficiency. He’s huge. He’s very athletic. He’s a better than 22% Assist Percentage guy from the Wing position. He’s a better than .8:1 Assist Percentage to Usage Guy. In his best season, his assist to turnover rate isn’t as good as you would like it, but it’s still close to 1:1. And as we’ll see in the next section, his rebounds, steal and blocks show his functional athleticism and intelligence is all there.
When looking for a player who might be a good passer from the Wing at the NBA level, George checks a lot of boxes.
2) Hayward has basically the same offensive profile as George, except slightly better passing stats and worse three-point shooting in his sophomore year.
3) Oladipo is an interesting guy as well. Since in many ways he profiles less well than many others on this list. But he has two things going for him. Relative youth for his experience. (He was a very young college junior.) And also supreme athleticism, which we already know if we’ve ever seen him, but which comes across in his rebounding and defensive stats. Athleticism provides opportunity.
Rebounds and Defensive Statistics
Per 40 Rebounding
1) The best things would obviously be if the player had great passing statistics. However, I just can’t shake the feeling that there’s some connection between rebounding, a player having solid if not better peripheral indicators (A:TO, AST% : USG, Unassisted Scoring, athleticism, height) and the player’s potential to pass in the future.
2) They don’t seem like necessarily related skills, but when I see Steph Curry greatly out-rebounding his athleticism, or Kyle Lowry out-rebounding his size, or Rajon Rondo, Jason Kidd, Magic Johnson, Lebron James, Larry Bird, Draymond Green, Russell Westbrook, Kevin Durant, Scottie Pippen, Clyde Drexler, Grant Hill, Brent Barry, Nate McMillan, Paul Pressey, Reggie Theus, Tracy McGrady, Anfernee Hardaway, Dwayne Wade, etc . . . Many of the games great passers are also great rebounders. Or at least good ones. (Though of course this is not true for all of them.)
3) Just look at TJ McConnell, whose rebounding stats are better than they appear on the surface because he just didn’t O-Rebound in college. He comes into the NBA. First year. No athleticism. 5.6 rebounds per 36.
4) I don’t know how to quantify this link, or what you would do to draw it out, since it’s far more complicated than a direct relationship. There just seems to be a relationship of some kind and if you look at this list, which was pieced together mostly by finding players who either had very good peripheral indicators (almost everyone) or players who had somewhat poor peripheral indicators for their position (Shumpert, Johnson-Odom, Bradley, Porter Jr, Beal, etc . . .), basically every good player with solid indicators and a season with 7 rebounds per 40 or more is some kind of passer in the league.
5) That’s Kawhi Leonard, Kyle Anderson, Draymond Green, Evan Turner, Gordon Hayward, Chandler Parsons, CJ McCollum, Victor Oladipo, Paul George, Jimmy Butler, Reggie Jackson, Marcus Smart, Khris Middleton. Good list.
6) The others above 7 Rebounds per 40:
Jae Crowder, who is on the lower end of promising in terms of AST%:USG Rate. He’s a very good ball-mover, but not a creator.
Otto Porter Jr, who has better peripheral numbers but much worse scoring numbers (in terms of creation.) Very good shooter. Low turnover player.
Bradley Beal didn’t have great peripheral or creation stats, but he’s also only 22 and might still improve. Who knows where he’ll end up?
Shumpert is Shumpert.
7) What you’ll also notice is that all of the players who are worse at passing, not just as compared to their position, but rather the guys we’d make more subjective claims about, like “He doesn’t have very good feel for the game” are consolidated at the bottom of the list.
On this list, Cory Joseph and Matthew Dellavedova are the only guys who were poor college rebounders with good feel for the game. Even then, they are not as great at passing the basketball in NBA as many of their brethren who rebounded better than them.
8) You can note this by comparing Dellavedova’s college numbers to those of Lillard as a pro and vice versa. Dellavedova is good, but Lillard is now as good or (more probably) better than him, despite having less natural ability in that part of the game.
9) So don’t ask me to quantify the link. But I will propose that one exists, which someone better than numbers than I might be able to draw out.
10) The Sports-Reference database is incomplete. I should mention it. It’s obvious looking at my table, but I should write about how it has affected my table. Sometimes I was able to supplement data from DraftExpress. A couple of the career rebounding numbers are guesses which are very probably close to the actual numbers. And some of the career data is actually culled from two or three seasons, rather than the player’s whole career, since that’s what Sports-Reference had available in terms of advanced stats.
Draymond Green and Jimmy Butler are perhaps examples of this. Which is to say, the table isn’t perfect, but it can give us a pretty solid idea of what’s going on.
Per 40 Steals
1) As much as I think there’s a link between rebounding, solid passing indicators and passing in the future, that’s how unsure I am about a link between steals and passing. If there is, I find it much harder to make out, just based on this list. But if athleticism is related to passing success and steals are a great indicator not only of functional athleticism but also intelligence, it would make cognitive sense that some link could be drawn.
2) The problem is that there doesn’t even seem to be a lower bound threshold as to how bad a player can be at steals and still become something of a passer. Rodney Hood is scraping the bottom of the barrel at 0.9 steals per 40 and anyone whose watched Utah or read Zach Lowe’s pieces about them knows he’s become surprisingly good there.
3) Of course, it’s definitely not a bad thing to be good here. The better players and passers do still tend to be better than worse. Having at least one season at better than 2 steals per 40 or maybe 1.6 steals per 40 does seem to be something of threshold, with everyone whose good besides Reggie Jackson (whose not the most natural passer) and Dellavedova being above that number. But of course there’s a lot of chafe above that number as well. And the best steals guys are very often not the best guys in passing.
Though if we had a wider search that went further back in time to when advanced stats for college players aren’t available, we’d find a number of good passers are very good at steals: Kyle Lowry, Steph Curry, James Harden, Ty Lawson among them.
Blocks Per 40
1) I’m not going to talk about blocks per 40. It’s perhaps important to see as an indicator of athleticism. The guys below 0.3 blocks per 40, at least the ones on this list, do all struggle on defense. But that’s only conclusion I’d be at all willing to draw.
Stocks per 40
1) My guess here is that the further down on the list you go, the less likely it is that the player will experience meaningful improvement upon their college passing in the NBA. For instance, both Dellavedova and Thomas are still capable passers, but both are down from where they were in college.
But even this would be subject to thresholds of frame+athleticism and stocks success more than it would be toward any sort of direct linear relationship. After all, Jimmy Butler and Damian Lillard are fairly middling here, even in their best seasons. 2.3 stocks. Khris Middleton is even worse than them. Rodney Hood is again near the bottom of the list. And Brandon Knight, who is by no means a good passer for a PG but still decent enough to play, is the bottom of the list.
2) Which is to say, you’ll either have to have a mind more keen than mine or be a computer to figure out the relationship between stocks and blocks and passing if there is one. Which, on the surface, it makes sense there would be. And perhaps should be.
3) As for my analysis in the future, I’m going to stick to rebounds and perhaps steals when I look at the players from the 2016 class who are most likely to have some passing ability. Though before we get there, in the next post, I’m going to quickly cancel out all of those players whom I don’t think will have great passing success at the next level, Jamal Murray among them.
Athleticism and Passing Part 1
Athleticism and Passing Part 2
