Monthly Archives: April 2014

Visibility Bias And Julius Randle’s Defense

25 Friday Apr 2014

Posted by in NCAA

≈ 27 Comments

Tags

, ,

One of my favorite players to analytically pick apart this season has been Julius Randle.  It seemed that most people were in accordance that he was largely overrated, but then Kentucky made a tournament run that confounded everything and inspired hope for his future.  Personally, I am not feeling too badly about my earlier synopsis and feel that all of my central hypotheses still hold.  I would absolutely not consider him in the lottery, so I may as well follow up on why I refuse to budge on my anti-Randle stance.

A common trend among people who take exception to my ranking of Julius Randle as a non-lottery pick is that they are not convinced that his defense will invariably plague him throughout his NBA career.  And it makes sense that some people would be skeptical, as defense is exceptionally difficult to pinpoint either statistically or by casually watching.  And even if they acknowledge that he may have been sub-par as a freshman, that is not enough to convince most that he will necessarily be bad in his NBA prime.  So how likely is he to mitigate this wart such that it is no longer debilitating?

His easiest out is simply that his defense is an overstated issue.  But all signs point in the direction of it not being so.  His steal and block rates are exceptionally weak for a lottery PF prospect.  Also his team was not great defensively considering the size and athleticism they boasted, and they were not better with Randle on the floor (or so I have heard and instantly believed.  Anybody know where to find UK on/off splits?).  And, you can watch for yourself as Randle makes a gigantic pile of mistakes in a single game.  He often has no clue what is going on and stands there confused as his assignment waltzes right past him. That lack of mental acuity doesn’t strike me as something that is likely to go away, nor will it be mitigated with marginal improvement. On some plays he was so slow to react that it seems Tennessee could have scored twice before he figured out what was going on.  While not every game is as rife with mistakes as this one, similar errors did persist throughout the season and tournament.  When the eye test, individual stats, and team level performance all strongly suggest that he is bad at defense, the most Bayesian conclusion is that he is almost certainly bad at defense.

So now that we all agree he is bad defensively, how likely is he to elevate his instincts to a more competent level?  His main concern is that he is too slow to discern the offensive play unfolding before his eyes and sometimes fails to react until the ball is going through the net.  I believe this deficiency heavily prices into his low steal rate, as players often generate steals by anticipating what will happen in advance.  Layne Vashro made an excellent post about evaluating potential, and his statistical analysis on the growth of steal, block, and rebound rates is grim:

These traits are something a player either has or does not have. Do not expect a prospect who cannot block, steal, or board to figure out how once he enters the NBA (not that this never happens of course). Instead, these traits should be viewed as a part of the baseline a player has to work from, much as height and leaping ability are popularly understood.

If we roll with the narrative that steals correlate with defensive awareness and instincts, Randle will almost certainly always have bad instincts. After all, it doesn’t make sense that a player may drastically improve his anticipation and awareness without seeing a bump in steal rate.  If increases in steal rate are outlier events, it logically follows that big increases in anticipation and instincts should be as well.  This would doom Randle defensively– if defensive instincts are barely more improvable than height or athleticism, then he is a stone cold lock to be a liability as an NBA player.

But to leave some margin for error: let’s be open minded and say that defensive instincts are as easily improved as the most readily improved skill: shooting (at least I assume this is the case, if there is evidence to the contrary I would appreciate hearing about it). Everybody makes a big deal about Marcus Smart’s shooting ability being a damper on his draft stock, but he isn’t even the worst shooter among guards in the draft. He made 30% of his 3’s and 75% of his FT’s in college. Imagine instead that he made 25% of 3’s and 50% of FT’s– would anybody still want to draft him in the lottery? It would likely be perceived as an insurmountable wart that distracts from every good aspect Smart brings to the table. While it is impossible to equate Randle’s defensive badness to shooting percentages, he is the worst defensive player among big man prospects in the draft. Even in the most impossibly optimistic scenario that defensive instincts can improve as much as shooting, Randle’s defense should still be perceived as an exceptionally costly wart. This perception only fails to be widespread because of visibility bias. When casual fans watch a game, they notice every missed or made shot and normally none of the defensive lapses. Further, this enables season by season tracking of shooting percentages that are not available for defensive acumen.

Consider:
1) Julius Randle’s defense is worse relative to his peers than Marcus Smart’s shooting
2) Shooting is likely more readily improvable than defense, and possibly by a large margin
3) In the instance that neither player drastically improves their wart, Smart has a much rosier upside comparison among a player who shares the wart with similar strengths (Dwyane Wade) than Randle (Luis Scola, David Lee).

All of the concern for Smart’s shot should apply tenfold to Randle’s defense.

As an interesting aside, concern for Randle’s shot should also be great than the concern for Smart’s shot. Floor spacing is quickly being recognizing as valuable. And with analytics becoming increasingly widespread in the NBA, it is worth pondering whether the league is moving in a direction such that players who cannot either hit 3’s or play defense will be coveted at all in the future. Randle’s shot is perceived as a positive as he hit 70.6% of his FT’s, and he has some potential to develop a 3 point shot in spite of only making 3/18 as a freshman. While the bar is lower for acceptable big man shooting, Smart has his defense and PG skills to fall back on and does have superior shooting splits to Randle. If Randle neither steps up his defense in a big way nor becomes a reliable 3 point shooter, it’s difficult to see him ever becoming an impact player in the NBA.

As I mentioned earlier, I could envision Randle becoming David Lee level good. This may sound alright to some people, but to me it is a horrific upside scenario that does not merit 1st round consideration. I do not think David Lee is a particularly useful NBA player because he doesn’t space the floor, and his offensive and rebounding value is consequently outweighed by his poor defense. If Lee is a prospect’s best case scenario, that prospect should be worth little.

There are undeniably a number of positive traits that Randle brings to the table.  He is strong, mobile, great at rebounding, and talented at converting difficult shots in the paint.  His strength enables him to get to the line where he is solid at making his free throws, and he also has some semblance of handling and passing to work with.  Further he appears to be competitive, hard working, and coachable.  But because he spectacularly fails at the most high leverage aspect of his performance (defense), this puts a massive damper on his upside.  He also has uncertainty regarding the second highest leverage aspect of his value (3 point shooting) that further inhibits his value.  These warts are going to be often overlooked because one is not readily visible, and the other is not yet accepted as common NBA wisdom.   But they drown out all of the positive qualities he brings to the table simply because none of his good traits are nearly as valuable as defense or spacing.

Hopefully this sheds some clarity on why I remain bearish on Randle in spite of his late season improvement and solid tourney showing.  I do believe he carries his fair share of bust risk, but I do not necessarily believe he is a lock bust.  The greater concern is that when his successful outcomes still are not that appealing, as he may post good stats as an NBA player without aiding his team too much in the W column.  If he ever averages something like 18 points and 10 rebounds with an 18 PER, I’d hope that nobody trolls me over my ranking of him.  So long as he keeps missing rotations on defense, I would never regret passing on him in the draft.

Playing The Round 2 Lotto

22 Tuesday Apr 2014

Posted by in Miscellaneous, NCAA

≈ 12 Comments

Tags

, , , , , , , , ,

Round 2 is the uneventful part of the draft, where the majority of picks amount to little or nothing.  But sometimes teams uncover solidly useful players such as Chandler Parsons, Isaiah Thomas, Paul Millsap, Kyle Korver, Danny Green, Omer Asik, etc.  The goal of drafting in round 2 should be to draft a starting caliber player such as the aforementioned names.  It is more likely that you can find a fringe bench player such as Chris Duhon, but those types offer little value since they can always be acquired with the veteran’s minimum.  Bearing that in mind, I’d like to review some of the highly rated players that I’d pass on and the unheralded players who I’d target instead:

Do Not Draft: CJ Wilcox (#32 ESPN, #35 DX)

I have no problems with Wilcox’s game, as he is a solidly good college player without any glaring warts.  But the underlying logic to rate him as a fringe 1st round prospect is so backwards that the thought of him going round 1 is somewhat grating to me.  CJ Wilcox is a good shooter, as he shot 39% from 3’s and 87% from FT as a senior.  He also has average size and athleticism for an NBA SG.  If you were required to trot out a round 2 prospect for bench minutes in 2014-15 season and hope he does not submarine your 2nd unit, Wilcox would not a bad pick.  But that is the opposite of the correct goal for drafting in round 2, as teams are not required to play their 2nd rounders.  And unless he was woefully misused in his 4 years at Washington, he is drawing dead to become a starting caliber SG.  He turns 24 in December and his skills outside of shooting are largely underdeveloped.  As a senior, he was roughly the same player he was as a freshmen and is not much of a handler, passer, rebounder, or defender.  Considering his age, it is exceptionally unlikely that he develops his non-shooting skills to NBA levels of competence, and it’s not like he’s a Kyle Korver or JJ Redick level shooter.  His upside is roughly Willie Green, the poster child for replacement level SG.

Instead Draft: Isaiah Sykes (unranked ESPN, unranked DX)

Sykes is the inverse Wilcox, in that he’s good at everything except shooting.  This was roughly Chandler Parsons’s appeal: he was a good handler, passer, rebounder, and finisher and then when his shot developed better than expected he quickly became a solid starter.  Sykes is 6’5 with a 6’11 wingspan and exceptional athleticism, and he offers defensive upside that Parsons lacked.  He racked up a solid steal rate in his final 2 seasons at Central Florida, and could be effective on this end playing for an NBA coach.  Also he is a good rebounder, ball handler, and passer, and his shot is not completely hopeless as he shot 34 for 103 (33%) from 3 for his college career, improving his 3p% each season.  The downside is that he shot poorly from the FT line (54% as a senior 56% overall) and was generally an inefficient and turnover prone player in his large offensive workload.  He turns 23 in December which gives him limited time to improve his shot, but his shooting splits are not too different from Parsons who was similarly bad at FT’s and merely took a higher volume of 3’s.  If his shot can take a mini-leap and he lands with an NBA coach who gives his game a nice haircut, he has enough positive qualities to become a solidly good cog in any NBA unit.

Do Not Draft: Jahii Carson (#38 ESPN, #42 DX)

Why is this guy a prospect?  He’s explosive, but he’s also diminutive at 5’11 and not exceptionally skilled.  He’s a solid but not great shooter and passer, and he struggles to finish inside amongst the trees in college.  He also is a near lock to be woefully bad on defense given his size and the fact that he did not accrue many steals to help atone for this.  He has developed a reputation as a selfish player and he is old for a sophomore as he turns 22 in August.  There is little to like and oh so much to dislike.

Instead Draft: Kendrick Perry (unranked ESPN, unranked DX)

Perry is also diminutive, as he recently measured 5’11.75 in shoes and 169 pounds at Portsmouth.  But he atones with exceptional explosiveness and length with a 6’6 wingspan.  Even though he’s a senior, he’s actually 4 months younger than Carson.  And while he shares Jahii’s burst with similar shooting and passing abilities, he is also better at everything else.  He’s a better finisher, rebounder, defender, and superior at protecting the ball.  He completely and utterly outclasses Carson in every way conceivable, and the fact that Jahii is on draft radar and Perry isn’t speaks to the inefficiency of consensus prospect rankings.

Do Not Draft: Cleanthony Early (#24 ESPN, #27 DX)

Well, this hype train sure got out of hand in a hurry.  I rather like Early as a college player, but considering him in round 1 is a hilarious overreaction to a well timed career game vs. Kentucky in the tournament.  Early appeared incapable of ever missing a jump shot vs UK, but you can trust me on this one: he sometimes missed shots in his other games.  He has solid size and good athleticism for a SF, but really what else is there?  He’s a solid but not exceptional shooter, and his finishing will struggle to translate when he is facing NBA competition rather than undersized Missouri Valley Conference bigs.  While his athleticism enables him to average 1 block per game, he projects to be a liability on defense overall.  And most troubling is this blurb from DX:

“The only small forwards drafted with a lower assist percentage in our database spanning back to 2001 (who played over 20 minutes per game in that college season) are Lazar Hayward, Al Thornton, Shabazz Muhammad, Dahntay Jones, Deshaun Thomas, Bobby Jones, Damion James and Maurice Ager”

Yikes, that is a frighteningly worthless collection of players.  And to make matters worse, Early’s pitiful assist rate came at age 22 (he turned 23 last week), and he should have learned to sometimes pass by now.  This calls his feel for the game into question, and it’s difficult to envision him becoming useful vs. NBA competition.

Draft Instead: Javon McCrea (unranked ESPN, unranked DX)

If only because his stats are frighteningly similar to those of Paul Millsap:
USG% O-Rtg eFG% FT% AST%
Millsap 26.9 114.3 57.6 62.3 8.2
McCrea 28.3 115 56.2 66.7 16.6
ORB% DRB% STL% BLK% Height
Millsap 18.1 23.9 3.2 6.7 6’7
McCrea 13.7 20.6 2.7 7.2 6’7

These stats are from each player’s final season of college, and McCrea will only be 3 months older as of draft night.

Millsap played a tougher schedule and is clearly the superior prospect with his advantages in rebounding and steals.  But McCrea’s assist advantage is not to be scoffed at, as assist rates correlate with feel for the game.  But falling short of Millsap’s goodness is not a bad thing, as he was woefully undervalued sliding to 47th overall in the 2006 draft.  More importantly, Millsap proves that undersized mid-major bullies do not necessarily flop in the pros.  McCrea’s assist and steal rates offer hope that he has the feel and length to become something in the NBA.  It will be interesting to see how he measures out pre-draft, because he offers intrigue as a 2nd round flier based on his stat stuffing and parallels to Millsap.

Do Not Draft: Rodney Hood (#21 ESPN #23 DX)

I have been bashing Hood all season, and finally DX and ESPN are starting to catch up as they no longer have him as a top 20 pick.  Nevertheless, he should never get picked in round 1 and I do not find him to be particularly compelling in round 2 either.  His only positive tool is that he has solid height for an NBA SF, and even that is in part due to his long neck.  He has subpar strength, length, quicks, and athleticism.  And to make matters worse, he has awful instincts defensively and doesn’t seem interested in working hard on this end either.  He projects to be somewhere on the scale of worst all time defender to a clear liability.  What does he bring offensively to offset this?  He’s a good shooter and a solid passer with decent feel for the (offensive aspect of the) game.  That is all.  He doesn’t rebound and isn’t much of ball handler or finisher.  It’s difficult to envision his offense amounting to enough to make his defense worth stomaching.  When at age 21 you are just the 6th best player on a team that cannot win a single tourney game, it may be a sign you aren’t going to excel in the NBA.

(Side note: as an avid Duke watcher my player ratings this past season are: Amile > Jabari >> Cook > Dawkins > Sulaimon > Hood.  Amile is the Amir Johnson of college: he goes unnoticed due to low scoring totals but makes a ton of good plays, not many bad ones, and has an overall surprisingly positive impact.)

Draft Instead: Artem Klimenko (unranked ESPN, #35 DX)

Klimenko is a complete and utter mystery box, which makes him a great player to target in round 2.  He is 7’1 with a 7’4 wingspan and good mobility, but at age 20 he has yet to face any level of challenging competition.  According to DX he is averaging 15 points in 24 minutes on 57% inside the arc against weak Russian competition.   This means little for his NBA projection, but at least he’s dominating the dregs like he should.  Also he shoots 74% from the line, which offers hope that he has some semblance of skill.  The primary question is whether he has the instincts and intelligence to maximize his physical tools and become a useful NBA defensive player, which is where he has the most potential.  If he does, then perhaps whoever gambles on him will acquire an Omer Asik level steal.  If not, at least you didn’t waste your pick on a player who has already strongly suggested that he lacks upside.  Teams can glean more regarding how worthwhile of a gamble Klimenko is via interviews and workouts.  But as a general concept: young, toolsy internationals with fuzzy translation are good targets as the talent on the board thins.  Giannis Antetokoumpo going 15th overall last year is a good example of this.

Closing Thoughts:

The players I listed as solid round 2 targets are some ultra deep sleepers, and it’s reasonably likely that all of them amount to nothing.  The vast majority of players who ever may amount to anything are already on DX/ESPN’s radar at this stage, and the best 2nd round picks will likely be players who are already on the radar and slide too far.  For instance: Spencer Dinwiddie (#42 ESPN/#49 DX) has become a bit of a forgotten man after his ACL injury.  But he still may declare nevertheless, and he may slide to round 2 and then become a Korver/Green level role playing wing.  Jarnell Stokes (#28 ESPN, #49 DX) also may make for a valuable role player with his unique combination of speed, strength, and skill.  But these players also may go in round 1 and the players that *should* merit 1st round consideration often make for the best 2nd round picks.

Since it is difficult to anticipate precisely where everybody will land when the draft actually happens, I constrained myself to scraping the barrel for this exercise.  So take this as a demonstration of my logic for uncovering diamonds in the rough as opposed to my list of favorite round 2 sleepers, as it will ultimately be the Dinwiddie/Stokes type sliders who are the slickest steals.

Jabari Parker: Is He The Next Melo Fellow?

11 Friday Apr 2014

Posted by in NCAA

≈ 15 Comments

Tags

,

As a Duke alum, I watched nearly every Duke game this year and have a number of thoughts on Jabari Parker, yet have refrained from writing about him.  This is largely because I see an interesting blend of positives and negative and had been reticent to commit to a strong opinion either way.  Now that the season is over, I’d like to lay out some troubling trends I have noticed as well as why they may not be fatal flaws.

First I would like to initiate a narrative that has gone largely unnoticed this year: Jabari Parker was a straight up chucker as a freshman.  It’s not that people failed to notice that he took a ton of shots; after all his most common comparison is Carmelo Anthony.  The fact of the matter is that when you launch a high volume of shots with the efficiency of a player like Carmelo, people are not going to complain about your chucking ways.  But I do believe that Jabari’s inefficiency flew a bit under the radar for a number of reasons.  First, let’s see how his usage and efficiency stats compare to those of his teammates as per sports-reference.com.

Usage% O-Rtg TS% AST%
Amile Jefferson 13.4 134.9 61.8% 8.3
Tyler Thornton 7.5 133.1 64.1% 18
Andre Dawkins 25.6 126 63.4% 5.6
Rodney Hood 23.8 122.5 59.0% 13.2
Quinn Cook 20.4 122.1 57.1% 27.7
Rasheed Sulaimon 20.3 120.7 55.2% 17
Jabari Parker 32.7 115 55.8% 8.6

On one hand, a 115 offensive rating on a 32.7% usage is impressive vs. the caliber of opponent Duke played.  On the other hand, it is not as if he was surrounded by garbage and Duke needed him to take every shot he could possibly get off.  Duke’s other top rotation players all had a significant advantage in offensive rating, and it would be nice if Jabari had a better assist rate considering all of the shooting surrounding him.  Other than Jefferson who did not attempt a 3, his other most common teammates all shot 37%+ from 3 with 4 of them hitting 41%+.

It’s easy to see why Parker took so many shots: he is the perceived best player on the team and carries with him a strong alpha male mentality.  It is clear he believes that it is his role to carry the scoring load, so it stands to reason that he should be taking the lion’s share of shots for his team.  But he took this to an extreme level.  Nobody ever accused Carmelo Anthony of being unselfish, yet he took on a less gargantuan role as a freshman for Syracuse (note that I am now taking usage/o-rtg from statsheet.com, which is why Jabari’s figures are different from the prior table):

Player Usage O-Rtg Ast% Opp D-Rtg
Carmelo 27.8 113.6 11.8 96.5
Jabari 31.7 111.5 8.6 100.7

Carmelo played on a much more defensive oriented team where taking a high volume of medium efficiency shots carries more value.  For reference, Duke’s team schedule adjusted O-Rtg was 123.5 vs. Syracuse’s 113.5, but their defense was much worse (102.3 vs. 91.3).  Further, Carmelo was surrounded by significantly less 3 point shooting as he only had two regular teammates who made 3’s, shooting collectively 35% behind the arc.  Yet he nevertheless posted a comfortably lower usage rate and higher assist rate.  Once you consider context, Jabari almost makes Carmelo Anthony look like Steve Nash.

The other disconcerting trend is that Jabari Parker was significantly more efficient against bad defenses.  While he is a good shooter for a freshman and has solid perimeter skills, he also operated quite a bit in the low post as he often played center for Duke.  His best performance of the season came against Boston College’s swiss cheese defense, as they start bigs listed at 6’8 219 and 6’7 207.  They have the #298 defense and are 238th in opponent 2p%.  Naturally Jabari bullied them to kingdom come, as he finished with 29 points, 16 rebounds, and 12/17 FG in a performance that included 6 dunks.  It was an entertaining show to be sure, but at the same time it was not against competition that remotely simulates NBA defense.  If you break up his performance to teams that are top 100 in opposing 2p% and played a top 150 schedule (essentially weeding out Vermont who was impenetrable by pitiful America East offenses), here are how his per 40 minutes stats look:

opponent Pts FGA eFG 2PA 2p% FTA AST TOV
top 100 22.5 19.4 42.4% 14.8 41.1% 7.6 1.1 3.2
not top 100 27.1 17.8 59.4% 14.5 58.7% 8.3 1.9 2.9

Note that the sample includes 500 minutes vs. good defenses and 574 vs. bad ones.  Granted, we are taking a small sample and breaking it up into two smaller samples, and one of his best performances barely misses the cutoff as UNC only has the #102 2p% defense in the country.  But even if you move his two UNC games in the tough sample, he still only musters a 44.2% eFG as compared to 59.6% in the weak sample.  And the fact that the performance drop off is largely driven by a drop off 2 point efficiency makes it less likely to be largely due to fluke.

The bottom line is that Jabari bullied bad teams and he bullied them hard.  This inflates his stats in a way that is not necessarily predictive of NBA performance.  He will still be an issue for smaller matchups in the pros, but they will become less common and there will almost always be a bigger help defender on the floor.  He still needs to develop his decision making and perimeter skills significantly to become an efficient scorer against NBA defenses, because his bullying did not work so well against tougher NCAA opposition.

For reference, here are Carmelo Anthony’s per 40 splits given the same criteria:

opponent Pts FGA eFG 2PA 2p% FTA AST TOV
top 100 24 19.5 48.0% 14.5 47.9% 7.5 2.1 2.8
not top 100 25.3 18.7 53.5% 13.1 53.2% 7.5 3 1.7

Note that Carmelo also faced a higher % of good defenses, with 842 minutes in the tough sample vs. 432 in the weak sample.  Naturally Melo’s performance fell off vs. serious defenses, but he padded his stats less vs. weaker teams and did not have a massive eFG% or 2p% chasm between the two splits.  Also while his assists and turnovers both suffered against tougher teams, his ratio in the tough sample is still much better than that of Jabari which implies that he may have a superior feel for the game.

Again, take these splits with a grain of salt due to sample size issues, but it aligns with my perception. Jabari relied moreso on rim scoring against undersized competition whereas Melo’s midrange dominance translates to higher levels of competition with ease.

While they appear to be similar prospects at a glance, freshman Melo is comfortably superior to freshman Jabari.  There is the possibility that Jabari merely needs time to adjust to being stoppable at the rim and adapt his game accordingly, but I would have felt better about this hypothesis if he had displayed some level of improvement down the stretch.  Instead he shot 6/16 on 2’s vs Clemson, 7/20 vs Virginia, and 4/11 vs. Mercer as Duke was upset in round 1.

Mercer is hardly a challenge in the paint, as they posted the 112th best 2p% defense playing the 197th toughest offensive schedule.  Yet they unwisely insisted on playing zone defense vs. Duke to stop Jabari, and it sort of worked.  I say sort of because Duke shot 15/37 on 3’s and rebounded 16/40 of their own misses, and they should be unbeatable by Atlantic Sun competition when this happens.  But Duke also punted defense this year in favor of a super offense, and allowed a 122 O-Rtg to Mercer.  So when Jabari shot 4/11 on 2’s with 0 assists and 4 turnovers (the rest of the team attempted just 14 2’s and committed 8 turnovers), I think it’s fair to pin a significant amount of blame on him for the loss.  Duke was given an all-you-can-eat buffet of quality 3 point looks for their 40% shooters, and Jabari diluted this by insisting on [not] getting his inside vs. the zone instead of trusting shooters to make shots.  I understand that it’s part of the alpha dog mentality, but it would have been nice if he had displayed a bit more macro level perception instead of going full cancer and playing his team out of the tourney against a vastly inferior foe.

So why am I not screaming at the top of my lungs that Jabari will be a bust?  There are a myriad of slippery aspects to any Jabari analysis, and I am not certain that these issues are indicative of any fundamental flaw that will invariably undermine him throughout his career.  He still has an intriguing blend of size and skills, and he will be forced to improve his decision making when he learns that trying to repeatedly dunk on players such as Roy Hibbert is difficult.  He’s such a fiery competitor that it’s not difficult to envision him finding a way to make his offense work in the pros, especially if he lands with a good coach.  And while his defense was not great, he did post excellent rebound numbers and solid steals and blocks.  So I am reticent to sour too heavily on Jabari, as there is much to like.  But I also think he has a wider range of outcomes than common narratives dictate, since he does need to overhaul his offensive approach to succeed as a pro.  It’s possible that he doesn’t peak any higher than a Jeff Green level combo forward who is a solid scorer but does not bring enough else to the table to be particularly valuable.

If he does elect to stay at Duke, I believe his sophomore season will be illuminating for his NBA future, as he will be forced to share the paint with possible 2015 #1 pick Jahlil Okafor.  He should spend more time on the perimeter, and he will no longer have the excuse of limited experience vs. defense that can physically match up with him.

Throughout the season I had Jabari in a close battle for the #2 slot on my big board with Dante Exum, but after a disappointing postseason I rate Exum comfortably ahead of him.  Now the question becomes whether I prefer Parker over a prospect such as Marcus Smart, and I am leaning in the direction of Smart for now.  There is enough to like such that Jabari will eventually settle into my 3-5 range, but his selfish ways and issues vs. stingy defenses remove some of his luster as a tanking prize.

Kentucky’s Luck Runs Out vs. UConn

08 Tuesday Apr 2014

Posted by in NCAA

≈ Leave a comment

Tags

, , , , , , ,

After eking out four close wins against quality competition, UK finally fell into a first half deficit that they couldn’t entirely overcome, as they lost to UConn 60-54.  It was an interesting tourney run for the Wildcats, and I would now like to comb through the details to point out what changed from the regular season to the postseason that propelled them this far.  They started playing well in the SEC tournament, so I will split each player’s stats per 40 minutes into the 9 SEC/NCAA tourney games vs their regular season performance.  Note that they tended to both run into better defenses and play at a slower pace down the stretch, so naturally everybody will have rosier regular season per 40 minute stats.

Willie Cauley-Stein

Season PTS eFG% REB AST STL BLK TOV
Regular 12.0 60.1% 10.2 1.2 1.8 4.6 1.6
Post 8.0 55.0% 10.3 0.7 3.0 6.0 0.3

Cauley-Stein’s postseason sample is too small to be particularly meaningful, but he did rack up a boatload of steals and blocks and only turned it over once in 120 minutes of play. More notable is how the UK defense struggled without him.  They did not perform at a high level vs. any of Louisville, Michigan, Wisconsin, or UConn, and their adjusted defensive rating dropped to 41st in the country as per kenpom.com.  Last year’s UK team bottomed out after the Noel injury and finished with the 129th defense, but among Calipari’s successful teams this is his worst defense that he has assembled.  From 2006-2012 all of his defenses finished top 15, and the 2010-11 UK team was the only iteration that was not top 9.  Cal’s last defense to perform this poorly was the 2004-2005 Memphis team that missed the tournament and finished with the 43rd defense.

For all of Kentucky’s size and athleticism, they are not particularly effective at preventing opponents from scoring. This is largely why I am not exceptionally high on this year’s crop of Kentucky prospects.  Their success down the stretch did not stem from suddenly pulling things together and playing great defense: it was almost entirely derived from decreased turnover rates and timely shot making.

Julius Randle

Season PTS eFG% REB AST STL BLK TOV
Regular 20.1 52.2% 13.6 1.8 0.6 1.1 3.6
Post 17.1 43.6% 13.1 2.0 0.8 0.8 2.2

Randle’s big progression down the stretch was cutting down on his turnover rate.  I have not charted statistics for this, but my suspicion is that this largely stems from fewer post up attempts, as he often coughed the ball up in traffic.  Regardless of the precise reason, this was a significant development for UK as their offense is difficult to stop when they aren’t turning it over.  They rebounded 41.9% of their own misses, so turnovers are especially costly in comparison to missed shots.  This is a positive sign for Randle, as he needs to find a way to score without being a turnover machine to succeed as a pro.

On the other hand, Randle’s warts persist and I still struggle to get excited about the prospect of drafting him in the lottery.  He drew 3 favorable defensive matchups in the tournament, as Kansas State, Michigan, and Wisconsin are all undersized, lack shot blocking, and are vulnerable inside.  Naturally Randle posted his three highest scoring totals of the tournament vs. these three defenses.  He had his lowest scoring output vs. the long and athletic UConn defense, finishing with just 10 points on 3/7 FG.  And his lower turnover rate is somewhat diminished by the fact that it came in tandem with poor shooting from the field.  He is still prone to defensive lapses, and there are still questions about his ability to translate offensively.  While I appreciate his competitiveness and growth throughout the season, he did not make a convincing case that he will be able to score effectively enough vs. NBA defenses to justify a lotto pick.  He has enough strengths for a freshman to be worth a flier in the 15-20 range, but I have a hard time envisioning him becoming a true impact player.

James Young

Season PTS eFG% REB AST STL BLK TOV
Regular 17.7 49.3% 5.1 2.3 1.0 0.2 2.4
Post 17.6 51.6% 6.0 1.4 0.6 0.3 2.1

James Young had a solid two games in the Final Four, and his performance reinforced my perception that he is the Kentucky player who should most seamlessly translate to the NBA.  He doesn’t bully smaller competition to the extent his teammates do, and he is able to get his shots off vs. defenses of UConn’s caliber with his combination of size, length, a quick trigger, and a knack for hitting contested shots.  His increase in eFG% down the stretch was largely driven by his 3’s finally starting to fall, as he shot 33.7% from 3 in the regular season vs. 41.7% in tournament play.  His 2p% only fell off from 47.8% to 44.8% in spite of higher volume (8.0 2PA/40 vs 6.2) and tougher opposing defenses, and he finally started to make his free throws hitting 83.3% after a 67.4% regular season.

The shooting upticks are a welcome sign for Young.  His regular season shooting stats were surprisingly mediocre considering how nice his form looks, and his NBA success hinges largely on him becoming an effective shooter.  He is still only going to be 18 on draft night, and between his age, form, and ability to get shots off vs. good defenses, he has quite a bit of room to grow offensively and could become a good offensive player in the NBA.

That said, there are reasons to temper expectations.  Like his teammates, his defense is not particularly good.  He has the tools to be good on this end, but seems to lack acumen.  And in spite of his ability to translate and room to grow offensively, he did not have a particularly effective season for a one way scorer.  If his shot doesn’t develop well, he might be an Austin Rivers level flop.  Even if his shot does develop well, he may not become significantly better than Nick Young.  He is likely worth a late 1st round pick for the scoring upside, but he has plenty of bust equity as well.

Aaron Harrison

Season PTS eFG% REB AST STL BLK TOV
Regular 17.1 47.5% 4.1 2.6 1.4 0.4 2.2
Post 16.0 57.9% 2.1 1.2 1.0 0.2 1.3
Season 2PA 2P% 3PA 3p% FTA FT%
Regular 7.9 48.4% 5.1 30.6% 5.9 81.4%
Post 5.5 42.2% 6.1 48.0% 3.8 67.7%

I’m including Aaron Harrison’s shooting splits to show that there was exactly one thing he did well this postseason: make 3’s.  His increase in eFG% and decrease in assists and turnovers is rooted in the fact that he attacked less and was used as a spot up shooter more, hitting 48% of his 3’s including a plethora of clutch shots.  He earned casual fan acclaim for his timely shot making, but overall I was not impressed by his tournament showing.

I was somewhat hopeful for Harrison earlier in the year because I felt his 3P% was suffering from bad luck, and that his overall game might look quite good once his 3’s started falling.  So while it is nice for his 3’s to violently progress to the mean, it isn’t encouraging when his production otherwise fell off a cliff.  And I do not believe that his clutch shooting is indicative of any innate ability to score against tight defenses – he simply spotted up for 3’s and happened to make them.  If anything I have cooled on him after his tourney play and do not feel that he is worth a 1st round selection.  He’s a 2nd round pick in my estimation.

Andrew Harrison

Season PTS eFG% REB AST STL BLK TOV
Regular 14.1 42.0% 3.9 4.6 0.6 0.3 3.2
Post 12.4 40.6% 4.3 6.1 0.8 0.4 4.1

With his brother taking on more of a spot up role, Andrew Harrison took on a greater portion of the PG duties and it shows with significant upticks in both assists and turnovers.  I strongly dislike almost everything about his game, as his sole strength appears to be bulldozing to the rim and drawing FT’s.  He is a horrible decision maker, as evidenced by his turnover rate and eFG.  There is plenty of room for him to improve as a college player if he elects to stay, but I simply don’t see his feel for the game ever becoming good enough for him to be a useful pro.  Somebody will try to salvage him with a 2nd round flier, but I wouldn’t bother with him.

Alex Poythress

Season PTS eFG% REB AST STL BLK TOV
Regular 12.8 48.6% 10.5 1.1 0.6 1.7 2.1
Post 12.5 69.0% 7.4 0.3 0.8 1.1 2.1

Poythress was UK’s unsung hero of the tournament, as he hardly missed down the stretch.  He shot 14/15 inside the arc in UK’s final 5 games, providing significant unacknowledged value.  It does not appear he will be entering this year’s draft, as his stock has eroded with his regression this year.  I am not particularly high on him but he may become worth drafting before all is said and done.

Dakari Johnson

Season PTS eFG% REB AST STL BLK TOV
Regular 16.0 57.6% 11.7 0.8 0.5 1.9 1.7
Post 12.2 55.6% 10.2 0.9 0.9 1.3 2.0

UK’s offensive rebound expert finished the season with a staggering 17.0% ORB rate.  This exceeds his 15.0% DRB rate.  His tournament was up and down, as he had some efficient games and other quiet ones.  He is draftable but I suspect he will stay in school for his sophomore season.

Marcus Lee

It’s not even worth posting his stat lines with just 156 minutes on the season.  He had a great 24 minutes vs. the soft interior defenses of Wisconsin and Michigan, and then posted goose eggs in 6 minutes vs. UConn’s significantly tougher defense.  There’s not enough information to have much of an opinion on him at this point.

Who Is The Andre Drummond Of This Draft Class?

04 Friday Apr 2014

Posted by in NCAA

≈ 3 Comments

Tags

, , ,

Andre Drummond is quickly becoming a classic example of a player whose RSCI rating seems to be more indicative of his pro value than his freshman performance.  Whenever a highly touted freshman underwhelms and his draft stock drops, it’s fair to posit whether that player might experience a career arc similar to that of Drummond.  Drummond is an outlier in that his circumstances that caused him to become so undervalued will likely not be replicated, but it is worth trying to learn from his situation to look for signs in 5 star recruits before souring on them too quickly.  So first I’ll walk through the qualities of Drummond that caused him to slide and how that changed in the NBA.

1) Unfavorable Circumstances
Drummond joined his UConn team late, and then coach Jim Calhoun was suspended for the first 3 games of the Big East season and later had to take a leave of absence for health reasons.  It is difficult to estimate the level of impact these circumstances had on Drummond’s performance, but it is easy to see how they may have been harmful.  Perhaps he would have gotten off to a hotter start if he had more time to prepare with his team in the offseason, perhaps he would have had a stronger finish if Calhoun was healthy and present to offer feedback on Drummond’s non-conference performance.

2) Trimming The Fat
One of Drummond’s greatest warts was his 50.9% TS, which is appallingly low for a man with his size and athleticism.  Yet as an NBA rookie he was able to skyrocket his TS to 57.8% without seeing a drastic drop off in his usage rate (21.7 > 17.2).  How is such a thing possible in a single season after the huge increase of level of competition?  Drummond simply stopped trying to score away from the hoop.  As a college freshman he shot 130/185 at the rim and 27/107 away from the rim.  As an NBA rookie he shot 204/318 at the rim and 4/24 away from the hoop.  This simple tweak had a huge impact as it increased Drummond’s eFG from 53.8% to 61.0%.

3) College Defense
UConn entered the season ranked #4 in the polls and 6th by kenpom.com, then losing in the first round of the tournament as an 9 seed.  They finished as just the #41 kenpom team with the #65 defense.  So it is easy to blame some of the disappointment on Drummond and assume that he did not make the expected defensive impact.  But at a closer look, he seemed to do quite well.  He had more than twice as many blocks as any of his teammates, and UConn finished the season with the lowest opponent FG% at the rim in the country.  They finished with the 3rd lowest 2p%, and the 5th lowest FT:FGA rate, all of which are excellent and largely attributable to Drummond who finished with more blocks than fouls.  UConn did have plenty of other size and athleticism, but their weaknesses came from poor perimeter defense, as they finished with the 319th defensive TOV% and opponents hit 34.4% of 3’s on above average volume.  Of course sacrificing in these areas makes it much easier to dominate paint defense, but the fact remains that the team succeeded in the areas where Drummond was expected to make an impact barring one:

4) Defensive Rebounding
Defensive rebounding is much more difficult to predict than offensive rebounding since it is largely context dependent, but Drummond’s turnaround is astonishing.  UConn had the 276th best defensive rebound% in the country and Drummond corralled just 15.5% of d-rebs, and then went on to rip down 27.2% of defensive rebounds as a pro.  Given his 14.2% o-reb rate in college, we shouldn’t be surprised that he upticked defensively as a pro but I am not sure that there was any signal that he would start pulling them down at nearly double the rate.  Perhaps this can be attributed to good scouting by the Pistons, perhaps it can be attributed to bad luck or bad assistant coaching at UConn.  But it is nevertheless an outlier event that gave Drummond a nice value spike as a pro.

5) Passion Questions
After Drummond’s somewhat underwhelming freshman year, scouts started to question his passion for basketball.  His explanation was that he simply was not the type to go out and beat his chest, but he nevertheless loved the game of basketball.  Whether it was a poor inference from observers or NBA money ignited his passion, it seems to not be a problem as a pro.

Now let’s look at some of the top freshman and see whether any similar circumstances may apply.

Andrew Harrison: Do any of these conditions apply to Harrison?  No, they do not – he is merely horrendous at the game of basketball.  We can safely move along.

Jabari Parker: He was likely in the best scenario of all freshman as he was able to play the 4/5 for Duke.  This was healthy for him in almost every regard since he was constantly surrounded by ball handlers and shooters and finished with more blocks and rebounds that he would have surrounded with more size.  That said he exceeded expectations so it is difficult to gripe.

Julius Randle: His coach normally sets up players for the pros quite well, his defense was horrendous, his defensive rebounding was top notch, and his passion seems to be present.  The one area where he may gain is from (literally) trimming his fat and slimming down physically.  Also he has trimmed a bit of turnover fat down the stretch which is encouraging.

Zach LaVine: None of these conditions apply to him, although his circumstances were unfavorable in a different way since he was buried behind superior players and seemed to have a poor relationship with his coach.  Perhaps he has more to flaunt than he was permitted to show at UCLA, so he may exceed expectations in a different way than Drummond.

Andrew Wiggins: People want to blame Wiggins’ lack of dominance on Bill Self, which is silly.  What is most important for Wiggins is that Kansas played an up tempo style and capitalized on his sole offensive strength: transition scoring.  The only unfavorable aspect for Wiggins is that he was surrounded by mediocre guards and spacing, which was certainly sub-optimal.  But I do not believe that this had a high leverage impact on his performance given his lack of offensive skills in the half-court.  And it is worth noting that Self has made past players such as Ben McLemore, Thomas Robinson, and Cole Aldrich look like college studs and consequently over-inflated their draft stock.

Kansas had their worst defensive season under Bill Self’s tenure, and that seems to be everybody’s fault but Wiggins.  Embiid was inexperienced and played a lower minute total than prior rim protectors Jeff Withey and Cole Aldrich.  The defense suffered with Embiid out and Wiggins appeared to be the only good defensive guard/wing on the roster.  I believe his NBA defensive projection is often overstated but he did perform well on this end in college.  And since all of his competent teammates were bigs, he was rarely used as a small 4 and may be slightly underrated by his d-reb% among other stats.

Wiggins does have questions with respect to his passion.  According to DraftExpress, from age 17 to 18 he grew an inch without gaining a single pound.  This would not be a big deal if he had instead focused on developing his skills, but they too are less developed than scouts had hoped they would be by now.  In tandem these are red flags that call his work ethic into question.  But the flip side is that perhaps he will put passion questions to rest by significantly improving his work ethic given the allure of NBA money.  This is the area in which he has the most potential to mirror Drummond.

Overall I do not believe that Wiggins has a boatload of Drummond equity, but he isn’t completely bereft of it either.

Aaron Gordon/Rondae Hollis-Jefferson: I am grouping these two together because all conditions apply identically to them.  These two players likely have the most fat to trim offensively because they threw up so many bricks from midrange.  Gordon shot 129/177 at the rim and 44/160 on non-rim 2’s.  RHJ shot 83/113 at the rim and 36/124 on non-rim 2’s.  I am fond of Sean Miller and think he is one of the best college coaches in the country, but he has an curious willingness to permit his players to fire away from midrange.  Their top 6 rotation players all took at least 40% of their FGA from midrange, well above the NCAA average of 29.3%.  I have questioned Gordon’s BBIQ for his shot selection, but at a closer look it may simply be a byproduct of coaching.  Granted, this doesn’t entirely parallel to Drummond as it is much easier to operate strictly around the rim as a center than it is as a forward, but both players can see nice efficiency upticks by passing up long 2’s more frequently.

Gordon and Hollis-Jefferson also deserve a ton of credit for Arizona’s leap defensively.  After swapping Mark Lyons, Solomon Hill, and Kevin Parrom for them and TJ McConnell, Arizona spiked from the #47 kenpom defense to #2.  And while the departed players were better offensively than defensively, Arizona only dropped from the #10 offense to #20.  It helps that players such as Nick Johnson and Kaleb Tarczewski had an additional year of seasoning, but Gordon and Hollis-Jefferson played large roles in Arizona having one of the more dominant defenses in recent memory.  They simply did not allow easy shots, boasting the best defensive eFG% in the country.

I would rate Gordon as superior to Wiggins in terms of NBA defensive potential, as he anchored a truly dominant defense.  Give Sean Miller credit for maximizing his talent, but this was by far his best defensive team ever.  Gordon appears to play defense with more intensity than Wiggins does, and while he may not be as fast or quick, he is much stronger.  It is simply much easier to feel great about the player(s) who led a coach’s all-time best defense over one who led a coach’s all-time worst defense.

Overall I’d say Gordon (and on a slightly smaller scale, Hollis-Jefferson) clearly has the most Drummond equity of any freshman in the class.  It isn’t a perfect parallel as shooting is more important for wings and defense is more important for centers.  Even if he mirrors Drummond’s arc, the impact will be lower leverage.  Again, this goes to show that Drummond is a unique case, so optimism for 5-star freshmen making huge rookie rebounds should always be tempered.  But considering both his strong finish to the season and his potential for further upticks, I am quickly reversing my stance on Gordon and once again believe he merits a top 10 pick.