Category Archives: The Predictive Power of the Draft Combine

Kevin Hetrick’s series detailing the predictive power of the draft combine. AKA Kevin’s Summer Project

Kevin’s Summer Project, Part 14: Hey, new data!!

The 2013 NBA combine is behind us.  After thirteen parts attempting to quantify what it all means, now it is time to declare winners & losers.  Let’s start with the crew deserving a bump in their draft status.

Winners

Shane Larkin – If his NCAA-leading points produced through pick & rolls (scores plus assists) didn’t impress you, or his 24 points per game during Miami’s ACC Tourney run, then how about his joining this crew: Derrick Rose, Russ Westbrook, Mike Conley Jr, John Wall, Nate Robinson, and Jerryd Bayless?  Those were the drafted underclassmen point guards with sprint speed below 3.15 seconds and no-step vert of 30” or more.  Ignore concerns about his tiny stature; this Hurricane is primed to do big things.

Otto Porter – Tall small forwards rate as the NBA’s most productive two-way players.  Who stood  tallest of the 2013 small forward class?  Of course, Otto Porter, also the draft’s most productive 19-year old.  I see a strong NBA career, despite middling athleticism tests…which aren’t consistently reflective of NBA success or failure for a small forward, anyways (actually, most of those correlations were negative).

Phil Pressey – He surpassed 3.2 seconds in the sprint and 11 seconds in the agility drill as a draft-worthy upperclassmen point guard.  That’s been a can’t-miss combination over the last thirteen years and a great source of late value.  Peyton Siva also bested these thresholds, but currently ESPN and Draftexpress include him outside the likely draftees*; at nearly 23, he has never been a particularly effective collegiate offensive player.  This serves as opportunity to say, don’t overrate players that otherwise may not warrant drafting, just because of favorable combine results (obviously an Olympic track athlete isn’t likely to find NBA success).

Nerlens Noel – Did I bump my head?  I am giving Noel the benefit of the doubt here.  The small group of very long, very speedy prodigy centers is fairly glamourous.  Nerlens’ 110” reach meets one criteria, and I have faith that his sprint speed bests 3.3 seconds.

Victor Oladipo – The complete list of underclassmen (Oladipo is still 21 as of February 1st, 2014) shooting guards with first-round talent and a 6’ – 9” wingspan, 31” no-step vert and 35” max vert in the last fourteen drafts is: James Harden, Gilbert Arenas, Jason Richardson, Joe Johnson, Eric Gordon, Ronnie Brewer and Jeremy Lamb.  I expect the young Hoosier will fit nicely into this list.  Ben McLemore also leapt over 31″ from a stand-still, exhibiting a trait often portending success for first-round, underclass shooting-guard talent.  Allen Crabbe barely missed the combination of qualifiers that Oladipo met.

Cody Zeller - His sprint speed ties the third best of the centers evaluated by this study.   His no-step vert rose the highest of the entire 2013 Combine.  Given the precedent of these traits predicting success in underclassmen centers, this helps solidify Zeller in the top-ten.  Norvel Pelle flashed an exciting blend of speed and length…who is he?  I don’t know, but as a late second round project, he may be worth a look.

Losers

Rudy Gobert – Here’s the shocker; my big, contrarian view.  Also, this may not be fair, because this project focused solely on NCAA players.  But heading into May 16th, reports of his wingspan were 7’ – 9”…now it’s a confirmed 7’ – 8.25”.  Why would this bump his draft stock?  As an anecdotal case, the players in the draftexpress.com database with wingspan greater than 7’ – 8” include: Mamadou N’Diaye, John Riek, Alexis Ajinca, Saer Sene, Michael Olowokandi, Chris Marcus, and Boban Majanovic.  Is this a list that inspires confidence?  Detailed here and here, as a general rule, increased size did not correlate to improved offensive or defensive success for centers; instead, elite athleticism generally corresponded to raised performance.

Follow me on this gross simplification, as I attempt to illustrate a point.  Of the players investigated in this study:

  • Gobert’s reach of 9’ – 7” exceeds the 70th percentile of centers by 3.5”.
  • His max vert of 29” falls below the 70th percentile by 4”.
  • OK.   Ignore the second bullet point.  Based on his sprint speed of 19.7 ft / sec, compared to 70th percentile speed of 21.2 ft / sec, it takes 1/5 of a second for the faster player to gain 3.5″.

While I’m not sure that solved anything, athleticism provided much more reliable prediction of NBA success for Centers.  Gobert’s sprint speed was third-worst of the 2013 combine, his agility second from the bottom, with max-vert tied for last.  He failed at the more important stuff.  I don’t think the events of May 16 and 17 did anything to positively differentiate from May 15th .

Shabazz Muhammad – At the one position where size rated as consistently beneficial, the UCLA freshman comes up short.  Turning 21 at the start of next season, combined with non-elite NCAA production, he is probably a reach in the top ten.

Late draft shooting guards – Snagging a “two” in the final quarter of the draft is a longstanding popular decision amongst NBA front offices; following that pattern, a recent draftexpress.com mock shows five coming off the board between picks 46 and 60.  Invariably, these picks have not worked out…potentially bad news for Ricky Ledo, BJ Young, Vander Blue, Brandon Paul, or Michael Snaer.  Rolling the dice on a fast point guard is historically smarter.

Myck Kabongo – In addition to the facts that he was not particularly productive at Texas, and that he technically qualifies as an upperclassmen (22 years old by February 1st), his athleticism results were gross.  Not a good combination for a prospective point guard.

Upperclassmen Centers – Over the last 13 drafts, this has typically been the domain of marginally performing players.  The exceptions are Joakim Noah, Roy Hibbert, Emeka Okafor, and Brendan Haywood.  Accounting solely for defense, Jason Collins can be added to the list.  Disappointing speed and leaping results from Kelly Olynyk and Jeff Withey do not help their case towards breaking into this tiny group of  successes.  Gorgui Dieng skipped the athleticism testing.

Neutral

Any sub six-foot point guard – Size showed zero inclination towards predicting offensive success for point guards, so if you like Trey Burke, Shane Larkin, Phil Pressey, Pierre Jackson or Isaiah Cannon, don’t be scared-off that they’re shorter than you.

Undersized big men - Richard Howell, Andre Roberson, DeShaun Thomas; if you liked these guys before the combine…carry on.

Anyone not listed here – Basically regardless of how amazing or disappointing the thing they did was…you shouldn’t let it dramatically change your perception of their draft stock.

*Casper Ware did not possess the necessary skill-set to get him drafted, hence he would not qualify in a re-do of this study in five years.  As a small, fast point guard though, I featured him with a picture in the second part of this series.  Due to that, I will always feel a small interest in his career.  This year, in his rookie professional season in Europe, he sits as the second leading scorer (20.4 ppg) and fourth-best assist person in Italian Lega2.  Despite struggling around the basket, he posted respectable true shooting of 55% to go with his bulk scoring.  His team currently battles in their playoffs, clinching a first round series yesterday, behind Ware’s 20 points on 59% true shooting.  Good luck in the semifinals, Casper.  Come on, NBA cellar-dwellars and higher-level Euro teams…next year, we can do better than the Italian second-division for Mr. Ware!!

Kevin’s Summer Project, Part 13: Bringing it all Home

As I continue wading through a summary of my findings, it seems again worth clarifying what this study is good for.  As a general rule, it is not intended as some code-cracking draft algorithm.  Speedy point guards frequently thriving doesn’t mean to ignore Damian Lillard, who posted a slow sprint speed.  The value it can provide is in separating a group of closely-spaced prospects; if there are five guys you like similarly, pick the one with the commonly successful athletic traits.  It also provides some insights into unearthing late-draft value, or conversely, avoiding rarely-successful player types with a second-round flier.   Finally, and overwhelming, the primary outcome is to not overvalue any of the pre-draft measurements.  But more on that later.

Despite wingspan ranking 18th, and height 26th of 51 centers, Jason Collins posted the fifth best defensive season of every year investigated by this project.  I wonder if we will ever hear from him again.

Despite wingspan ranking 18th, and height 26th of 51 centers, Jason Collins posted the fifth best defensive season of the fifteen-hundred investigated by this project. I wonder if we will ever hear from him again.

Last week, we recalled that point guard speed and shooting guard explosiveness proved most-likely to portend offensive success.  Looking at the small forwards, this series noted the combination of their offensive and defensive success, with a prevalence for the largest of the group to shine brightest.  The union of size, athleticism & skill inherent in the job-posting for “small forward” provides unmatched two-way success from the 2000 – 2010 drafts.  In recent years, that continues thanks to Paul George, Kawhi Leonard, Chandler Parsons, and Jimmy Butler.  Oddly, this was the one position where size proved most-meaningful, while the athleticism results largely trended towards negative correlations.  Both of these were contrary to all other positions.   This seems reasonable.  Let’s say there are three basic ingredients to a player’s success: skill, size, and athleticism.  As the smallest players on the court, the guards that supplement their outstanding skill-level with elite athleticism prove more prone toward success.  The front-court players that combine their prodigous size with elite athleticsm tend to separate from the pack.   The “big” small forwards, players skilled-enough to be labeled as wings, yet approaching the enormity of their larger foes, check two of the three boxes without even considering their level of athleticism.  Elite speed or hops are just a cherry-on-top.    As the biggest players suitably skilled to be back-court players, the group possesses a formidably unique blend of attributes.   Because of this, I like Otto Porter’s chances of being can’t-miss near the top of the draft.  More of this type of talk following the combine though.

Moving into the front-court, power forwards generally provided non-definitive results.  On offense, wingspan and reach offered some high-points, but nothing remarkable.  For center offense, speed proved most reliable, with elite combinations of reach and quickness providing star-potential.  The defensive results for the big men provide the most shocking outcomes of this study.  Size doesn’t matter.  Look at the table for power forward height and all size correlations for centers.  Apparently effort and smarts, as well as strength & athleticism, overshadow height, wingspan, and reach as the broadest factors influencing bucket-stopping big man dominance.  A 6′ – 10″ power forward is tall.  You know who else is tall?  A 6′ – 8″ guy.  Looking at the spectrum of players drafted over an eleven-year timeframe, within the context of RAPM, incremental increases in length across the front-court positions do not notably impact defensive performance.  This contrasts known draft-day adages and also the selection patterns of NBA teams.   Ironically of the five positions, height correlated least-well with draft position for small forwards, and of underclassmen, strongest for power forwards and centers.  This is the opposite of how these players actually produced in the NBA.

For point guards on defense, speed again ruled supreme, however size did prove beneficial.  All forty-five upperclassmen correlations between Points Stopped and size measurements were positive.  Similarly, shooting guard defensive results reflected a prevalance for longer players to succeed more-readily.  Explosiveness, again evidenced through their leaping measurements, provided even better correlations though.   While the height & wingspan of these players warrants some consideration, of course Tony Allen taped-in with average 6′ – 3.5″ barefoot height and 6′ – 9″ wingspan; he’s pretty good at defense.  Certainly on your draft board (everyone has a one, right?), a small boost is warranted for length in a guard prospect, but given the relative importance of guards on offense compared to defense, don’t get carried away.

To wrap these ideas together:

  • If you only think of the pinnacle (Lebron, Durant, Carmelo), this is obvious, but top-to-bottom, tall small forwards are the NBA’s best players, providing strong two-way impact.
  • Length is important for guards on defense, but did not prove beneficial offensively.  If you otherwise can’t differentiate two comparable players, certainly give the longer guard the benefit-of-the-doubt.
  • For big men, increased length lightly corresponded with improving power forward offense, with reach providing the best correlations.  On defense, and for centers  at both ends of the court, size measurements frequently, and surprisingly, calculated as very-low and often negative correlation with performance.  Over two-thirds of the correlations between height and Points Stopped  ended negative for the front-court players.   There are so many factors influencing effective defense, that the impact of an extra inch here, two inches there, doesn’t prove as a primary means of separation between players.  When you start saying that “Player X doesn’t have size to play power forward; he’s a tweener”, or “Player Y can leverage his long arms into basket-protecting dominance”; perhaps stop, take a deep breath, and count to ten.  There may be bigger factors impacting his road to success.

This is the second-to-last installment of this series.  The NBA Draft Combine will be held May 15th to the 19th.  Come back in the aftermath, for a take on which work-out warriors deserve the praise, whose tape-measure failings you can blissfully ignore, and if any players really boosted their draft-day disposition.  Until then.

Kevin’s Summer Project, Part 12: The Beginning of the End

Today, I will briefly discuss relationships between Center defense and pre-draft measurements, before delving into the series summary.  The NBA regular season is nearly over, when playoff talk becomes the rage for the half of the League’s teams…and the lottery serves as solace for the others.

The results of the defensive correlations mirrored offense .  Solid athleticism, measured through max vert, no-step vert, agility drill, and sprint speed, offered relatively strong indication of future defensive performance; of 100 correlations, 85 proved positive, with a few eclipsing 0.50.  Size measurements though, distinctly differed.   Check out the table below:

Center Table

Pretty underwhelming.  Looking at underclassmen, the three shortest wingspans were Chris Kaman, Spencer Hawes and Al Horford.  The three longest?  Hassan Whiteside, Eddy Curry, and James Lang.  Joakim Noah’s wingspan?  85.25”.  Hasheem Thabeet?  90.25”.  While Javale McGee, DeAndre Jordan and Brendan Haywood reached the highest, Noah and Horford ranked in the bottom four of fifty players.  Ok, obviously I am cherry picking information, but size as vital didn’t hit the mark.  Staying away from size would be stupid, but prioritization doesn’t hold water either.

And that is all I want to say about Center defense.  Pre-draft camps start in May; it’s time to start summarizing and reconsidering the outcomes of this series.  Over the next two articles, I will re-explore the predominant trends exposed.

For point guards on offense, speed inclined towards the highest probability of success, particularly at three-quarter court sprint under 3.20 seconds.  Over the last two seasons, recent draftees George Hill and Darren Collision set career-highs for offensive win shares; Ty Lawson won two NBA player-of-the-week awards.  Drafted upperclassmen with this speed and agility-drill-time below eleven seconds universally exceeded draft-day expectations, including Isaiah Thomas in 2011.  No player met these combined thresholds in 2012, but it is definitely something to keep an eye on in May.  Of underclassmen point guards, drafted players with speed faster than 3.15 seconds and no-step vert above 30” always succeeded.  A highly elite group, John Wall certainly made a leap this year.  Kemba Walker missed qualifying by 0.01 second, but appears headed towards a productive career.  Again, no one added themselves in 2012, but any 2013 addition to this elite group will deserve the strong look that they will surely get.

Of shooting guards, the first trend presented was the relative lack of success at the position.  That held true this year, with recent #3 pick OJ Mayo hovering around league-average PER, and lottery-pick Xavier Henry appears headed out of the league.  You know who rules though?  James Harden, but of course, he resides in an elite underclassmen crew of high-leaping long-wingspan players (31” no-step vert, 35” max vert, 6’ – 11” wingspan) .  Who joined that group in 2011 and 2012? Only Jeremy Lamb; he didn’t play much, but with a 14+ PER at age 20, and apparently Sam Presti’s-faith, let’s keep an eye on those measurements in 2013.

While not a die-cast trend, two upperclassmen exceeded those leaping and length measurements in 2011 and 2012: Marshon Brooks and Orlando Johnson.  Johnson is interesting in that despite those awesome athletic tools, he failed at one of the tell-tale marks from this study; an upperclassmen shooting guard with agility drill slower than 11.2.  Only two of seventeen drafted upperclassmen exceeded two career offensive win shares (OWS).  Over the last two drafts, three players joined them; the other two combined for negative-1.2 OWS this year.  Johnson posted 0.6, and considering that I live in Indianapolis, hopefully he follows the relative success of John Salmons rather than his slow-shuffling cohorts.  That said, cast a somewhat leery eye at a prospect sporting this profile.

You know what didn’t prove crucial for either guard position on offense?  Size; between Isaiah Thomas’ play and the back-to-back-to-back rookie of the month awards for 6’ – 3” Bradley Beal and 6’ – 2.5” Dion Waiters, this remains a reliable conclusion.  Don’t let an unimpressive stretch of the measuring tape completely sour you on a favorite guard prospect this year.

That’s it for today.  Come back in two weeks, to run through the remainder of the series, and see if any prior trends lost steam over the course of the season.   Then, onto the 2013 pre-draft camps…

Kevin’s Summer Project, Part 11: Power Forward Defense

Josh Smith's nearly 40" vertical sends alot of opponent shots into the stands.

Josh Smith’s nearly 40″ vertical sends alot of opponent shots in the opposite direction

In Part 5 about Power Forwards, lack of strong correlations between offensive performance and pre-draft measurements was discussed.  To a large extent, that rings true again here.  Of 240 correlations from:

  • Two age groups
  • Four seasons plus “Peak”
  • Eight measurements
  • Three player classifications (All; First-round picks; 2004 & after)

Ninety-nine produced negative results, while only five exceeded 0.40.  Four of those belong to the same subset: underclassmen vertical leap.  As reflected in the table below, this measurement provided exclusively positive results.  But while it functions as the sole reliably indicative number, it also never exceeded 0.50, which early on I set as the threshold for things getting interesting.

Table 1

As this series nears a conclusion, leaping manages to separate from the pack as the most reliably indicative trait for NBA success.  Scrolling through previous articles, those measurements served as a foundation for the discussion of:  underclassmen point guard offense, underclassmen shooting guard offense & defense, center offense, and now, underclassmen power forward defense.  Combined with a propensity for speed to rear-it’s-head as prevalently strong, “explosiveness”, particularly for underclassmen, proves most likely to result in NBA success on offense and defense.  The combination of elite athleticism and young-age merging towards high-marks does not surprise.  Presumably, the nineteen & twenty year olds that warrant drafting are skilled for their age, and when combined with world-class jets & hops, it presents a monstrous combination.

So that is not a shock, but the results of their brethren size-based measurements are.  First, when considering height, this is a player trait typically held in high regard.  Certainly, how tall a player is gains pub from media and fans alike; Player X doesn’t have ideal size, he won’t be able to defend the opposition, etc.  Viewing actual draft-day decision making, a similar preference emerges; correlating draft position with height provides positive values for all positions and age groups, peaking at 0.37 for underclassmen centers.  This is ironic if you recall Part 6, featuring graphical representation of the  strong negative correlation between underclassmen center height and offensive performance.  Excluding small forwards, only 47 of 115 OWS-to-height correlations finished positive, with 0.38 as the apex.  Regarding defense, the results were initially more promising; for guards, 38 of 50 positive correlations with peak of 0.50.  Unfortunately though, these positions prove least likely to make significant defensive impacts.  Moving into today’s focus, the power forwards, barefoot height offered no inclination of future defensive aptitude.  This is reflected in the table below:

Table 2

The results are overwhelmingly negative, with underclassmen exhibiting particularly notable numbers.  In this case, Josh Smith, the overwhelmingly dominant player of the group, stood 6′ – 7″ at draft time.  Other “short” power forwards with decent defensive results include Paul Millsap, Thad Young and Ty Thomas; “tall” player with lesser outcomes are Jason Smith, Troy Murphy and Charlie Villanueva.  Similar to Centers, an over-valuing of height is at-play; for power forwards, correlations between draft position and height proved highest, at 0.30 for underclassmen and 0.26 for the elder group.

Using Offensive WS and Defensive RAPM as the measuring-sticks, the only position where height warrants special attention is that tall wings appear more prone to thrive.  In guards and big men, height receives too much popular credit; i.e. as long as a player is within a reasonable range for their position (no 6′ – 3″ power forwards), this trait can be lightly regarded.

Focusing solely on power forwards again, all non-leaping correlations rarely produced strong values.  Aside from vertical jump, no measurement provided a correlation above 0.30 in season three, four, or peak; including rookie and second years, nothing cleared 0.40. The diversity of roles at this position, and the success of players like Young and Millsap, keep many traits from resembling importance for either offensive or defensive success.

To summarize:

  • Underclassmen power forwards that jump high were most likely to excel at defense.  With 90% of this project published, No-Step and Max-Vert appear to be the “winning” measurements; not that any of this is fool-proof.
  • Height proves overvalued by media, fans, and NBA front offices.  This becomes apparent through both the offensive and defensive data.

Kevin’s Summer Project, Part 10: Small Forward Defense

This series is wearing me out, but I’m in the homestretch.  This article, then power forwards & centers complete the positional analysis for defense. Prior to pre-draft activities, I’ll roll out a summary post, and follow that with articles about the applicability of these results to the outcomes of the 2013 pre-draft camps.

I will start by discussing a difficulty experienced throughout this project though.  I grouped players by positions; it seemed pointless to compare the size and athleticism of 6-footers against bigger, slower front-court mates.  The traditionally defined positions provided an effective and easy means to organize players into meaningful sub-sections.  Of course, this raises the issue of the defined positions not being neatly applicable to all players, particularly “tweeners” and guys capable of occupying various roles.  Today’s article exemplifies that conundrum better than many, discussing the dearth of high-quality shooting guards, and also Andre Iguodala’s elite small forward status.

Iggy defense

Is he a shooting guard, or a small forward? Answer: Both?  Neither? Regardless, he plays hella good defense.

Frequently, as I assigned positions to players, decisions arose.  Determining a clear-cut method for these assignments was not as concise as separating players by age based on a calendar date.   Instead, I perused a variety of sources; in this case Basketball-reference and Draftexpress listed Iguodala at Guard / Forward.  Very early in his career, he paired with Allen Iverson in the back-court.  Later, he often teamed with a combination of Andre Miller, Willie Green, and Louis Williams.  According to 82 games.com, he played the following percentage of the Sixers minutes:

  • Rookie year: 50% at SG, 16% at SF
  • Second Season: 37% at SG, 39% at SF
  • Third Season: 41% at SG, 34% at SF
  • Fourth Season: 17% SG, 61% SF.

At the end of the day, I assigned him to “small forward”.  Decisions like this obviously affect the results, as calling Iguodala a “shooting guard” impacts the correlations.

Ultimately though, this project never intended to pose as “OMG, the code is broken…the NBA draft has been solved!”  Primarily, the goal is providing a means to compartmentalize the results of the pre-draft activities, and to gauge whether certain strengths and weaknesses are properly valued at draft time.   In those regards, I am content with my outcomes; providing general rules of thumb on which characteristics deserve attention, while also improving grounds to ignore much of the measurement data glut.

In Part 4 , I discussed the superior offensive production of small forwards when compared to their brethren at other positions (For a description of Points Stopped, see here).  While that is not true defensively, the shift from the backcourt to this group of players definitely presents a jump in defensive performance.  Of the players lumped into the guard categories, only ten total seasons of greater than one-hundred Points Stopped were recorded; the small forwards registered 31 total such seasons. Standout performers included Andre Iguodala, Kevin Durant, Shane Battier, Luol Deng, Lebron James and Rudy Gay.  Iguodala’s three best seasons ranked in the top-fifteen explored. That trend continues in recent drafts, as Paul George, Gordan Hayward and Landry Fields rank top-five for Win Shares from the 2010 class, with Kawhi Leonard and Chandler Parsons occupying top-four spots from 2011. Unshockingly, the position best-combining size and skill occupies a valuable spot on offense and defense, providing an above-average proportion of the NBA’s best players.

Continuing another theme from the offensive data, athleticism test results were non-vital for defensive success.  For all upperclassmen, 23 of 25 correlations were negative for speed, agility, leaping, and bench press.  Of first round picks, it was 21 of 25. Underclassmen results evenly distributed around zero. Durant, Deng and Battier excelled defensively with marginal athleticism (relatively speaking, of course; these are world class athletes), while Joey Graham, Al Thornton, and Rodney Carney showed jaw-dropping  ability to navigate a cone-drill, sprint in a straight line, and jump, then got drafted in the mid-first round, only to never provide strong offensive or defensive production.

Again similarly to offense, increasing size showed consistently positive trends, but without highly impressive marks. Between two age groups, three player cross-sections, and three measurements, ninety correlations were calculated; 83 were positive, but only 29 exceeded 0.25, with three surpassing 0.50. The three highest all involved barefoot height for upperclassmen first round picks; one data set is represented in the graph below.

Graph 10

Shane Battier dominates this graph, but Danny Granger and Mike Dunleavy also neared 100 Points Stopped as the 2nd and 3rd tallest players on the list.  If this graph was perpetuated into the future, Chandler Parsons moves to the far-right side and supplements this nicely.  I attribute this to the largest “small forwards” combining the size of front-court players, with a skill-set allowing them to be labeled a wing; it is a best-of-both-worlds.

On a final note, height and reach generally provided better correlations with defensive production than wingspan.  Wingspan generates the most publicity, particularly compared to reach, but through three positions this measurement does not discernably separate itself from the others.  As can be seen in the table below for a variety of small forward sub-sets, the correlation of increased wingspan with improved defensive performance rarely exceeded 0.25, and 0.30 serves as the summit.  Iguodala, Battier, Dunleavy and Corey Brewer measured average or worse, and each performed admirably on defense to varying degrees.

Table 10

With that, let’s call it a day.  Onto the summary for small forward defense:

  •  Well-considered, yet somewhat arbitrary decisions were made for assigning some players to one of the five positions.  I can live with this.
  • Moving from the Guards to the Forwards, a large leap in defensive production exists.
  • Correlations of athleticism tests with defensive performance resulted in overwhelmingly negative correlations, particularly for the underclassmen.
  • Size correlations offered the best results, though no specific trait proferred consistently strong values.
  • As an example, through three positions and 80 defensive correlations, wingspan only exceeds 0.30 on twelve occassions, and only surpasses 0.40 once.

Kevin’s Summer Project, Part 9: Shooting Guard Defense

Kobe wasn't drafted between 2000 - 2010, but the other guy was.  Tony Allen locks people down.

Kobe wasn’t drafted between 2000 – 2010, but the other guy was; Tony Allen locks people down.

Three months ago, I covered Shooting Guard Offense…I’m really dragging this out.   The high points of that article included: shooting guards languish offensively; like their PG counterparts, size provided little benefit; upperclassmen with slow agility times rarely succeed; and long, explosive underclassmen almost always do.  So, which of these trends hold-up for defense?  Do other conclusions emerge?

Regarding the first item; shooting guards also served as the most worthless defensive position.  Two weeks ago, we exposed the dearth of impactful point guard defenders, but this group looks worse: the best SG is worse than the best PG; the tenth best SG season slides in as the thirteenth best PG.

With that out of the way, these guys’ defense provided the strongest correlations yet from this series.  For all underclassmen, 32 of 40 correlations were positive.  For upperclassmen, it was 37 of 40, with the three negative correlations generating from bench press.  This is fairly stark contrast with the offensive outputs, when I noted how the frequency of busts resulted in poorly-defined trends.    The results push more extreme when looking at first-round picks only; every non-agility drill correlation proved positive  for underclassmen, while half exceeded 0.25.  Compared to the offensive distribution around zero noted in Part One, this result looks noteworthy.  Certainly at this position, size and athleticism warrant closer viewing on defense.

The prevailing trend toward physical acumen instigating impressive defensive performance, while minimally impacting offense, makes sense.   Scoring buckets is highly skill based: shooting form, while spotted-up and maintaining it on the move; nifty ball-handling; finishing in traffic, with both hands; court-vision and timing on the pick & roll, and in transition…none of these skills are captured by size & athleticism measurements.  Defense though, certainly man-to-man, is greatly aided by overwhelming the opponent with strength, length and speed.  In the post, muscle helps, while on the perimeter, speed and strength prove important, whether attempting to stay in-front of a speedy opponent, or fighting and trailing a guy through a series of screens.  Obviously, smarts, fundamentals, and effort also play a huge role in solid defense, but speed, leaping, and long arms help make-up ground when recovering or helping from the weak-side.

Coming back full-circle though, these players struggled at making a defensive impact.  So, how much should be read-into the defensive correlations?  The primary difference between the offensive and defensive data lies with the results generated when analyzing the size measurements.  Results were appreciably higher for defense, particularly for underclassmen.  For the entire group, of thirty total correlations, twenty-six were positive (including all 15 for upperclassmen), but only four rose above 0.25.  Viewing only first round picks provides increasingly meaningful numbers though.

 While the upperclassmen marks typically end positive, they are definitely unremarkable.  The younger players though, offered markedly better defense with increased length.  Early on, I described 0.50 as the value when I started paying attention.  As their careers progressed, this group approached (or reached) that threshold for both height and reach.  

Table 1

That table doesn’t quite match the strength of the leaping measurements though, which again most-indicated future NBA success for the two-guards.  The table below compares favorably with anything produced in this series.  Most are solidly positive, with an impressive high mark of 0.70, reflected in the graph below.  Uber-athletic Ronnie Brewer and Jason Richardson help prop this up, while floor-bound Dajuan Wagner, Joe Forte and Antoine Wright hold-down the other end.

Table 1

Graph 1

All twenty speed correlations ended positive, however none exceeded 0.36.  Nine of ten agility correlations for underclassmen were negative, but all ten were positive for the older players.  As noted earlier, upperclassmen with poor agility rarely succeeded, with only two players clearing two career offensive win shares.  That theme repeats itself on defense, as only those same two players cleared 15 Points Stopped in a season, and neither exceeded 38 Points Stopped.  I’m still operating under a theory that exceptional agility drill times are more important for role players, a status more likely for older draftees.   Overall, this upperclassmen group as a whole left much to be desired. Only Dwyane Wade and Tony Allen performed better than 55 Points Stopped in at least two seasons.  Of the 51 upperclassmen shooting guards drafted, only four players reached this level in any season.

Conclusions:

  • Shooting Guards provided a poor rate of success from the 2000 to 2011 drafts.  Sure, Dwyane Wade, Brandon Roy and James Harden kick-ass, but the list of players not living up to expectations runs much longer.
  • Generally, the upperclassmen busted.  Aside from Wade, Roy and Allen, the other 48 players combined for 342 Points Stopped over the seasons investigated by this study.  Compare this with Tony Allen’s 300 total Points Stopped in 2010 – 2011 and strike-shortened 2011 – 2012.  Dwight Howard tallied 438 Points Stopped in 2007 – 2008.
  • Size and Athleticism proved more important defensively than offensively, particularly for underclassmen, where the size measurements fared better than in previous portion of this study.  The No-Step Vert correlations for the 19-to-21 year olds reached relatively high levels.
  • Poor agility times doomed upperclassmen Shooting Guards.

Come back in two weeks, to find out if pre-draft measurements continue to prove more valuable on defense.

Kevin’s Summer Project: Part 8, Point Guard Defense

Size, speed and smarts combine to provide defensive results for Kirk Hinrich

Size, speed and smarts combine to provide defensive results for Kirk Hinrich

This post will be a quick diversion from pre-All Star activities, and begins delving into relationships between pre-draft measurements and individual defensive performance.  To update, the base defensive metric is unique to this project (now referred to as Points Stopped), and the group evaluated were players drafted from 2000 to 2010 from the NCAA or high school, over the first four years of their NBA careers, through 2010 – 2011.  Eight pre-draft measurements were correlated versus that metric: barefoot height, wingspan, standing reach, no-step vertical leap, maximum vertical leap, bench press reps at 185 pounds, three-quarter court sprint speed, and agility drill time.

Today, point guard defensive performance gets explored.  In November, their offensive production  was viewed; speed proved most important, with the other athleticism traits also providing consistently positive correlations.  Size attributes offered nothing suggesting importance related to point guard offense.

For defense, the first item to address is that point guards are not very good at it.  Of seasons included in this project, the second-best point guard output was Mike Conley with 127 Points Stopped; there were 32 better seasons by a center, 30 superior campaigns by a power forward, and 23 stronger annums by a small forward.  Of the sixty point guards, five players posted five total seasons with more than 100 Points Stopped relative to the “30-win player”.  Of the fifty-one centers; seventeen guys notched a total of forty seasons meeting this threshold.  Comparatively, eleven point guards contributed nineteen total seasons of 4+ offensive win shares, compared to five centers for eight total seasons.  This disparity does not mean that all point guards were horrendous at defense, but they lacked the ability to differentiate themselves.  The standard deviation of seasonal Points Stopped for point guards is 36, compared to 86 for centers.  With negative-160 Points Stopped over four years, Aaron Brooks occupies the title of the one player whose defense completely offset his otherwise significant offensive contributions.

Surely no one is shocked by these outcomes, but it is important relatively, when comparing the defensive correlations to the offensive results.  Few point guards are defensive difference makers, and defensive correlations need lightly regarded compared to the corresponding offensive numbers.

For the underclassmen point guards, correlations were very weak; of 120 total (forty each for all players, first round picks, and 2004 to 2010 draftees), only eighteen exceeded 0.25, with none higher than 0.50.  Of the size correlatoins, only 2 of 45 rose above 0.25.  The four consistently reliable performers were Chris Paul, Mike Conley, Russell Westbrook and Raymond Felton.  Three of those four stand shorter than six-feet in socks, with wingspans less than 6’ – 6”.  Conley, Westbrook and Felton tested as some combination of fast, agile, or explosive leapers…and Chris Paul is a basketball genius.

So the lessons learned about elite point guard prospects remain unchanged from the prior conclusions: do not place significant value on size; elite athleticism provides a springboard to success; and always draft basketball geniuses when available.  Also, Raymond Felton was a tolerable, to nearly average defender, early in his career.  I did not know this.  In his fourth season, he paced the ninth-ranked defense in minutes and steals, while snagging their third most defensive boards…RAPM approved.

Looking at the upperclassmen, speed again definitely proved most important. The table below reflects the across-the-board positive correlation results, while the graph depicts one particular outcome.

Table 1 

Graph

Chris Duhon and George Hill raise up the top-left corner, while Kirk Hinrich and his 3.10 speed were selected in 2003, and provided the strongest, most consistent defensive performances of the upperclassmen.

Table 2Contrary to the offensive side of the ball though, point guard size mattered for the upperclassmen.  All forty-five correlations between Points Stopped and  height, winspan & reach were positive.  Although none of the values exceeded 0.50, the tally for first-round picks is displayed in the adjacent table.  Players with solid length, adding Earl Watson to the mix of Hinrich, Hill, and Duhon, bolstered these results; the under-six-foot-in-socks crew never provided a season with more than 45 Points Stopped.  Overall though, none of the correlations are overwhelming, nor is cumulative player performance.  Aside from Hinrich, Watson and Duhon, only one season of greater than 55 Points Stopped occurred.  Based on this, the outcome fails to offset the previous inclination towards speed trumping size, with strong play on offense by Ty Lawson, Jameer Nelson, and  Darren Collison helping to build a solid case for that.

I think this topic has been pummeled; let’s wrap up:

  • With the lone exception of Chris Paul’s 2008 – 2009 season, young point guards lack the ability to make major positive impacts on defense, and few seperate themselves from their contemporaries. 
  • Between that, and the defensive correlations being less pronounced than for offense, the prioritization of speed and athleticism still holds merit at this position. 
  • Especially because speed again provided the strongest correlations.   Size can generally be lightly regarded.
  • For upperclassmen though, a slight tendency exists for size to provide defensive benefit.  As discussed in part two, this relationship likely occurs due to older rookies probability of succeeding primarily as role players.  In this case, the extra length aids the group towards defending multiple positions and finding court time because of their versatility.

Next time, this series progresses to shooting guard defense, where the cumulative player performance is even more appalling than for their backcourt mates.

Kevin’s Summer Project: Part 7, A Metric for Defense

i-31

What’s a pic of this guy doing in an article on defense? Not sure, but I’ll try to justify it.

Time to switch focus to defense.  This presented the tricky task of finding a catch-all defensive stat without major flaws.

Steals, Blocks, Defensive Rebounds, and their associated rates all prove too simplistic and skew towards various positions.

Defensive Rating over credits team defense; i.e. the poorest defenders on the elite defensive teams typically rate as superior to the best players on the lesser teams.  This ruled out the defensive win shares built from it.

Opponent’s PER can provide cover to players where their team hides them on defense; when Ben Gordon rated as an elite shooting guard, it seemed non-worthwhile proceeding there.

On-Court / Off-Court numbers fluctuate based on a player’s back-up as well as the quality of teammates and opponents…they do build the foundation for where I headed though.

Adjusted plus minus (APM) intrigued me since becoming aware of it; the number attains a sensibly complex simplicity for evaluating players.  Many things happen on a basketball court that are not capably measured in a box score; in order to ascertain who contributes to victory best, why not run a massive regression with every player as a variable and every line-up matchup as an equation?  Unfortunately, quick online searches only find APM data appearing from 2006 – 2007 to the present, and the numbers are not split into offense and defense.  Also, the numbers are statistically noisy over sample sizes even as large as one season.

Occasionally throughout the 2012 – 2013 Pro Basketball Prospectus, regularized adjusted plus minus (RAPM) gained mention, most notably as it related to leading the field in a group of other stats-based projection systems during the 2011 – 2012 season.  Given my mild inclination towards APM, I was intrigued.  Scrolling through the vast wall of numbers available here, the defensive data matched my understanding of individual performance better than anything else prior.  I became a big fan, developed a number that met the needs of this project, and now we will wander into the murky world of plus-minus stats.

If you just want to know which players rated best at defense, skip the next few paragraphs.  If you want to read about an arbitrary creation of a defensive metric…carry on.  The number derived can loosely be defined as “points stopped compared to a 30-win player”.  Thirty wins became my threshold, which equates to 36.6% wins.  From there, with “A” equal to the NBA’s average offensive rating for a season, I solved the equation:

0.366 = R^14 / (A^14 + R^14), where “R” equals the 30-win player (this equation format is common to calculating expected win percentages, while using a team’s defensive and offensive ratings).

The “R” was always 4 to 4.2 points less than “A”, which when halved, accounted for the defensive value of the “30 win player”.  So, a “30 win player” allowed two more points per 100 possessions than an average player.  With 2 as an example “R” value, the “points stopped compared to a 30-win player” calculates as:

Total Possessions / 200 * (Defensive RAPM per 100 Possessions + 2)

With all of that said, the point of this series is not to advocate for a defensive metric.  Basically, this number gives players credit for the quality of their defense, but also for how much they played.  This was important, as the Offensive Win Shares provided similar context, and in both cases I was trying to avoid overvaluing players that were great at one end, but couldn’t get on the court because of their atrocities at the other.  Finally, the number was primarily contrived so that the vast majority of players provided positive values, but that some small percentage offered negative results.  This is also similar to OWS, with approximately 15% of the players rating sub-zero.  Finally, the rating does not compare players solely to their position, but instead a composite 30-win player.

So, take that for what you will; the defensive parts of this series build upon it.  The obtained results generally matched universal understanding of who played strong or weak defense.  When surprises arose, I reached out to my fellow bloggers for confirmation or refutation (Special thanks to Hickory High, 3 Shades of Blue, The Two Man Game, Queen City Hoops, The Brooklyn Game, Daily Thunder, and Warriors World).  The definite majority provided a response along the lines of “Sure, that’s a reasonable enough result.”

At the end of the analysis and correspondence, I decided the number performed satisfactorily enough to write about.

Today, I will not delve into the relevance of draft measurements as it relates to defense.  Instead, this article ends with three “All Kevin’s Summer Project Defensive Teams” and a little discussion about one surprise.  From that, your opinion of using RAPM may grow or fade.  Remember, these only represent seasons meeting the following screens:

  • Drafted from 2000 to 2010
  • First four seasons post-draft
  • Only includes the 2000 – 2001 through 2010 – 2011 seasons.
  • Player was from NCAA or a High School Kid (pre-2004)
  • That participated in pre-draft measurement combines

Per position, the most quantitatively productive defensive seasons revealed by this study were:

First Team

PG – Chris Paul, 2008 – 2009, 194 points better than a 30-win player.

SG – Dwyane Wade, 2005 – 2006, 158 points better.

SF – Andre Iguodala, 2007 – 2008, 250 points better.

PF – Josh Smith, 2007 – 2008, 393 points better.

C – Dwight Howard, 2007 – 2008, 438 points better

Second Team

PG – Mike Conley, 2010 – 2011, 127 points better than a 30-win player

SG – Ronnie Brewer, 2008 – 2009, 138 points better

SF – Kevin Durant, 2009 – 2010, 248 points better

PF – Josh Smith, 2006 – 2007, 262 points better

C – Dwight Howard, 2006 – 2007, 365 points better

Third Team

PG – Kirk Hinrich, 2005 – 2006, 121 points better than a 30-win player

SG – Monta Ellis, 2008 – 2009, 121 points better

SF – Andre Iguodala, 2005 – 2006, 232 points better

PF – Chris Bosh, 2004 – 2005, 259 points better

C – Jason Collins, 2004 – 2005, 359 points better

Generally speaking, these results met expectations.  Monta Ellis serves as the primary ‘red flag’.  This lead to three conclusions.  First, as expected, the guards offered the least impressive defensive seasons.  Much of the group did not ascert themselves strongly here during the early years of their career.  The tenth-best season was only 80 points better than ”30 win player”, or less than one point per NBA game.  Players qualitatively much better than Monta Ellis, lacked in quantity; one season, Tony Allen provided 94 points of benefit in 2600 defensive possessions, compared to Ellis’s total in 6400 .  Per possession, in his “best” defensive seasons, Ellis was near average; he benefits from playing a lot.

This leads to conclusion number two.   I need to adjust for pace.  Golden State easily played faster than every other team in 2006 – 2007.  While my method avoids over-valuing players that rarely found the court over four seasons (but managed to produce a highly efficient 300 minute campaign), it allows guys whose teams played at blistering speed to benefit.  Factoring for pace, 2007 – 2008 Ronnie Brewer sneaks into the third shooting guard spot, supplanting Ellis (note to hardwoord paroxysm: I have some re-work to do…could be three weeks again).

Finally, Monta’s good defensive seasons were the final two years of his rookie contract.  RAPM considered him a “30 win defender” his rookie season.  In his second and third campaigns, when he earned $660 and $770,000, his defense rated better, at nearly league average per possession.  As soon as the ink dried on his 6 year, $66 million extension, the credit received from RAPM dropped towards “30 win defender” status again; whether this exact scenario actually occurred, at least it is a narrative I can believe.  Playing for a contract happens, right?

It’s time to wrap this up.  To my knowledge, a perfect defensive catch-all does not exist; RAPM work bests.  With some minor modifications, I developed a number that meets this Series’ needs.  Come back next time, for an assessment of which pre-draft measurements most impacted point guard defense.

Kevin’s Summer Project, Part 6: Center Offense

628x471

I must be tired. Can Jeremy Tyler re-claim glory forecasted by high school expectations?

A few interesting tidbits of information will kick off this post.  First, the position of Center provided the largest positive and negative correlations between pre-draft measurements and Offensive Win Shares.  Also, upperclassmen centers serve as the lowest performing group of the ten Position & Age pairings.  Really, it’s not close; of the twenty-eight drafted players, Joakim Noah, Roy Hibbert*, Emeka Okafor and Brendan Haywood prove best.  Duking it out for fifth best…Jason and Jarron Collins and their career single-digit PER’s.  As a general rule of thumb, don’t draft older centers; if they were not considered selectable at a younger age, they probably will not pan out.  In the last twelve drafts, there has been one instance where that decision results in extreme disappointment.   Their performance warrants limited discussion below.

With that; how about starting with the big negative correlation?  One of the most dramatic patterns emerges for barefoot height of underclassmen centers, with short players faring better.  Adding seven-footers to the frontline is a longstanding staple of building an NBA roster.  In the increasingly perimeter-oriented NBA though, that tenet devolves into myth.  Check out the steeply sloping graph below:

Graph 1

Ultimately, the correlation is not extremely strong.  There are three data points skewing the graph negative: Dwight Howard, Al Jefferson, and Al Horford.  The giants include late-bloomers Tyson Chandler and Chris Kaman, as well as injury-riddled Greg Oden.  The most noteworthy aspect involves the correlation of this measurement with draft position; at 0.37, it ranks fourth highest of eighty calculations.  Average draft position of the eight tallest underclassmen centers is eight; for the corresponding shortest, the mean selection is 22nd.  B.J. Mullens and Patrick O’ Bryant rank among the skyscrapers, yet underwhelm.  The trend is also apparent with upperclassmen, highlighted by number two pick Hasheem Thabeet, but beyond him, a list of the thirteen-tallest upperclassmen centers is very ugly.  Trust me on that.

Some NBA decision makers overvalue size in their bigs, when the definite trend runs toward speed and athleticism reigning supreme.  During these eleven drafts, twenty-four underclassmen centers were drafted; the top ten for sprint speed includes: Dwight Howard (fastest), Javale McGee, DeAndre Jordan, Al Jefferson, Greg Monroe, Tyson Chandler and Al Horford.  Only two players of this ten failed to live up to draft day expectations: Steven Hunter and the aforementioned Greg Oden.  Despite this dominance, the correlation between speed and draft position was zero, and the average sprint-time for the top-half of draftees was slower than the bottom-half.

Although only featuring ten data points, the graph below reflects this pattern, in this case after the 2004 hand-check rule changes.

Graph 2

Howard, Horford, Noah, Okafor, and Jefferson bolster this graph while a trio of marginal (or busted) lottery picks round out the bottom.  This disposition towards quickness still holds up through recent drafts.  Of underclassmen in 2011, Nikola Vucevic** ranked among the speediest, actually finishing among the top-ten centers in the 2000′s.  Andre Drummond ran the fastest of the underclassmen centers in 2012; both guys left draft day expectations in the dust.  Before going overboard though, I will note that Brook Lopez and DeMarcus Cousins are recent, slow-footed big men that buck the pattern.

Why have fleet-footed bigs overtaken the favored giants?  On offense, obviously big men increasingly venture away from the basket.  The speed aids in effective face-ups, and also monstrous rolls to the basket and shifty offensive rebounding.  Some of these guys utilize their speed towards elite defense; the more court time due to that skill, the more OWS can be racked up.

Scrolling the database for a measurement combination exhibiting “can’t miss” status, the pairing of top-shelf speed and reach rarely fails.  Since 2000, the NCAA / high school centers with sprint speed lower than 3.30 and reach of 9’ – 2” (110″) are: Dwight Howard, Emeka Okafor, Javale McGee, DeAndre Jordan, Greg Oden, Al Jefferson, Nikola Vucevic and Jeremy Tyler.  I’m not going to blame Greg Oden for his non-success, and Jeremy Tyler…diamond in the rough?

Well, that covers all five positions for offense.   I’m ready to move on to defense and contradict everything from these first six parts .  In conclusion regarding center offensive performance:

  • Centers that turn 22 prior to February 1st of their rookie NBA season almost exclusively provide zero offense.
  • NBA teams are overvaluing size in centers.  A tendency exists towards preferring 7-footers, but that does not work out very well.
  • Like their tiny point guards mates, speed offers the most importance for offensive success.  A combination of reach and quickness results in large propensity for production, and unshockingly, several of the best alley-oop finishers in the league.

Come back in a few weeks for investigation into linkage between pre-draft measurements and defense.

* This is a post-issuance edit.  I forgot Roy Hibbert.  An important note on this whole study is that the data solely focuses on players with available pre-draft data.  Roy Hibbert did not endure the pre-draft measurements and athleticism tests.  Other players I have noticed this about include: Rajon Rondo, Louis Williams, and Lebron has size measurements only.  Surely there are many less successful players that also were injured, strategically declined their combine invite, etc.  As a whole, I have viewed this as something that likely evens out, relative to the 50 – 100 players that have measurements. The oversight of Roy Hibbert was gross negligence on my part though.

** He was 21 as of February 1st of his rookie NBA season, which is the definition of underclassmen that I used for this project.

Kevin’s Summer Project, Part 5: Power Forward Offense

Well, it is an ideal week for this post, with the world in a holiday-based lull.  Power Forward easily provided the weakest correlations between OWS and pre-draft measurements.  Across the entire PF draftee set, of 80 correlations, only four reached 0.25, with a high value of 0.32.  Of eighty first-round correlations, the amount above 0.25 only increased to ten, with a maximum of 0.43.  The strongest marks involved wingspan and reach, but of underclassmen one standard deviation below average, Blake Griffin and Kevin Love register for wingspan, with Griffin and Paul Millsap nearing the bottom for reach.   Sooo…article over?   Not exactly; during the first parts in this series, I was able to focus on point guard speed, shooting guard leaping, and small forward size.  But the overwhelming trends belie the criticality of pre-draft measurements as it relates to NBA offense.  Power forward provides the first real opportunity to talk about ‘no result’ as the result.

What is the antonym for “workout warrior”? I don’t know, but this one stands poised to sign his second big NBA contract.

I root for and blog about a fairly miserable team.  In evaluating prospects, my tendency relies on trusting previous production; i.e. the numbers; both production based, but also age and strength of schedule.  The murky waters of “the more-open NBA game will allow Harrison Barnes’ athleticism to shine”, or “Thanks to his wingspan, Jeremy Lamb projects as a high quality defender”, or “Jared Sullinger is too short and fat to play in the NBA” (none of these are real quotes, to my knowledge) always leave me curious.  Any of these things could be true, but how definite is the precedent?

And this research project is confirming, at least for offense, that most pre-draft measurements do not inform late-breaking, player-evaluation-shifting information into the process.  Maybe this was already obvious, but invariably as follow-up to the combine, glowing reports are penned about the week’s workout warriors, and discussions become increasingly inundated with the recent surprises and confirmations of the strengths & weaknesses of the year’s class.

Examples of upside-down results include:

  • Of 102 power forwards, the seven shortest included Thad Young, Dejuan Blair, Jason Maxiell, and Craig Smith.  These four tallied average peak season of 3.1 OWS, an above-mean value compared to the entire group, from an average draft position of 28.
  • In 2006, Chicago utilized the #4 pick to select Tyrus Thomas, an amazing athlete, who of recent drafted power forwards ranked 95th percentile for no-step vert, 98th percentile for max-vert, 60th percentile for agility, and 84th percentile for speed.  Early in 2010, the Bulls parted ways with Thomas, shortly after selecting Taj Gibson at #26, who tested 8th percentile for no-step vert, 14th percentile for max-vert, 48th percentile for agility, and 17th percentile for speed.   How much did athleticism color Chi-town’s opinion in 2006?
  • Some players that ‘failed’ at every measurement passed the test of time; of the drafted group, Paul Millsap tested at 12th percentile for height, 61st percentile for wingspan, 26th percentile for reach, 44th for no-step vert, 37th for max vert, 38th for agility, and 40th percentile for speed.  At draft time, Kevin Love, the current king of this position, was near or below the median for height, wingspan, reach, and no-step vert, without reaching the top-quarter for max-vert, agility or speed.  Both were highly accomplished rebounders, skilled offensive players, and guys that used their smarts to make an impact on defense.

Has Kevin Love toned-up and slimmed-down since his UCLA days?

Obviously, these anecdotes exist at every position; I am just raising them here.  A primary outcome of the first half of this series is suggesting to not take the combine results overly seriously.   Every draft-able player possesses a variety of skills, smarts, size, athleticism, and intangibles; they will use their combination of particular traits towards success or failure, but very few combine-measurables offer anything remotely resembling a panacea.   There is not a perfect statistics-based draft projection system, but at least my future writing can be grounded in the context of the importance of combine results.  (Of course, defense remains non-discussed; perhaps those articles turn this series on its collective head.)

To close this out, I will hypothesize on the reasons for today’s result consisting of non-results.  One idea probably leads to the other.  First, based on my positional assignments, there were 102 NCAA or high school power forwards selected from 2000 through 2011; twenty more than any other position, and double the number of centers.    Second, and probably the reason for the first, is that “power forward” encompasses the most diverse range of skills.  At the margins of the traditionally defined positions, “point guards” and “centers” receive the most concrete roles.  With a few exceptions, “point guards” are the smallest players, the best ball-handlers, and will lead your team in assists.  There are “scoring point guards” and “distributing point guards”, but both groups fill this same list of descriptions.  “Centers” are tall, ideally seven-foot; they defend the paint, rebound, and hopefully provide scoring near the basket.

The wing positions become more diverse; however they are frequently scorers, whether as a leading guy or a spark-off-the-bench.  Superstars fill many roles, but the remaining group generally gets parsed into ball-handling slashers or spot-up shooters.  Defensive specialists trickle into these positions, but more often than not, they serve as three-and-D guys; dual specialists with a defined role at both ends.

At “power forward”, a multitude of specialists and skills find their place.  Obviously, multi-faceted stars like Kevin Love and Chris Bosh exist, but traditional bruisers like Carlos Boozer and David West still make a nice living.  Matt Bonner and Steve Novak provide excellent floor-spacing; tough guy enforcers like Reggie Evans keep surfacing; defensive specialists, including Taj Gibson make their mark; and players like Josh Smith, Serge Ibaka, and Ryan Anderson blur the lines in-between.  Paul Millsap, the 6’- 6” forty-seventh pick in 2006 can tally 6000 points, 3500 rebounds, and 466 blocks with career PER of 18.7, nearly matching 6’ – 10” Lamarcus Aldridge, the second-pick in the same draft, with career production of 8200 points, 3500 rebounds and 464 blocks (career PER = 19.6).

While no one position is uniform, the variety of ‘power forwards’ covers the most expansive grounds, making it difficult to clearly define a prototype, or specify important size & athleticism traits.

Today’s conclusions:

  • For power forwards, solid relationships between pre-draft measurements and offensive performance do not exist.
  • This was also frequently true for the other positions.
  • For power forwards, this is due to the diversity of functions served.  Are there more dissimilar role-players than Steve Novak and Reggie Evans?  Well, they play the same “position”.

Next up, we will finalize looking at offense for centers, before moving to the other end, where I anticipate things getting more intriguing.