Flips & Tricks: Cavs-Knicks, Behind the Box Score
December 8, 2016Joe Thomas isn’t going anywhere as long as Hue Jackson’s in Cleveland
December 8, 2016Andrew Wiggins does not turn 22 years old until Feb. 23, 2017. He is in the third year of his young professional basketball career. He is averaging over 22 points per game for the youngest team in the league, one that is struggling to take an expected leap into relevance.
Yet an article from Dean Demakis is making the rounds that declares Wiggins a certified bust already. In general, basketball analytics don’t look too fondly upon the Minnesota Timberwolves youngster. This isn’t a shocking new development of any kind on the internet. Many will infamously recall the article by FiveThirtyEight’s Neil Paine in December 2014 that compared Wiggins’ career prospects to the mostly mediocre James Posey.
As I read Cathy O’Neil’s Weapons of Math Destruction, I’ve started to become a bit cynical of some of these mass-applied statistical approaches in the sports world. Yes, analytics have assisted in an improved evolution of sports and many fascinating strategy changes. Yes, we can learn a lot from them. But that doesn’t mean these statistics have to be an end-all, be-all evaluator of skill, talent and impact.
The statistical case in these arguments aren’t that difficult to see, even to an untrained eye. Wiggins, who had inefficient shooting percentages in his lone year at Kansas, remains a largely inefficient professional scorer. His career true shooting percentage – which accounts for free throws and three-pointers – is only 53.1 percent. The NBA average is right around 54 percent in the modern game.
Wiggins, who is 6-foot-8 with go-go arms, also fails to gobble up any other statistics that would impress a basketball-focused model. Per 36 minutes of playing time for his career, he averages 4.1 rebounds, 2.1 assists, 2.3 turnovers and 1.0 steal. How does that compare to the average 2016-17 NBA player, regardless of position, entering Wednesday? Those numbers were 6.5 rebounds, 3.3 assists, 2.1 turnovers and 1.2 steals.
But something still smells a bit fishy to me here. I’m not faulting the statistical work of Demakis or Paine, per se. I yelled at those who yelled about Paine’s work nearly two years ago. In thinking over this more and more, my more refined statement is this: We simply have to be more careful about the absolutism of such statements. And we should re-assess if these statistics are really saying what we think they say with the certainty we think they’re portraying.
A prime counter-example is this: Allen Iverson. In the mid-aughts, I recall reading much of David Berri’s Wages of Wins Journal website. Iverson was oft-criticized by Berri and others in the then bourgeoning basketball analytics community, despite being the 2001 NBA MVP and a 2016 Hall of Famer.
The way we talk about sports in 2016 is far different than the way we talked about sports in any earlier era.
From a rate-based analytics perspective, Iverson was never considered as good as he was by the common fan. The same goes with Isiah Thomas, the Detroit Pistons star and 2000 Hall of Famer. His career true shooting percentage was 51.6 percent. One could point out countless similar examples in baseball too, where the stars of decades past are now looked at with a more critical eye by today’s statistics.
But is this totally fair? Keeping with Iverson and Thomas, would anyone ever argue they were a bust compared to their draft expectations? If they were playing today, there’d be think-pieces galore arguing against the logic of the common fan and critiquing the players’ inefficient shot-taking habits. We’d be debating the merits of plus-minus-based stats and PER and points all over again. Many would try to argue these two were below-average contributors, just like Wiggins. The truth however likely lies somewhere in the middle.
The way we talk about sports in 2016 is far different than the way we talked about sports in any earlier era. In the olden days, the only stats that mattered to the casual fan were just the ones listed on the back of trading cards. Nowadays, we are armed with an indefatigable number of statistics and statistics about statistics and even more statistics about statistics about statistics. It creates a polarizing debate between the self-declared informed or the uninformed, with relatively little room for nuance behind a purported bible of truth-bearing numbers.
Despite all these grandiose claims with statistics, basketball analytics struggles to properly weigh the value of usage and role. Neil Paine himself wrote in October 2015 about the many different versions of usage and how analysts (including myself) never seem to agree on the best one. Wiggins, Iverson and Thomas all carry or carried larger than average scoring-producing roles for their respective teams. How could/should a model properly account for that?
And no matter what, defensive statistics will remain a near-impossible thing to measure in any sport. With several players moving constantly at all times, the never-ending amount of spatial-analysis data is overwhelming. Plus-minus-based statistics are inherently flawed for a number of reasons. To assume that these numbers are meaningful and impact-driven seems to be violating many of O’Neil’s rules of Weapons of Math Destruction in her book.
Partially because of the still-debated Kevin Love trade – and even his fascinating career statistics – Wiggins will remain under a close-watched microscope for his entire career. He received too much prep hype and was a No. 1 pick in a tight draft. Being called the “Canadian Jordan” was never fair, just like it’s never been fair to constantly try to anoint a new MJ for decades. The career success of LeBron James, compared to expectations, is not something that happens for every single top pick. That can’t be the case.
Still shy of 22 years old, there is still much uncertainty about how Wiggins’ career could develop. But he doesn’t have to be a multiple-time MVP, sure-fire Hall of Famer and efficient stat-gobbler to still be a productive basketball player for a long time. The Timberwolves will improve as they age. So will Wiggins. It screams as being far too pessimistic and critical to declare that analyzing his game is a solved case already. And it seems almost too trustworthy in the numbers to say that that he’s not helpful at all to the value of a basketball team.
Various links from around the sports interwebs:
- Who’s Next? NASCAR’s unique search for their next superstar reaches far and wide [Joe Posnanski/NBC Sports]
- Basketball’s Nerd King: How Daryl Morey used behavioral economics to revolutionize the art of NBA draft picks [Michael Lewis/Slate]
- Steph Curry might be the grittiest player in the NBA [Kevin Pelton/ESPN]
- How do Kawhi Leonard — and Steph Curry — train their brains? Strobe lights (yes, really) [Tom Haberstroh/ESPN]
- From tanking to a title: Kyrie Irving and Tristan Thompson are evolving – and leading [David Zavac/Fear The Sword]
- Rich Paul and the Art of Building a Powerhouse Sports Agency From Scratch [Zach Frydenlund/Complex Sports]
- Why are the New York Rangers good while the New York Knicks are bad? [The Ringer]
- Bill Simmons Isn’t Too Big To Fail [Kevin Draper/Deadspin]
- Sean McDonough’s Long And Winding Path To Becoming The Best Play-By-Play Guy On TV [Martin Rickman/UPROXX Sports]
11 Comments
https://media.giphy.com/media/3oz8xwxgzsgFqYVVWo/giphy.gif
Yep, that’s the tough part is that everyone needs to understand that “what we know” is likely to continue to evolve so that we can consider that nothing is ever certain. Working in on probabilities and trying to understand deficiencies of the current metrics is laudable.
Like the Britton for AL Cy Young debate that centered around WPA -> personally, think that metric is extremely flawed to use on relievers as it inflates their overall value (when used by itself). There is value in pitching many more innings. There is value in an Andrew Miller usage-model that prevents those later potentially high leverage innings (or allows them to happen instead of being in worse shape). All very malleable and we need to use many different angles and continue thinking things through.
Here’s one analysis:
http://imgc.allpostersimages.com/images/P-473-488-90/97/9730/M6RA500Z/posters/kevin-love-with-the-nba-championship-trophy-game-7-of-the-2016-nba-finals.jpg
I didnt recall that heinous draft outfit. Clearly he’ll end up a scrub.
https://uploads.disquscdn.com/images/374fe57d4a99ffc31139ba0780f2c9ad4df1a8d9d03273200ec9d14e0c402d3a.jpg
Especially now, the idea that there should be any real debate about who won that trade is simply ludicrous.
I’m under the impression that this is one of the all-time classic Draft outfits, no?
Google is paying 97$ per hour! Work for few hours and have longer with friends & family! !mj228d:
On tuesday I got a great new Land Rover Range Rover from having earned $8752 this last four weeks.. Its the most-financialy rewarding I’ve had.. It sounds unbelievable but you wont forgive yourself if you don’t check it
!mj228d:
➽➽
➽➽;➽➽ http://GoogleFinancialJobsCash228MediaTodayGetPay$97Hour… ★★✫★★✫★★✫★★✫★★✫★★✫★★✫★★✫★★✫★★✫★★✫★★✫★★✫★★✫★★✫★★✫★★✫★★::::::!mj228d:….,…
What did you guys wear to the draft then? Oh that’s right you’re douche’s with no talent. How far did your basketball ball career take you jackasses? All the way to a recliner and a laptop so you can tell your dumb friends how your comment owned someone who couldn’t give a crap about you. Sprouting and regurgitating things you think are smart that you heard from someone else and now claim as your own.
Punks.
Internet nerds are peasants.
I think you’re right that it’s much too early to call Wiggins a bust, but it’s misleading to group him with Allen Iverson without mentioning that the analytics liked AI just fine. (He ranks just outside the Top 50 in career PER and in the Top 100 on all the other box-score analytics.) Even Zeke had a good run in PER.
I only mention this because before I read the first Basketball Prospectus I thought AI was a total fraud precisely BECAUSE of his conventional stats, specifically his FG%—it was his high PER and Hollinger’s explanation of usage rate that forced me to admit he was probably better than I thought.
You would appear to be the biggest one of all…