This week's mailbag features your questions on the value of the Celtics' head coach, playoff defense and hot 3-point shooting.
@kpelton how many nba players would you take to start a franchise ahead of brad stevens?
- yusef (@ryunyusef) May 3, 2018
With Stevens leading a Boston Celtics team playing without All-Stars Gordon Hayward and Kyrie Irving to a 2-0 lead in the Eastern Conference finals, his value has become a popular source of debate. FiveThirtyEight's Nate Silver put the bar very high indeed, finding nearly 50/50 results on a poll asking respondents to choose between Stevens and MVP contender Giannis Antetokounmpo after the first round of the playoffs.
As Ben Falk laid out in a subscription piece on Cleaning the Glass last week, even determining the general value of coaching remains difficult, let alone assessing individual coaches. Yet, as Kevin O'Connor pointed out on The Ringer, there are compelling arguments for the value of a great coach like Stevens: Not only would a team get him (presuming extensions) for longer than any individual player, rewarding him with a high salary doesn't affect the team's salary cap, as is the case with a star player.
If we imagine the NBA decided to put all its players and coaches in a draft, and let the 30 teams pick among them, I think Stevens would probably go in the top 30 or 40 -- that is, late in the first round or maybe early in the second. That would put him in the same range as current or future All-Star players.
At the same time, I'm chastened by Stan Van Gundy's departure from the Detroit Pistons last week. If not ever regarded quite as highly as Stevens, Van Gundy was once one of the league's most celebrated coaches. Remember, Van Gundy was perhaps the Golden State Warriors' top target to replace Mark Jackson -- a job that eventually went to Steve Kerr -- before getting the opportunity to run basketball operations in Detroit.
The evidence is pretty clear that basically no coaching candidates are good enough to offer a dual role as head coach and front-office executive in order to land them. Something similar is true of Tom Thibodeau with the Minnesota Timberwolves. While it's less clear whether Thibodeau's coaching is being undermined by his front-office role, two seasons into his Minnesota tenure, Thibodeau has yet to produce the same defensive results he saw with the Chicago Bulls.
Remember, not that long ago it was Thibodeau's example we were using to discuss the upper bounds of value of coaches (and by we, I mean me). That's a pretty good sign that coaches' skills aren't as consistent or transferable as those of star players.
While I think Stevens is more likely to fall in the Gregg Popovich category of creating lasting value by changing with the style of play, odds are he won't look quite as valuable down the road as he does now. Rarely are coaches as good, or as bad, as they seem in the moment.
"I know you often cite that the reason a team lost is not due to how the winning team played, but how poorly the losing team shot when compared to their shot quality as measured by Second Spectrum's measures. However, is there a team in the NBA where their opponents consistently underperform their expected effective field-goal percentage based on shot quality? If there is, would this be indicative of good defenses being able mentally and physically wear down their opponents to the point where even open shots don't end up going in?"
-- Brett Hulihen
This topic is, of course, relevant to the Celtics' playoff run. Over the course of their seven games against the Philadelphia 76ers and Cleveland Cavaliers, they've held opponents to an effective field-goal percentage (eFG) 6.1 percentage points worse than their quantified shot probability (qSP) of 52.8 percent, as measured by Second Spectrum. That figure takes into account the location and type of shot, nearby defenders and the ability of the shooter. So how sustainable is Boston's shot defense?
For your first question, the answer is yes. Shot quality alone can't explain the differences between teams' shot defense. Second Spectrum also calculates a metric called quantified shot making (qSM, which is the difference between a team's qSP and its actual eFG). Teams' defensive qSM in 2016-17 explains about 25 percent of their same rating in 2017-18, which suggests this is a real skill and not just a random effect.
As for why this happens, I don't think it has much to do with good defense causing misses on open shots. If we look instead at how opponents have shot against teams on wide-open shots (no defender within 6 feet) taken from at least 10 feet out, as available on NBA Advanced Stats, the relationship between 2016-17 and 2017-18 is much weaker. (It is worth noting that the Celtics are an outlier; they've allowed the lowest eFG in the league on open shots both seasons.)
Instead, I'd explain it by pointing to factors that are more difficult to capture in shot-quality metrics: Was the shot rushed before a defender closed out? Was there actually a closeout rather than a defender simply being in the area without closing out?
All of those things are real parts of shot defense -- but they're probably limited in terms of impact. During the regular season, Boston ranked second in the league with a qSM of minus-1.1 percent -- meaning that, in this particular measure, the effect of their defense was less than one-fifth as strong in the regular season as it has been in the postseason. So I'm skeptical the Celtics can keep up the kind of shot defense they've produced recently.
"This Aron Baynes 3-point shooting is crazy. It has to be completely unprecedented, right?"
-- Ben Falk
Ben took a break from his site to pose a question about Baynes' hot streak from 3-point range. After making just three triples in 21 attempts during the regular season, Baynes has already made 10 on 20 tries during the playoffs. Seven of those came during the Boston Celtics' 4-1 series win over the Philadelphia 76ers -- more, as Marc D'Amico pointed out on Celtics.com, than either Marco Belinelli or Robert Covington made for Philly.
With at least four more games to go, Baynes already has posted the third-largest improvement ever for 3s made from the regular season to the ensuing playoffs:
Fisher doesn't really belong here; he shows up largely because he played more games in the playoffs than the regular season because of an injury. The other stories are more interesting. Russell had made just 29 3s through his first three seasons before breaking out with 25 in the 1996 playoffs. The improvement propelled him into a starting role the following season.
Perkins had never made more than 19 3s in his first eight seasons before being encouraged by George Karl to let fly after a midseason trade to the Seattle SuperSonics. He made 19 the rest of the regular season, 30 in the playoffs, and 99 the following year as he developed into a stretch 5 far ahead of his time.
The elder Dunleavy was still adjusting to the 3-point shot in its second season and went 1-of-16 during the regular season before going 6-of-15 in the playoffs. Within two years, he'd lead the NBA in made 3-pointers.
That leaves Anderson and Wade as the two players who did not see the 3-point evolution carry over. Anderson, 1-of-8 in the 1992-93 regular season, simply got hot at the right time with six 3s in 11 attempts. He made 22 the following season, but never more than 34 in a season even after the line moved in temporarily starting in 1994-95.
Wade had made as many as 88 3s in a season earlier in his career, but never at a high percentage, and he was 7-of-44 during the 2015-16 regular season before famously declaring that he would actually practice shooting 3s. His 12-of-23 shooting from long distance was instrumental as the Miami Heat rallied to beat the Charlotte Hornets in the first round before taking the Toronto Raptors the distance. When Wade started 2016-17 by making 10 triples in 19 attempts over his first five games, it appeared practice might make (near-)perfect. However, Wade shot just 28 percent the rest of the season.
Of course, what separates Baynes from these other players is they'd at least made some 3s before, if not necessarily so many. Baynes had previously made just four total 3s in his career. So this is definitely an unprecedented development. Based on other similar stories, it's possible Baynes has actually added the corner 3 to his full-time repertoire, but it's also realistic that this is nothing but a well-timed fluke.