Time for another edition of the NBA mailbag.
Throughout the NBA season, I will be answering your questions about the latest, most interesting topics in basketball. You can tweet your questions using the hashtag #peltonmailbag or email them to peltonmailbag@gmail.com.
This week's edition includes the Brooklyn Nets' best star pairing, the effects of baseball-style series on this NBA season and finding a Mendoza line for NBA 3-point shooting.
"Which pairing do you feel will be the most successful this season for the Nets: Kevin Durant-Kyrie Irving, Durant-James Harden or Harden-Kyrie?"
- Ashwin Dias, Mumbai
It's been intriguing to watch how coach Steve Nash has distributed the minutes for these players together. In the four games during which all three of Brooklyn's stars have played thus far, they've logged 124 minutes (60% of the team's total) as a trio. Of the three possible duos while a third star rests, Nash favors the Durant-Irving pairing that has been working together since the start of training camp. They've played an additional 32 minutes (15%) in those four games as a duo, with just 11 minutes (5%) for Durant and Harden and six minutes (3%) for Harden and Irving.
Despite a dominant win over the Oklahoma City Thunder on Friday without Durant, I think it's clear that Harden and Irving are the weakest duo of the three options -- both because their skill sets tend to overlap more than either lead ball handler with Durant and because the Nets badly need Durant's defensive presence. So I'm not surprised Nash has shied away from pairing them.
The real question, then, is which All-Star guard better complements Durant. My instinct is to say Irving, since Harden's game has trended away from the pick-and-roll and toward isolations over time. Not since he played with Durant in Oklahoma City has Harden been paired with such a versatile screen setter; whereas Irving had experience with someone broadly similar in LeBron James with the Cleveland Cavaliers. But maybe that's the wrong way to think about it, given Brooklyn has rarely used Durant as a screener for either guard (just 35 total pick-and-rolls with Irving and 25 with Harden, per Second Spectrum tracking).
Still, the Nets' coaching staff seems to like the Durant-Irving chemistry. Even before adding Harden as a scorer capable of running an offense without either other All-Star on the court (Harden has gotten 28 of the 32 minutes in the four games together during which Brooklyn has played with just one of the three), Nash was synchronizing the minutes for Durant and Irving rather than staggering them to keep one on the court at all times.
Ultimately, it might just come down to who's the better player between Harden and Irving, and that's been Harden over the course of their careers. There is a degree of chemistry between Durant and Harden dating back to their Thunder days, and -- with the caveat that more of these minutes came with other starters because Irving missed the first two games after Harden's debut -- that duo has a sizzling +15 net rating in 71 minutes together. So for now, I'd say Harden, though we'll get a better idea of the answer to this question with a larger sample size that makes such lineup metrics meaningful indicators.
#peltonmailbag What is the influence of much more "baseball style" series in this season. Is it a good news for underdogs because of more chances to get a 1-1 in these two consecutive games between two same teams?
— milan2005 (@milan20056) January 26, 2021
So far this season, there have been 34 of these baseball-style series completed with two games between the same teams in the same location. Home teams have gone 17-17 (.500) in the first game and 24-10 (.706) in the second game. Specifically to your question, we've seen 19 sweeps out of 34 matchups, which is to say the team that won the first game has gone 19-15 in the second game (.559).
For comparison's sake, we can look at what happened when two teams met twice in the same location during the pre-stoppage portion of the 2019-20 season -- which were scheduled with many games in between. Last season, the team that won the first of those two meetings went 122-93 (.567) in the second game. So splits haven't become noticeably more common thus far this season during series.
In general, the idea that it's harder to beat an opponent multiple times in a row is something of a canard; I think people are influenced by the fact that by definition the team that won the first game is less than 100% likely to win the second game, which is to say regression to the mean. This notion has been debunked by Football Outsiders when it comes to playing a division opponent a second time in the NFL regular season; and Jordan Sperber found college basketball teams that win both head-to-head meetings in the regular season are overwhelming likely to win a third in a conference tournament.
To be fair, we do see evidence of the losing team having an advantage in the next game in NBA playoff series, known in gambling as the "zig-zag theory." But I suspect this is more due to the additional motivation of being in worse shape in the series than the ability for the loser to make adjustments.
I also wouldn't be surprised if we eventually do see more splits of the baseball-style series as we get a more robust sample size simply because home-court advantage is smaller without fans in most arenas, meaning road games are closer to a toss-up than in years past. Thus far, the home team going 41-27 (.602) in series is something of an outlier. In all other games, home teams are a game below .500 at 103-104. It's unlikely the series are increasing home-court advantage; it's probably just random noise in a small sample.
"What is the Mendoza line for 3-point shooting?"
- Kevin Arnovitz, Los Angeles
Kevin asked me this question on our recent "Small Sample Size Theatre" episode of Brian Windhorst & The Hoop Collective, and I was unhappy with my answer. I blurted out 33%, which is more or less the point below which additional 3-point attempts are probably a negative instead of a positive. But that's not the spirit of the Mendoza line, which refers in baseball to whether a player's batting average is better or worse than .200, a minimum bar for major league-caliber. (It's named for 1970s and 1980s journeyman infielder Mario Mendoza, whose career batting average was .215.)
Per Baseball-Reference.com, just 8.4% of all batters with at least 100 at-bats over a season between 2016 and 2020 have finished it hitting .200 or worse. If we look for the equivalent percentage of seasons with at least 100 3-point attempts between 2015-16 and 2019-20, we find that 30% comes closest to this rate (112 of 1,200 seasons, 9.3%). Conveniently, that was my second and more considered guess on the podcast.
The next question is which player best embodies this line as Mendoza did in baseball. Because 3-point percentage is clearly not as central to player value in basketball as batting average in baseball, this reveals a paradox: the players who shoot the most 3s while hovering around 30% accuracy are often Hall of Famers who can get away with it.
Most seasons below 3-point Mendoza line
• Charles Barkley: 8
• Corey Brewer: 8
• Jerry Stackhouse: 8
• Russell Westbrook: 7
• Dominique Wilkins: 6
• Allen Iverson: 5
Source: Stathead.com; minimum 100 3-point attempts
I would probably nominate the "Brewer Line," after the longtime defensive specialist who cracked 30% accuracy from 3-point range just once in the nine seasons during which he attempted at least 100 3s.