Pelton mail: How much Embiid is enough for Philly?

ByKEVIN PELTON
October 17, 2017, 11:56 AM

— -- This week's mailbag features your questions on first-round options, an early LeBron MVP snub, point differential analysis, and more.

You can tweet your questions using the hashtag #peltonmailbag or email them to peltonmailbag@gmail.com.

There's too many variables to mount a good answer to this question, so let's answer a simpler version: How many games -- and more importantly, minutes -- would Embiid need to play for the Philadelphia 76ers to pass the eighth-seeded Detroit Pistons in projections using ESPN's real plus-minus (RPM)?

Embiid starts out projected for 50 games (the minimum for any player who doesn't have a specific injury going into the season) and 25 minutes per game, a total of 1,250 minutes. Philadelphia's projection is 2.2 games back of Detroit. To make that up requires us adding about 650 minutes from his total and taking them away from Jahlil Okafor, the worst rated by RPM of the Sixers' three centers. That would mean Embiid playing basically a full season (76 games) at 25 minutes per game or a more reasonable 63 at about 30 minutes per game.

I'm a bit more hopeful about Philadelphia than RPM is. As Zach noted in picking them to make the playoffs, the 76ers actually outscored opponents by 3.2 points per 100 possessions with Embiid on the court last season, per NBA.com/Stats. With the additions of J.J. Redick and Ben Simmons, I think lineups with Embiid could be even better this season, and the arrival of Amir Johnson (plus eventually a healthy Richaun Holmes) gives Brett Brown more effective Embiid alternatives than Okafor.

Subjectively, I think if Embiid plays at least 55 games the Sixers are a likely playoff team.

Weirdly, the NBA did not change the date for teams to pick up team options on the third and fourth seasons for first-round picks on rookie contracts along with the deadline for rookie extensions. So teams will still have the first two weeks of the regular season to evaluate those players before making a decision by Oct. 31.

Looking at third-year players, there are several interesting decisions looming. No. 29 pick Chris McCullough -- traded from Brooklyn to Washington in February as salary filler -- seems like a lock to have his option declined. And the Memphis Commercial Appeal reported last month that the Grizzlies plan to waive No. 25 pick Jarell Martin to solve a roster crunch with too many guaranteed contracts.

Looney could be on the bubble. Same with Josh Huestis, the 29th pick of the 2014 draft, who's heading into his third season because he spent his first year in the G League. And then there are several lottery picks whose more expensive fourth-year options aren't sure things. The Philadelphia 76ers have to decide whether declining the $6.3 million 2018-19 option for Jahlil Okafor makes him more or less tradable. Has No. 5 pick Mario Hezonja shown enough to warrant the Orlando Magic picking up his $5.2 million option? And the Chicago Bulls must decide on No. 14 pick Cameron Payne, who won't make as much ($3.2 million) but might not be worth the roster spot.

Third-year rookie options are usually closer to a sure thing; Boston Celtics guard R.J. Hunter was the only player not to have his option picked up last season, when he was waived by the Celtics. The lone 2016 draftee who looks to be in some danger is No. 17 pick Wade Baldwin, who could be cut by Memphis to make room for Mario Chalmers on the roster.

"We often hear a team outperformed or underperformed their point differential. I'm curious if the increase in offensive output league-wide may be skewing some of the results we see from the Pythagorean wins formula. Have we seen an increase in teams under- or overperforming their point differential over the past five years, or am I just dreaming stuff up here?"

-- Frankie Rodriguez

The first thing worth noting, as always, is that the recent rise in scoring is really returning it to historical norms rather than breaking new ground. Last season's 105.6 ppg were the most since 1990-91 -- but would at that point also have been the fewest in the 15 seasons since the ABA-NBA merger.

There is some relationship between scoring and the Pythagorean win expectation for teams. The higher the score, the larger the exponent in the formula should be. (This is why, at an extreme, the formula in basketball raises team's scores to much higher values than the approximate square that made the baseball version look like the Pythagorean theorem, giving the term its name.)

That noted, I don't see much evidence we're seeing point differentials do a worse job of explaining teams' records. There have been 15 teams with a difference of at least five wins between their Pythagorean expectation and their record over the previous five seasons, which is a tad higher than the average five-year period since the ABA-NBA merger (13.9) but not unusually so.

"My question has to do with Steve Nash's two MVP awards. Shaq has often said that Steve Nash robbed him in 2004-05, his first season in Miami, and I've often read that LeBron was more valuable in 2005-06, leading a subpar Cleveland roster to 50 wins while averaging 31/7/6 at age 21. I was wondering if you could use advanced stats to examine those two seasons of Nash's performance as it relates to O'Neal and James, respectively."

-- Muj Syed

Let's look at the whole field of candidates using the metrics I would rely on now: my wins above replacement player metric (WARP), wins generated using ESPN's real plus-minus (WAR) and value over replacement player using Basketball-Reference.com's box plus-minus (VORP):

The first thing that becomes clear is even if you don't think Nash was MVP, O'Neal probably wasn't the right alternative. In his first season with the Miami Heat, O'Neal wasn't nearly as productive as he had been with the Los Angeles Lakers, when he lapped the league in WARP in both 1999-2000 (when he won his lone MVP) and 2000-01 (when he was beaten out by Allen Iverson).

Instead, the top two best players seem pretty clearly to be Garnett and James, neither of whom received a single first-place vote. (James finished sixth in the balloting; Garnett finished 11th.) Both of their teams narrowly missed the playoffs, and that was a bridge too far for MVP voters then (and probably still now).

Somehow, despite Garnett nearly duplicating his 2003-04 MVP campaign, the Timberwolves were barely better with him on the court, according to 82games.com. The Cavaliers were predictably terrible with James on the bench, but his net plus-minus (plus-9.7 points per 100 possessions) paled in comparison to Nash ( plus-15.5, putting him in the league's top five).

In a season without any obvious winner, I'm ultimately probably OK with the voters rewarding Nash for his pivotal role in turning around the Suns. In 2005-06, however, the case for Nash is not nearly as strong despite his value stats being similar.

In a season incredibly deep in star performances -- the five players who topped 19 WARP are tied for the third most in a season dating back to 1977-78 -- James, who finished second to Nash in voting, stands out. (Incidentally, the other player with 19-plus WARP was Elton Brand.) James carried a Cleveland team that was outscored by 90 points in the 605 minutes he spent on the bench to 50 wins by playing an incredible 3,361 minutes. (Last season's leader in minutes played was Andrew Wiggins at 3,048.)

This is not revisionist history: I picked James as MVP for SI.com in 2006, and I'd stand by that now.