APBRmetrics

According to reports I've followed from the sportswriters serving as potential voters on the 2017-18 MVP award, it appears it's down to two players -- James Harden and LeBron James.

Harden is the overwhelming favorite due to the advantage he holds with his teams success, the Rockets currently own the top record in the NBA and are projected for 66 wins. James' team is currently ranked 4th the Eastern Conference, projected to win 51 games.

Last year, Russell Westbrook won the MVP by 135 points over 2nd Place James Harden. Westbrook's Thunder were a 6th seed in the Western Conference, winning just 47 games.

The Offensive comparison

James
(proj.)
Total pts = 2271
Total ast = 752
Total off reb = 99
Total tov = 349
TS% = .621

Harden
(proj.)
Total pts = 2229
Total ast = 636
Total off reb = 41
Total tov = 318
TS% = .620

Westbrook
(2017)
Total pts = 2558
Total ast = 840
Total off reb = 137
Total tov = 438
TS% = .554

2017 Westbrook has the higher volume of production, but is clearly the most inefficient of the bunch. Yet, Westbrook finished last season with nearly 70% of all first place votes. Harden finished with 47 less first place votes than Westbrook last season with numbers that are overall similar to this year's campaign. One difference, the Rockets acquired Chris Paul, and are on pace to win 11 more games than last year. The Rockets are 15-8 without Paul this season (65%). In other words, the Rockets are basically the same team they were last year (55-27, .67 win%) without Paul in the lineup. Paul, in nearly 60 games this season is clearly explains the 10 or 11 win improvement for the Rockets.

James and the Cavs lost Kyrie Irving in the 2017 offseason. The Cavs are currently on pace for the exact same record as last year. The one constant, is that James has elevated his play to compensate for the absence of the All Star Point Guard.

Like Harden, James' most effective teammate missed over 20 games. The Cavs went 13-9, unlike the Rockets minus Paul, showed no drop-off, with James again picking up the slack in his absence. The Cavs have had 18 different starters this year, James and JR Smith the only players to start at least 60 games. JR Smith has been a liability for the Cavs, whereas Capela and Ariza -- the 2nd and 3rd most consistent Rockets' starters -- have made positive contributions for the Rockets, especially on the defensive end.

The Defensive Comparison

Critics have come out in full force against James' defense this year, some in same breath praising Harden on that end of the floor for the first time in his career. The Cavs have one of the worst defenses in the NBA. But it's important to point out that most leading defensive metrics such DPM or Defensive Rating, don't have way to isolate individual defensive performance apart from team defense. For example, having a solid rim protector can cover a lot of flaws in perimeter defenders if you just take in the advanced stats.

The Cavs have been awful defensively, 2nd to last in team Defensive Rating. But that doesn't necessarily translate into LeBron being terrible. LeBron is far in away the Cavs best shot blocker (71), the next closest is Jeff Green (32). James is arguably the Cavs best perimeter defender as well, George Hill and Cedi Osman honorable mention. The average offensive player that LeBron guards shoots .459 from the field against the rest of the league, when James guards them, he holds them to .446, 1.3% below their league average. In Harden's case the average player he defends shoots .458 from the field regularly, when Harden defends them they shoot slightly higher at .461 from the field. Harden also has 142 defensive fouls in 460 less minutes than James (including 20 more shooting fouls) -- James has committed only 100 defensive fouls in a league leading 2977 minutes played.

Harden plays with a bonafide shot blocker in Clint Capela. Meanwhile Eric Gordon is easily the perimeter defender on the Rockets this season. But NeNe, Mbah a Moute, along with Capela and Gordon have more so than Harden helped the Rockets to top defensive Team this year. Ariza, Tucker, and Paul although inconsistent, have arguably been better than anyone playing the majority of the 2017-18 season on the Cavs roster not named LeBron James. The Rockets currently have no players playing as poor defensively as Kevin Love, Tristan Thompson, and Rodney Hood.

In short, the Rockets have stout defensive players that help provide the content for the defensive metrics like Defensive Rating and Defensive Win Shares which inflate Harden effectiveness on the defensive side of the ball. When we attempt to contextualize base on team personnel and isolate each players defensive performance through the player tracking data it appears to show James having regressed from an elite defender to slightly above average, and Harden slightly below average.


So James has had the most offensive impact, both through total minutes on the court and offensive efficiency. James is also the better defensive player. And with Westbrook setting what may be a new precedent last year, winning the MVP while playing on a 6th seed, let's see if the sportwriter's -- armed with more advanced stats in the last four years than all of prior NBA History -- will actually select the league's most valuable player for the MVP award.

You can pretty much throw stats out the window when discussing MVP. It's primarily narrative driven. There's nothing very compelling about the best player of his generation leading a team to a 42-38 pythag record in the Eastern Conference. Harden will more than likely win this in a landslide.

...James has had the most offensive impact, both through total minutes on the court and offensive efficiency. ..

More minutes, yes. Where do you get the rest of this?
ORtg
121 - Harden
118 - LeBron

OBPM
9.7 - Harden
7.6 - LeBron
http://bkref.com/tiny/XzIeI

ORPM
7.0 - Harden
5.3 - LeBron (after also Curry and Paul)
http://www.espn.com/nba/statistics/rpm/_/sort/ORPM

RPM-Wins ranks them Harden-Westbrook-LeBron, accounting for minutes.
In both offensive and total Win Shares it's Harden-LeBron...

When Harden missed 7 games in midseason, the Rockets went 4-3. They outscored opponents by 3.6 PPG, and these teams were on avg -1.7 in SRS. So Hou was just +1.9 SRS without Harden (and with Chris Paul).
Subtract that 1.9 over 7 games from their season 8.6 * 80, and you get SRS = 9.2 in the other 73.
The difference is +7.3 PPG when Harden is available.

LeBron hasn't taken any rest days, but he did seem to tank in January. Without that bad month, he'd certainly be up there as a co-frontrunner.

shadow wrote:You can pretty much throw stats out the window when discussing MVP. It's primarily narrative driven. There's nothing very compelling about the best player of his generation leading a team to a 42-38 pythag record in the Eastern Conference. Harden will more than likely win this in a landslide.

So the fact that he's the best player of his era should be held against him?

Mike G wrote:

...James has had the most offensive impact, both through total minutes on the court and offensive efficiency. ..

More minutes, yes. Where do you get the rest of this?
ORtg
121 - Harden
118 - LeBron

OBPM
9.7 - Harden
7.6 - LeBron
http://bkref.com/tiny/XzIeI

ORPM
7.0 - Harden
5.3 - LeBron (after also Curry and Paul)
http://www.espn.com/nba/statistics/rpm/_/sort/ORPM

RPM-Wins ranks them Harden-Westbrook-LeBron, accounting for minutes.
In both offensive and total Win Shares it's Harden-LeBron...

When Harden missed 7 games in midseason, the Rockets went 4-3. They outscored opponents by 3.6 PPG, and these teams were on avg -1.7 in SRS. So Hou was just +1.9 SRS without Harden (and with Chris Paul).
Subtract that 1.9 over 7 games from their season 8.6 * 80, and you get SRS = 9.2 in the other 73.
The difference is +7.3 PPG when Harden is available.

LeBron hasn't taken any rest days, but he did seem to tank in January. Without that bad month, he'd certainly be up there as a co-frontrunner.

If both players played the same amount of minutes, I'd reevaluate my conclusion that LeBron should be the frontrunner in the MVP race. But there's a 450+ difference in MP in favor of LeBron. My preferred formula borrows from AWS as you may know.

((((PTS+.8*ORB+.5*AST-.8*(FGA-FG)-FG-.35*(FTA-FT)-.5*FT-1.2*TOV))))

LeBron James (81 GP) offensive score = 607.40
James Harden (71 GP) offensive score = 519.95

*Not included: James has 65 secondary assists to Harden's 57
James has 23 offensive fouls to Harden's 27

In my assessment James has easily had the biggest offensive impact.

D-rell wrote:
So the fact that he's the best player of his era should be held against him?

No. The point is MVP isn't a lifetime achievement award. LeBron can be the best overall player in the era without necessarily having the best individual season every year. No one typically ranks Barkley or Malone ahead of Jordan on the all-time greatest players list, but both of them managed to win an MVP over Jordan.

One way to mesh the "narrative" approach with the "statistical" one is to estimate the impact a player has on his team's chance of winning a title.
The Cavs (according to b-r.com) have a 0.7% chance to win it all. This of course does not account for "playoff LeBron" to replace the season version, but his season has been pretty solid. It also figures the Cavs' whole season strength, rather than where it is now, in an 11-2 run.
So one can argue LeBron moves the needle from zero to a few percents.

Houston has the highest prob% at 35.7, based on season SRS of 8.60 and likely playoff competition.
https://www.basketball-reference.com/fr ... _prob.html
In games without Harden, their SRS was about 2 or 3. Such teams and their title prob% -- Por (1.3), OKC (1.7), SAS (1.7), Ind (0.9)

Without Harden, the Rockets would seem to be another team chasing the Warriors, back in the pack with these others.
Cavs without LeBron might win a playoff game.

Jordan was season MVP and Finals mvp in the same year 4 times.
Very few have done it even twice -- Bird, LeBron; Erving in the ABA

If Houston were truly a 2 or 3 SRS team without Harden, that would imply he's worth upwards of 19 wins by himself. I find that hard to believe. I think if you replaced Harden with a league average shooting guard the Rockets could still win up to about 55 games. They'd essentially be the CP3 Clippers with better shooting and roughly the same level defense. They would only lose about 10 wins by replacing Harden's minutes with a league average SG by my quick minutes weighted RPM calculation.

Average players make 8.2 wins per position, for a total of 41. The Rockets won 65, but their MOV suggests 61 would be expected.
A team with +3 mov will win around 50 games. So Harden might have added 11 wins, or 15 if he was responsible for winning most of the close ones.

You predicted 55 wins, and then Chris Paul missed 24 games; Capela 8, Gordon 13, Ariza 15, Anderson 16, Luc 21, Nene 30 ... And they won 65. How did they do it?

When did the voting close? It should be closed before playoffs start. I assume the announcement date is choreographed for later but when? Wait longer and it might get awkward.