Big NBA MVP questions: Should Giannis be the front-runner?

Where does the NBA's MVP race stand nearly two months into the 2018-19 season?

It's never too early for MVP narratives to take hold, as what was once a topic for the spring now has become a season-long discussion. And indeed, betting markets and straw polls suggest that Milwaukee Bucks forward Giannis Antetokounmpo is the early MVP leader on the strength of his team's surprising start, with LeBron James the other leading contender in pursuit of his fifth Maurice Podoloff Trophy.

Yet narratives don't always match statistical reality, so let's ponder some of the big questions about the MVP with the help of advanced metrics, including whether Giannis is the right front-runner.

Can a center be most valuable?

The leaderboards in most advanced stats at this early stage have a decidedly tall slant. Basketball-Reference.com's value over replacement player (VORP), for example, rates six centers and four other players listed as power forwards (one of them is  Kevin Durant, who has actually played more minutes at small forward) among the league's top 14 players. Meanwhile, the PER leaderboard -- once we've filtered out low-minutes players -- is even more heavily tilted toward the frontcourt, with 19 centers in the top 30, including reserves Montrezl Harrell, Nerlens Noel, Jakob Poeltl, Dwight Powell, Domantas Sabonis and Jonas Valanciunas.

These results reflect a topic I addressed ahead of the NBA draft: Because the floor is so well spaced, it's never been easier to get productive play from centers. Indeed, consider the average PER by position (as defined by Basketball-Reference) weighted by minutes played:

Across all positions, the average PER is always set to 15.0. However, that varies widely across positions now, with centers naturally dominant and wings much weaker. It's not just the very best centers who are most productive, but also the ones who are cheaply available to teams -- including Noel and JaVale McGee, both of whom are in the top 30 in PER after signing one-year deals for the veteran's minimum as free agents this summer.

When we're determining value, then, position has to be taken into account. To be as relatively valuable on a per-minute basis as a small forward with a 25.6 PER (like Kawhi Leonard), a center would need a PER of 32.2 -- something that's never been achieved in NBA history.

In other words, it's probably not possible for the best center in 2018-19 to be more valuable than the best wing player, which has important implications for the MVP chances of not only Joel Embiid (who's not actually among those high-rated centers, ranking 17th in PER and 26th in VORP) but also Anthony Davis (mostly -- though perhaps not exclusively -- a center).

Should Giannis be the front-runner?

Although I used PER over average to make that point, the better way to define value in the context of MVP is as compared to replacement level -- the production we'd expect from a free agent signed for the minimum salary -- so as to properly credit players for their durability.

My wins above replacement player (WARP) metric now reflects a much higher level of replacement for centers, and to a lesser extent power forwards, as compared to wings in particular. With that adjustment, here's how WARP ranks this year's most valuable players (projected to 82 team games so as not to disadvantage players whose teams have played fewer games thus far).

While LeBron tops this list, surprisingly it's last year's winner who's in a virtual tie with him.  James Harden's stat line is basically identical to what he posted in 2017-18, but because his Houston Rockets have stumbled to an 11-12 start after winning a league-best 59 games last season, he's gotten zero attention as a possible repeat candidate. It will be interesting to see whether that changes if the Rockets turn things around. Damian Lillard, whose Portland Trail Blazers have fallen off after a strong start, is in a similar position after finishing fourth in last season's MVP voting.

The most interesting MVP candidate that position-adjusted WARP identifies is  Paul George, who also ranks third in wins produced based on ESPN's real plus-minus (RPM), which isn't adjusted by position. George has played a key role in the Oklahoma City Thunder's league-leading defensive rating and has helped keep their offense afloat with 2016-17 MVP Russell Westbrook missing eight games. According to NBA Advanced Stats, Oklahoma City's offensive rating drops to a dismal 94.7 points per 100 possessions with George on the bench.

Those takeaways noted, let's dig deeper into the comparison between Giannis, LeBron and KD -- only two of whom can make the All-NBA first team because they all play forward. Why does Antetokounmpo lag behind the other two stars?

The first reason is the lower replacement level at small forward, where Cleaning the Glass estimates James has played 66 percent of his minutes and Durant 65 percent as opposed to 12 percent of Antetokounmpo's minutes.

The Lakers' positionless lineups show the challenge in using position to determine value. Cleaning the Glass considers LeBron the small forward when he plays alongside power forward Kyle Kuzma, but in practice the defensive matchups with those lineups depend more on the opponent than any rigid position. And offensively, James tends to operate more as a point guard (particularly during Rajon Rondo's absence, as ESPN's Brian Windhorst recently broke down) than as a forward.

However, I think Durant's example reinforces the importance of positional replacement level. His ability to defend small forwards has allowed the Warriors to start Jonas Jerebko -- another big man who's excelled while making the veteran's minimum -- at the 4 in Draymond Green's absence rather than going deeper into their weak wing rotation.

The other key factor here is minutes played. Not only has Giannis missed a game due to injury, he's playing fewer minutes per game (33.8) than either LeBron (34.8) or KD (35.7), meaning Milwaukee has relied slightly more on its bench than those teams. While the difference isn't dramatic, along with position it helps explain why Antetokounmpo can rate better on a per-minute basis than James and Durant, yet still land behind them in terms of value.

When it comes to RPM, Antetokounmpo is actually third of the group in terms of per-possession rating and just 12th overall in RPM wins. As a result, I don't think Giannis should be considered the MVP front-runner based on advanced stats.

How many games does Curry need to play? How about Kawhi?

Though playing time is a relatively small difference between the MVP front-runners, it's a huge factor for  Stephen Curry (who has missed 11 of Golden State's 26 games) and a large one for Leonard (he's missed six of 26). When Harden won last season, the 10 games he missed were the most for an MVP since Allen Iverson (11) in 2000-01. That would basically imply Curry can't miss any more time and have a chance to win.

However, history should be considered a guide rather than a hard-and-fast rule, as Westbrook winning the MVP on a 47-win team offers a recent reminder that voters can always rethink MVP tradition. So instead, let's ponder the question of how many games Steph and Kawhi would need to play to be most valuable statistically. If we flip the leaderboard to WARP per 82 player games rather than team games, Curry shoots (pun intended) to the top:

Nonetheless, in part because he's played just 33.3 minutes per game, Curry will be hard-pressed to lead the league in WARP. Harden would beat him if they both played the remainder of the schedule at their current pace, and LeBron's 82-game pace isn't far behind. Voters might be willing to reward Curry even if he's slightly less valuable if it's strictly due to injury, but he can't miss many more games.

Since he hasn't been as dominant when he's played, Leonard's task is even more difficult. While he cracks the top 10 on a per-player game basis, Kawhi is still behind both Giannis and LeBron and would have to rely on the narrative power of leading the Raptors to the NBA's best record. They're currently three games up on the Denver Nuggets for that honor, with the Warriors four games back.

Curry staying healthy the rest of the season might actually have more impact on other candidates for MVP than his own hopes. If Golden State surges ahead of Toronto with Curry (the Warriors are 12-3 in games he's played this season and 5-6 without him), that could hurt Leonard's candidacy. It also becomes a lot more difficult for voters to pick Durant if he's not even the most valuable player on his own team -- something that was evident when Golden State struggled early on in Curry's absence, though it did give Durant the opportunity to put the team on his back during the Warriors' final four games without Curry, when they went 3-1 with an overtime loss in Toronto.

Whether Curry can stay healthy the rest of the season is just one of the many questions to be answered about the MVP race, which should still be considered wide open with more than two-thirds of the season left to be played.