A couple of years ago, N'Golo Kante might have been the most valuable soccer player in the world.

In August 2015, Leicester City signed the relatively unknown French midfielder from Caen for £5.6 million. The season before his arrival, Leicester finished with 41 points. In Kante's only season with the club, Leicester finished with 81 points, winning the most improbable championship in the history of modern sports.

Then, in July 2016, Chelsea signed the by-now-well-known French midfielder for £32 million. The year before Kante's arrival at Stamford Bridge, the club finished with 50 points, by far the worst tally since Roman Abramovich bought the club back in 2003. In Kante's first season with the team, Chelsea finished with 93 points, winning their fifth Premier League title. Leicester, meanwhile, held onto all of the other key contributors to their championship team, but without Kante, they almost slipped right back down to where they were before he'd arrived, finishing with 44 points.

Ignore any other context and it looks like Kante alone was worth something like 40 points a year. That's obviously not quite true, but it certainly seemed like Kante was one of those rare players who could significantly elevate his team's performance without putting the ball into the back of the net. And in the summer of 2017, a group of researchers from the University of Salford and University College of London published a paper that supported this very idea.

They created a number of models to determine a player's "plus-minus" rating, that is comparing how a team performs when a player is on the field with how it performs when he's not. These models are tricky and imperfect because unlike in basketball, most teams don't spend a significant amount of time with their starters off the field; therefore you get things like Manchester City's Claudio Bravo being the second-highest rated player in the world for the 2016-17 season. However, the results of the research also seemed to verify what everyone at Stamford Bridge and the King Power Stadium was thinking about Kante.

- Marcotti: Soccer has always been bigger than sports
- Ogden: Does England have a traveling fan problem?
- Williams: Arsenal used to be invincible. What happened?

"The paper presents a method for estimating how important each player is to a team's success," Ian McHale, one of the paper's authors, said over email. "For the seasons under consideration, Kante was found to be the player contributing the most to a team's success. And this doesn't just mean Leicester and Chelsea, but a hypothetical team made up of any set of players: Kante would contribute the most."

Except, in the two seasons since then, something has happened that had never happened before ... and it's happened two seasons in a row: The team Kante plays for didn't win the Premier League. In 17-18, Chelsea finished in fifth and last year, they landed third.

"Kante has plummeted down the rankings in the last 12 months, probably because he was played out of position for much of that time," said McHale.

Much of the discussion surrounding Kante over the past year-and-a-half echoes what McHale suggests. Since Chelsea signed Jorginho, a relatively immobile player who needs to sit in front of the defense and constantly have the ball at his feet in order to be effective, Kante has been deployed higher up the field under both Maurizio Sarri and current manager Frank Lampard. This discussion between BBC pundits Alan Shearer and Ian Wright serves as a good summary of the general sentiment:

Jorginho or Kante as Chelsea's holding midfielder...

Do you agree with @IanWright0's assessment? pic.twitter.com/hRfFVTK4k3

— Premier League (@premierleague) November 26, 2018

However, Sarri claimed he needed a different player profile in his holding midfield role. "In that position, I want a player able to move the ball very fast... N'Golo is very useful for us, but this one is not his best characteristic."

In August, Lampard echoed those thoughts. "This idea that he wins the ball probably as well as anybody in world football doesn't mean that he has to sit in front of the back four and do that. He also has too much in his game to drive forward with the ball, to lead midfield areas and win the ball back high up the pitch. That's what I want to give him the freedom to do."

Even when France won the 2018 World Cup, Didier Deschamps opted to play Paul Pogba, of all people, in the holding role with Kante given the kind of freedom that Lampard talks about: license to win the back higher up the field.

Three separate coaches have envisioned a more aggressive role for Kante. Could they all be wrong?

Last season under Sarri, all of Kante's defensive numbers declined except one. In his sole season at Leicester, he made 5.22 tackles per 90 minutes and 4.65 interceptions, both of which led the Premier League. In his two seasons under Antonio Conte, he averaged 3.67 tackles and 2.45 interceptions and last year, he was down at 2.18 tackles and 1.28 interceptions. In the Leicester and Conte years, he won possession in midfield an average of 4.85 times per 90 minutes, ranking in the top three in all three years. Under Sarri, that number dipped down to 3.26.

What didn't dip, though, was the number of times he won possession in the attacking third. That rose up to 1.02 per 90 after not being above 0.63 in either of the previous three seasons. In fact, only Mohamed Salah and Richarlison won more possessions in the final third than Kante last season. The difference, of course, is that Kante's a midfielder and they're both attackers. So when Kante wins possession that high up the field, there tend to be more players ahead of the ball and therefore more players capable of capitalizing on re-gaining the ball. Why is this important? Because winning possession in the attacking third has also been shown to correlate with, well, winning games.

On top of that, Kante found his way into the penalty area more than he ever has before. He took 73 touches in the box last season after taking 58 touches in the box in the previous three seasons combined.

Winning possession in the attacking third and finding space in the opposition box are two of the most valuable things that a player can do that don't directly involve taking or setting up a shot. Kante may have been playing a new role, but he did more of these things than ever before. He also set or matched career highs in shots, chances created, goals and assists.

However, the value that Kante provided in years past was that he did the defensive work of multiple players and allowed Claudio Ranieri and Conte to often only play two midfielders instead of the customary three you see employed by the best teams in the Premier League. That made room for another attacker: basically another player who could win the ball back in the attacking third and make runs into the box.

The running joke at Leicester was that they "played three in midfield, Drinkwater in the middle with Kante either side." At Chelsea, Eden Hazard said of Kante's omnipresence: "I think I'm playing with twins."

Those teams were at least partially built around Kante's unique skills. His value wasn't necessarily in what he did; it came from everything he allowed his teammates and managers to do. But it's still not clear that one role is better than the other.

In Conte's last season with Chelsea, with Kante in his supposedly more natural role, they won 70 points on a plus-24 goal differential. In Sarri's one season, with Kante playing higher up the field, Chelsea won 72 points on a plus-24 goal differential. In terms of their expected goal differential last season, Chelsea were the clear third-best team behind Manchester City and Liverpool. Kante has only started three Premier League games so far but it seems like we're in for more of the same this season.

Kante hasn't really played enough to define his role within Lampard's preferred system, but early signs suggest that he's going to be used as something of a hybrid; his number of possessions won in midfield is up from last year, but he's already scored two goals. And so, perhaps the truth of the Kante conundrum is that he doesn't have a best position. He's never been a pure holding midfielder. He's never been a box-to-box runner either. He's always been somewhere in between but, in playing either role, he's proven to be indispensable.

Remember the year Leicester won the title? Kante won the third-most possessions in the midfield among all Premier League players. His teammate, Danny Drinkwater, was No. 1.