A Different Approach to the MLS MVP Race

What does it really mean to be MVP?

What’s the best way to determine the MVP of a season?

It is not an easy question, and there’s a lot of different approaches. Some people simply look at the goals and assists table - after all, goals win games, and the points earned there make the difference on the table at the end of the season.

With expected value numbers, we can look at what “should” have happened. Who should have scored more, who’s out over their skis, etc. We can also take into account the performance that happens when shots are not being taken. After all, it’s a little unfair to let the attacking players have all the glory; why can’t a defensive midfielder be the most valuable player? For that reason, many of the more data friendly pundits and writers will have expected value models contribute to their MVP rankings.

And then, of course, there’s all sorts of esoteric beliefs that writers and voters hold. For example, a player needs to play a certain percentage of their team’s minutes. Maybe the most valuable player needs to be a team leader as well. There’s more ways to dissect what value means than there are players on the field at any given game.

I decided to come up with a quick and easy metric to evaluate who I think the leading MVP candidates are. Not the “true” most valuable players by a specific model; not the ones whose underlying numbers are so much better than everyone else. Instead, it’s a combination of expected and actual performances.

The Breakdown:

It’s a simple formula, with each part mattering:

First, I look at the percentile rank of a player’s individual performance in a game vs the typical at their position and home/away. This gives me an idea of where each individual performance stands. If a player’s performance is 95th %ile or above, that’s one (1) MVP Point. This means that any decent player is going to earn points regardless of position.

Second, I look at goals. While xG is more predictive of future performances, there’s no question that goals are the most descriptive measure of a team’s performance. If a player is scoring a lot of goals on fewer xG, they’re having a good season, and a much better one than the player who’s situation is reversed. If you want to talk about transfer value or predictive power, those discussions are more important for future games. Each goal gets one MVP Point.

Another aspect of goals is when they are scored. Alexi Lalas (of all people) had a good idea that I kind of like. He votes for the player who scores the most goals but only counts one goal per game. In his mind, more consistent scoring is more valuable than concentrated scoring. I tend to believe that, so for each 1st goal gets a bonus MVP Point.

Finally, I do take into account assists. Not second assists - the pass before the pass is more than adequately counted in the G+ performance measure. At the same time, creating a goal from your pass is still important, and players with a lot of assists are having a good season. If that’s because they’re passing to great finishers, it still comes out in the wash.

With that in mind, here’s the current MVP ranking for the season:

No real surprise to see Messi at the top of the list, though I feel like the media has to pretend that it’s a closer race than it is. Sam Surridge and Anders Dreyer are number 2. Surridge is a big time scorer, but he’s not as “in the discussion” as these other players - at least from my general impression. Dreyer has made a big splash this season but the loss of Milan Iloski, who he’s been assisting, may impact him down the stretch.

Rounding out the top 5 are Evander and Denis Bouanga. Bouanga had a quiet start to the season but is hitting peak form, and may end up right next to Messi if he continues at his current pace. Evander is someone who gets a lot of noise, but doesn’t always look quite as good on the underlying numbers. It will be interesting to see if he goes up or down.

MVP Points Through History

Now, how does this measure compare historically to voting? We have G+ data going back to 2013, so we can start from there:

2013: Mike Magee wins while also first in MVP Points with 46. Other finalists: Robbie Keane was 3rd in points, and Marco Di Vaio was 4th.

2014: Robbie Keane takes it this year, but was actually second. Bradley Wright Phillips edged him by virtue of 16 points from goals but was not even a finalist. Lee Nguyen and Obafemi Martins, the other finalists, were 3rd and 4th respectively.

2015: Sebastian Giovinco takes the award this year, but is 2nd to Kei Kamara on MVP Points. Like last season, a close match goes to the more well rounded player. Kamara is a finalist, along with Benny Feilhaber. The voters were being nice to Feilhaber who did not land in the top 15. BWP is 3rd this season.

2016: David Villa - Giovinco is actually first this season, but I think that voters don’t like going back to back. Piatti is second and Villa 3rd, but neither Giovinco or Piatti are finalists. Instead, it’s BWP (4th) and Klestjan (14th - a trend of being nice to Americans)

2017: Diego Valeri - much like last season, Villa is the clear leader in MVP Points but fails to go back to back. Valeri is second, so it’s not a robbery. Nemanja Nikolic is 3rd in points and rounds out the finalists.

2018: Josef Martinez was undeniable here with a massive 18 point lead over BWP in 2nd place. He won easily with Almiron and Zlatan the other finalists. Zlatan was 3rd but on ~1000 fewer minutes.

2019: Carlos Vela has the best individual season in MLS history, racking up an unreal 91 MVP points (double that of some previous winners!). He wins MVP with Zlatan and Josef Martinez taking 2nd and 3rd both in Points and voting.

2020: The weirdest season - Pozuelo is the MVP, but only 6th in MVP points. The top 5 are:

• Diego Rossi

• Gyasi Zardes

• Jordan Morris

• Raul Ruidiaz

• Chris Mueller.

That’s a very odd group of players. It’s strange that voters did not take the opportunity to go with an American here, when they were in striking distance of the top. COVID really did a number on the season, and the MVP award probably should have just been canceled like they tried to do with the Supporter’s Shield.

2021: Another odd season where the Revolution set the points record despite not being all that great of a team. Carles Gil won the MVP award - really more of a team trophy at that point - despite being a shocking 27th on the MVP Points list. He is far and away the lowest ranked winner of the trophy; no one else is outside the top 6. The actual top, Taty Castellanos, walked away from the season with a Golden Boot and MLS Cup trophy. I’m sure he’s not complaining. Daniel Salloi and Hany Mukhtar were MVP Finalists, and in the top 3 Points earners.

2022: A three way tie at 54 points sees Mukhtar declared the MVP (getting a bonus from previous seasons) while accompanied by Sebastian Driussi and Daniel Gazdag. Driussi was a finalist, Gazdag was not, as his teammate Andre Blake was there instead.

2023: Luciano Acosta wins the award and tops the group in points, though it’s close; Denis Bouanga is just a point behind and Cucho Hernandez is just 3 back. Bouanga will be a finalist, but the other spot goes to 4th place Thiago Almada.

2024: Lionel Messi wins going away, but he’s not in first place. He’s fourth! The actual leader is LA Galaxy talisman… Gabriel Pec. Didn’t expect that? Neither did I. Puig is actually third, and the dominance of both may have led to them vulturing votes away from each other. Denis Bouanga is second, not even a finalist this time.

There you have it - historically, MVP Points have tracked very closely with the actual winners of the MVP trophy. If history is any indication, the MLS MVP will be whoever comes in second behind Messi (since he got his last year, before he was truly in the lead).

Anyone Else?

It’s also interesting to use the MVP Points framework for other awards - for example, Alexander Freeman leads the league for 22U22 eligible players with 16. Gerardo Valenzuela has 11 points on just 688 minutes. These players might be in the running for the top spot when those awards are released at the end of the season. Quinn Sullivan and Owen Wolff are in the running if you include previous years.

As a metric, I like this one; it tracks closely with both what the numbers say, and general public opinion around players.