Tracking data is the next frontier in soccer analytics. Now that we have access to the location of every player on the field updated 25 times per second, we can measure things like off-ball runs and defensive positioning that were mostly invisible when using only event data. Tracking data holds the potential to unlock the game’s secrets. This article is not going to solve soccer. It is going to solve something arguably even more important: who the best teammates are based on their goal celebrations.

When Metrica released their first few games of anonymized tracking data over a year ago, I first dreamed of using tracking data to revolutionize how we look at goal celebrations. However, with just a handful of games available, and unknown players and teams, this dream would have to wait. But now, thanks to Major League Soccer and Second Spectrum, American Soccer Analysis has access to tracking data from the last couple of years of MLS play and the dream of measuring player celebrations can now be realized.

The sheer size of the data is daunting. Each game has about three million rows of information, on its own, enough to turn a six year old MacBook Air into a searing hot chunk of aluminum choking on the firehose of data firing off its cooling fans in vain. Luckily, in order to determine the best goal celebrators, we don’t need all of this from every game. I restricted my analysis to the 20 seconds after each non-penalty goal scored so far in the 2021 MLS season to date. This way we are looking at a much more palatable 2.1 million rows, something that even an aging laptop can handle without melting into slag or forever spinning the pizza of death.

I’m going to focus on a few simple key performance indicators (KPIs) to measure celebration performances in MLS for individuals and as a team:

Maximum speed attained during the celebration period
Minimum distance to the goal scorer
Distance travelled to reach the goal scorer

The goal with the highest team median maximum celebration speed was Cecilio Domínguez’s game winner in the 67th minute in Colorado (4:55 mark of the video to the right) that gave Austin the first lead in their history. After scoring, Domínguez wheeled off toward the corner flag followed by his teammates reaching a median top speed of 14.8 mph. Nick Lima, having started out almost 50 yards away, reached a maximum speed of 16.5 mph to reach his compatriot. The only Austin players not to participate were goalkeeper Brad Stuver and center back Jhohan Romaña. Stuver gets a pass as the keeper, but Romaña just ambles around half-field, never approaching the winger who he previously played with at Club Guaraní prior to arriving in Austin.

But maybe celebrating just isn’t Jhohan Romaña’s thing. Typically, the farther a player is from the goal scorer, the less likely they are to go celebrate. However, of the 373 MLS players on the field for at least five goals scored to date, Romaña is an outlier in the amount of social distancing he leaves with his goal scoring teammate. At 55.7 yards, Romaña’s median distance from the center of celebration is over 12 yards farther away than Minnesota’s Brent Kallman and Orlando center back and goalkeeper Rodrigo Schlegal. Kallman and Schlegal occasionally join in the celebrations, while in his five goal celebration opportunities this season Romaña has not.

Austin also has one of the best celebrators in the league in Matt Besler. Unlike his occasional center back partner, the veteran’s median distance traveled to celebrate a goal is 45.6 yards, second only to Nashville’s Jalil Anibaba’s 50.38 yards. The distance between a player and the goal scorer is weakly correlated with their maximum speed and, with their long travel distances, Besler and Anibaba attain the 6th and 7th highest median top speeds during celebrations. This holds true for MLS legend Nouhou Tolo. His long runway allows Nouhou to attain a median maximum speed of 14.5 mph. The fastest so far this season is Columbus winger Luis Díaz who reaches a median top speed of 14.8 mph. Díaz is notable as he starts out much closer to the goal scorer than most of the other goal celebration speedsters.

These simple KPIs are only scratching the surface of what can be done with tracking data to determine the best celebrators, and therefore best teammates. There are obvious limitations, sometimes a player will be called over by a coach to confer during a celebration, limiting their opportunity to join in. Late game goals for a trailing team often result in muted or non-existent post-goal celebrations. Perhaps Jhohan Romaña and others remember the 2018 World Cup where multiple teams misinterpreted a restart rule and left a player on the field lest the other team take a quick restart and score (I checked with MLSRefStats to make sure this wasn’t a thing) and just wants to be extra careful. I look forward to others picking up the baton on this type of analysis, imagine applying ghosting to compare a player’s movement to an optimal celebration. The possibilities are endless.