This article is part of our infrequent Replication Project series where we take an important analytics finding from yesteryear and see if it still holds up with modern data. You can find other articles in the series here.

Way back in 2011 a somewhat remarkable thing occurred in the nascent soccer analytics blogosphere. On May 20, Chris Anderson posted the provocatively titled The Uselessness of Free Kicks in the Premier League, which argued that since free kicks are rare and not often converted they are “not particularly effective devices for scoring in the Premier League.” A few weeks later, Anderson lent space on his own blog, Soccer by the Numbers, to Ian Graham, now the outgoing Director of Research at Liverpool but then at Decision Technology, to argue the opposite. “Why Shots From Free Kicks Are A Good Idea, Or At Least Not A Bad One” showed that while free kick shots rarely result in a goal, the correct comparison is to other shots outside of the box. In that comparison, free kick shots are about twice as valuable as open play shots from outside the box. So with an additional decade of data, have things changed since then using MLS data and who was right?

Note: go listen to the fantastic Post Script podcast, by ASA’s own Tiotal Football and ASA Alumnus John Muller that chronicles the history of soccer analytics (and tactics) blogging which motivated this post.

What’s changed?

First off, how much is different between the 2010/11 Premier League season that Anderson (the first half) and Graham used for their analyses and in the TAM era of MLS (2015-2022)? Turns out not all that much.

Shots from free kicks for First half of the 2010/2011 Premier League Season and all mls matches from 2015-2022.

Like in 2010/11, the last eight years of MLS data show that a large chunk of games do not feature a single free kick shot, almost exactly one-third. Furthermore, about 70% of games have one shot or fewer. Overall, the exact percentages are slightly different, but let’s chalk that up to Anderson having fewer than 200 games in the data set versus the 4,395 in the MLS one. The fact remains that direct free kicks shots are rare. Anderson also reported that only 3% of goals come from direct shots from free kicks. If anything, that is overstating the prevalence of free kick goals, as 2.3% of goals came from direct free kick shots in MLS.

Anderson spends most of his article chronicling how the Premier League teams fared with free kick shots during the first half of the 2010/11 season. TL;DR: not well. Most teams rarely put free kick shots on target and almost never score. This again holds up in a larger sample size from MLS. Across 153 team seasons, 36% of the time a team scores zero free kick goals directly off a shot, and 70% if the time they score one or fewer. Unless your team has Sebastian Giovinco or Lucas Zelarayán, you are not scoring more than four in a season.

Per Graham’s definition, Normal shots = all shots that are not Free Kicks.

In Ian Graham’s defense of the free kick, he uses a couple different plots to show that direct free kicks are still a valuable commodity. First, Graham shows that the correct comparison for a free kick shot is not all shots, but shots outside the box. Indeed, direct free kicks and all other shots outside the box are on average approximately the same distance from the goal line, both back in 2010/11 and in MLS from 2015-22. Next, he calculated the goal conversion rate and showed that free kick shots were converted at rates roughly twice that of shots taken outside the box, which again remains true. Again the absolute numbers are a little different, but one season of data is noisy so small differences are expected.

So who was right? And are direct free kick shots good or not?

Like many headlines, “The Uselessness of Free Kicks in the Premier League” is probably a bit more inflammatory than the actual content of the article. Anderson’s conclusion is: “[A direct free kick’s] relative infrequency, coupled with dodgy odds of turning them into goals hopefully have done one of two things: either they have convinced you that they're not particularly effective devices for scoring in the Premier League, or at least they will make you think twice next time you get excited when your team is awarded a free kick not too far from the penalty box.” This isn’t quite the uselessness advertised in the headline. In 2010/11 and today, free kicks are not “effective devices for scoring.” If you are expecting your team to score lots of goals directly from free kicks, you are not going to have a good time. So despite the headline, I’m going to call Chris Anderson’s overall conclusion as technically correct, which is the best kind of correct.

Importantly, Ian Graham emphasized that the correct comparison for direct free kick shots is not a hypothetical better shot, like one in the box, but shots taken outside the box. He concludes his article by writing: “4.5% of direct FKs are converted compared to only 2.6% of normal shots from outside the box. So, in an average match situation a player might nearly double his chances of scoring a goal by diving for a FK rather than taking a shot from open play.” This is unambiguously correct.

Interestingly, while Graham doesn’t say which data he used in the article, he was almost certainly using what today we’d call an Expected Possession Value (EPV) model: “While we were developing the Castrol Rankings, we looked at free kicks and came to the conclusion that winning a free kick increased a team's chance of scoring a goal.” This is the key point on whether a free kick is useless or not. While we don’t have the Castrol Index data, we do have our own EPV model, goals added (g+), for MLS.

With g+ we can see the value accrued from fouls suffered which by definition result in a free kick. Looking across the entire field, penalty box excluded, we see that suffering a foul increases your team’s chance of scoring more than your opponent’s everywhere except the most central areas of the field. Furthermore, getting fouled near the box, as you’d expect, is where the value is the highest. So, it’s clear that getting fouled, and earning a free kick, in the final third is good, at least on average.

So who is the best in MLS?

Finally, to be a bit self-indulgent, I decided to look back at my very first article at American Soccer Analysis, The Art of the Free Kick and How to Giovinco, that was published exactly five years ago. There, I showed that when a free kick is within 35 yards of goal, you should fire away to maximize the chances of scoring. That has not changed. But more importantly, we now have five more years of data to determine who is the best free kick taker in MLS.

Amazingly, Diego Fagundez has scored five free kick goals in only 25 attempts, blowing away the best free kick taker back in 2018, Didier Drogba (who also attempted 9.5 free kicks per goal in USL). Fagundez rarely attempts free kick shots, averaging only 0.13 per 96’ played, over 12x fewer per 96’ than the free shooting Sebastian Giovinco and less than half of any of the other top takers. Curiously, Fagundez has had three seasons since 2015 where he has attempted a single free kick shot - scoring on his only attempt in 2016 - and only attempted 10 in 2022, converting two. So, Josh Wolff, maybe let Diego be Diego on free kicks. Of note, Atlanta’s Thiago Almada, is currently at the Fagundez Line, also scoring a goal every five free kicks. So if Almada continues to score bangers, and isn’t sold before he reaches 25 free kicks, we may have a new free kick champion.

Conclusion

It is clear that earning a free kick in areas where a direct shot is possible is valuable in soccer. However, the rarity of free kicks and their low conversion rate make them unreliable (but not useless) as a means of generating goals. That said, it’s pretty handy to have a player that can score free kicks and perhaps grab a few bonus points over the course of the season. Direct free kicks, on average, produce higher goal rates than playing balls into the box, which perhaps changes if you’ve got a Julian Gressel. Maybe one day someone will look more granularly where shooting or serving is better. Regardless, I’ll be back in 2028 to check up on free kicks and what MLS players are the best at taking them.