In Defense of Long Shots
With analytics guiding many of our modern football conversations, certain metrics have had incontrovertible effects on how we view aspects of the game. While gatekeepers of orthodox footballing practices will demand for players to shoot as often as possible, analytics, and its poster boy, expected goals (xG), has subverted that traditional wisdom for greater selectivity in shots. With a statistically tempered approach, shot locations have skewed from speculative to sensible.
This is all well and good, I work with data myself and am obviously a big believer in this approach — but I also think we’ve pushed ourselves too far into an extreme. By ‘us’ I mean the analytics community and anyone else pushing forward this application of xG. With this article I wanted to put together a more balanced argument in favor of longshots’ sporadic usefulness, something in between the two extremes of Merson/Souness and… whichever name represents analytics for you.
I’ve been wanting to write this one for awhile now, initially approaching Statsbomb’s Euan Dewar with this idea well over a year ago. In the end, I put it off. This being mostly an argument based on play-by-play decision making, there still isn’t a great amount of data to back-up my point and there won’t be until position data is more commonplace. Instead, I’ve resigned to having a long debate on the topic with our friend Nathan Ashley Clark, before writing the very subjective article you see here on the thoughts we came up with.
The players (1000+ minutes) who shot from range the most times last season in the Premier League (anyone with over 0.6 outside the box shots per 90), hit the target 26.4% of the time. This figure is above average when compared to that of all the players to have taken a single long-shot, which helps push the idea that players who take more long-shots do so because they’re better at them.
The Premier League’s elite range shooters from last season (Anthony Martial, Marcus Rashford, Riyad Mahrez) hit the target slightly over 50% of the time. Obviously the better at shooting from distance you are, the more that should weigh into your decision making habits in the final third. On top of this, long shooting isn’t simply about scoring success. Taking into account the aforementioned 26% success rate, this means one in every four long-shots has a solid chance of gaining a lot of extra value through rebounds and set-pieces that may come from them.
Almost a third of all goals (31%) in the main European leagues (top 5 + Portugal) come from set-pieces and corners make up the largest chunk of that (12.3%). Set-pieces are incredibly important because, taking into account the fact that they’re repeatable situations, they can be practiced to a much higher degree than the wider events of a game which is full of unpredictable actions.
A shot from outside the box giving rise to a corner, particularly if you’re a team that puts a lot of work into set-pieces, dramatically increases the value of said shot. One could even make the argument that those teams could utilize an above average long-range shooter, encouraging him to try his luck specifically to increase the likelihood of dead ball situations.
From a macro perspective, these types of shots are also a useful means to force your opposition to defend outside of their own box. Using them to keep your offensive approach interchangeable and even as a decoy for other attacking moves is extremely important for a dominant side that spends most of the time in their attacking phase and needs to find diverse routes towards goal.
From a more defensive standpoint, we also see a lot of clearance-type long shots which initially appear as almost mindless strikes usually after a set-piece. However, their primary purpose is to prevent a possible counter-attack while your side is in a disadvantageous position and not necessarily to score. The compilation of Ander Herrera shots below is legitimately very funny but also includes a ton of these situations, since Herrera was tasked with this role for a long portion of his United career.
Another argument in defense of long shots is that sometimes you are simply a lot worse than your opponent and, despite a shot from a less than advantageous position theoretically being a bad call, you just have to try your luck because you may not have the capacity to out-play the opposition into giving you a higher quality opportunity. If Andros Townsend doesn’t decide to take that shot against City, Crystal Palace would not have taken 3 points from the Etihad in December.
Finally, the most subjective argument of all: long shots are just so much fun. I grew up thinking they were the best part of football and now I tweet about how they represent poor decision-making. The truth is that every fan who bought a ticket is still going to be ready to shout ‘shoot’ at their midfielder whenever he gets the smallest bit of space in front of him despite the increasing influence of analytics in the sport — and there might be a good thing in that. The excitement that comes from the entertainment value of a play (even if that is Marcos Rojo ignoring seven passing options to take a shot that goes a bit over the bar), really can change the momentum of a game.
The rise of Expected Goals within football discourse has helped clarify long shooting for us. In general, there should be fewer long shots in football and we’re moving towards an era where their volume is reduced. As a rule, looking for passing options that could help you progress into a higher quality chance is almost always the better option.
However, long shooting is something that has to be considered within the broader context of decision making. The value of a long shot does not arise within the context of immediate goal scoring: when considered within the immediate context of “will I score from this shot” they’re obviously bad, but the various contexts in which they arise are much more varied than an xG model allows. As such, they should not, nor will they ever be, dismissed from our game.
