Oxford and Defending Set Pieces: The problem wasn't truly solved last season, and it's holding them back this year

Oxford dominate the match possession, create better chances, and consistently pose the threat of running away with the match. And then conceded at a set piece. Sound familiar? 

It was United's undoing yet again at AFC Wimbledon, first letting Jack Rudoni header home from Anthony Hartigan's free kick to cancel out Sykes' first half opener, before allowing Nightingale to complete the turnaround from a corner. 

Rudoni's second of the game three minutes later following a smart cutback and finish capped the game and exposed another worrying Oxford capitulation, but the bottom line is: this match was won and lost on set pieces, and until Oxford are able to defend them, they are not going anywhere. 

It is only emphasized by the seeming strength of Oxford's set piece work last year. 

There seemed a period last season when Oxford were not just able to consistently defend them, but could also use their own corners and freekicks as prolific attacking weapons.

Oxford were excellent at attacking set pieces in League One last season

John Mousinho had cracked the problem, and Oxford were thriving because of it. 

Taking a look at stats via Opta's StatsPerform tool, however, shows that Oxford's defensive numbers from last season are not as strong as they may have seemed, and they have only grown worse this year. 

They conceded 15 goals from set piece scenarios (free kicks, corners, throw-ins etc..) which was only the thirteenth best record in League One and showed that they were actually pretty average at defending corners for a team which finished in the top six. 

In fact, no other team in the top six conceded more goals than Oxford from set pieces. 

Oxford were the worst team in the top six at defending set pieces.

The raw numbers, though, still show that they were considerably better than what they are producing this season. 

Oxford averaged 0.33 set piece goals conceded per game last year, roughly once every three matches, whereas this season, that number sits at 0.5 set piece goals per game, once every two games. 

Oxford's set piece goals conceded per game is comfortably higher than in the previous two seasons (Not including penalties).

Those numbers also neglect penalties, which have contributed three league goals this season. Whether a penalty kick counts as a set piece is an argument for another day, but if these are included, then the set pieces conceded per game rises from 0.5 to 1. 

75% of Oxford's league goals conceded this season have come from set pieces or penalties.

Delving deeper into the stats further underlines, however, that Oxford had never really solved the issue of defending set pieces, despite the better success last season. 

In 2020-21, Oxford conceded 171 shots from set piece situations, which was the seventh highest in the division. 

This season, they have already conceded 24 shots from these situations, the highest of any team in League One. It comes with the caveat of having played a match more than most other teams in the division, but further underlines a trend which seems to be going in the wrong direction for Oxford.

What is also notable is that Oxford are one of only four teams in the top twelve on the list of most shots conceded from set pieces to be in the top half of the table. All of the current top six place in the lowest eight teams for shots conceded from set pieces, indicating that this perhaps is what has kept Oxford apart from the top teams in the division. 

Equally worrying is Oxford's Expected Goals Conceded (XGA) from set pieces, a measure of how many goals a team can reasonably expect to concede from the chances they have allowed from set plays. It is the sixth highest in the division, indicating that they concede plenty of high quality chances from dead ball scenarios every game. 

Oxford are one of the poorest teams in League One at defending set plays by these metrics

Again, though, we can see that the numbers did not look very good last season either. 

They placed fourth for XG against from set plays at 17.5, which is actually higher than there final total of 15 (Opta counts the playoff matches in their statistics), indicating that United were lucky not to concede a few more goals from the chances they conceded at set plays last year, perhaps why we viewed their defending as more solid than it was.  

Despite conceding more goals from set plays this year, Oxford were actually expected to concede marginally more goals last season according to their SP XG.

This year, they have actually been a tad unfortunate, given that they have conceded three goals from an XGA from set pieces of two, but they are still clearly being held back by their issues when dealing with these plays. 

Oxford's XGA from open play is the tenth lowest in the division, at 3.5, indicating that, without their issues with set pieces, Oxford would be one of the stronger defensive sides in League One. 

This is backed up by Oxford's set play to open play XGA ratio of 0.26, which means that over 25% of Oxford's Expected Goals Against comes from set play scenarios. That is the fourth highest ratio of all teams in League One to have conceded five or more goals. 

Oxford have conceded a high percentage of goals from set plays (Teams who have conceded less than five goals distort the percentages)

Oxford are also struggling for penetration from their own set pieces this campaign.

Last season, Oxford recorded the fourth highest XG for from set pieces in League One. This year, they have had the fourth lowest. 

They are yet to score from a set play (not including penalties) in the league this season, and their XG from set pieces of 0.9 suggests that they have not warranted much either. 

Their open play to set piece XG for ratio is just 0.09, the joint worst in the division, and indicates that only 9% of Oxford's total XG this year has been created through any form of set piece. 

Oxford create an extremely low percentage of their chances from set pieces [Image via: https://dataviz.theanalyst.com/season-reviews/2021/?competition_id=11&season_id=2021]

Whilst the struggle to attack set pieces is not as imminent a problem, United's sudden drop off compared to last season, coupled with even worse results defensively, does suggest a more systematic problem for Oxford and set plays. 

Perhaps it is because Oxford are being beaten in the air far too frequently, with an aerial duel win percentage of just 48% this year, although it was only marginally higher at 51% last season. 

An easy solution to set piece problems is a lack of experience and knowhow, but this would appear to be irrelevant in Oxford's case. 

The average career appearances of every player in Oxford's squad is 77.5, the third highest in the division, behind Portsmouth (87.08) and Wigan (83.04). Between them, Oxford's squad have played 2,249 first team matches, the most in the division, and indicates that it is not as simple as saying that there is a lack of experience or backbone in the team. 

It is very difficult, therefore, to pinpoint exactly what is going wrong for Oxford with set pieces at the moment, but something is clearly not working,

The expected goals numbers suggest that United have been slightly unfortunate, and that perhaps we will see the number of goals Oxford concede even itself out as the season goes on to levels similar to last season. 

What the data tells us, though, is that last season's numbers were not as impressive as may have first seemed, and Oxford have in fact been one of the worst teams at defending them for a fair while now. 

In this case, it becomes a problem. 

Oxford cannot afford to lose games which they are otherwise controlling through cheap set pieces, and if they have any hope of bettering the sixth place finish they achieved last season, they will need to find a way to stop throwing away set piece goals in an extremely physical division.  

Many of the stats used during this article have come via Opta Analyst's interactive season statistics, which can be found here.