Heading for a Fall - From the Archives 2/24/17

It appears that the Sportdec blog, where I initially posted this, has vanished into the ether. So I figured I'd repost this here.

Southampton seem to be the forgotten team in the Premier League this year. Though the title is pretty much Chelsea’s already, there is a very competitive race for the top four positions and an intriguing relegation battle as well, so those few teams in the middle are going under the radar. That said, despite a new manager in Claude Puel and the yearly sale of half the roster to Liverpool, the Saints are once again playing decent football. Though they are currently just 13^th in the table, they have the 9^th best Goal Difference (in a tie with Stoke), the 7^th best Shot Difference, and the 6^th best Shot on Target Difference. Those numbers tend to closely track league position - after all, teams can’t win unless they score more goals than their opponent and they can’t score goals without taking shots - so it’s a little surprising to see Southampton as far down as they are. It seems they are being let down by finishing at both ends of the pitch, with their own players seemingly unable to convert basic chances and opposition players channeling Lionel Messi. Fraser Forster’s foibles have been covered elsewhere, but less has been written about Southampton’s own finishing woes, which is odd since Southampton have an unusual attacking profile which might help explain some of it.

How bad has Southampton’s finishing been this season? They are dead last in converting non-penalty Shots into Goals, at just 6%. If you just look at how frequently they convert non-penalty Shots on Target, they’re still last at 18%. Big Chance conversion is 32%, last yet again. If you look at Expected Goals (xG), a statistic which accounts for shot location and type to give a scoring probability to each shot, they are scoring about nine goals fewer than we would have expected through their first 21 matches. The interesting thing is that Southampton’s finishing was decidedly above average last season, ranking 7^th in conversion of Shots, 2^nd in Shots on Target, and scoring two goals more than their xG would suggest. While finishing is something that tends to be quite variable, the way Saints attack most likely adds to this effect.

One of the things that defines Southampton’s attack is their reliance on crosses. Southampton average 27 crosses a game, more than every team except Manchester United. This constitutes 5% of their total passes, quite high for a team which sees as much of the ball as they do. Low possession teams such as West Brom and Crystal Palace use a lot of crosses in attack, but no team averaging more than 50% possession uses them more frequently than the Saints. As you might expect, the emphasis on crossing is reflected in the number of headed shots Southampton take, where they again rank 2^nd in the league. Last season, this led to a league-leading 15 headed goals, but this season they have a league-low three. This gap, while surprising on its own, is even more amazing when you consider that Southampton have both increased the number of crosses per game this season and averaged more headed shots per game. They converted 15% of their headed shots into goals last season; this season they have converted just 4%. If their prior conversion rate had held this season, Southampton would have scored seven more goals.

There are several possible explanations for this drop-off. Personnel turnover might play a role, particularly Graziano Pelle’s departure as he was one of their prime aerial targets with five headed goals last season. However, Charlie Austin was theoretically an excellent replacement, as in seasons where he has played at least 1000 Premier League minutes he never scored fewer than four headed goals. This season, he has just one in 870 minutes. Shane Long (5 headed goals last season and just one this season) and the now-departed Jose Fonte have also seen their conversion rates fall off. It’s possible that the quality of cross delivery has declined, although that seems unlikely given the top crossers (Tadic, Soares, Ward-Prowse, and Bertrand) were there last season as well. As stated above, finishing does tend to be highly variable, so a large part of it could simply be luck, as unsatisfying as that answer is. Headers tend to be scored less frequently than other types of shots, so an attack geared towards generating them is perhaps more likely to be subject to this kind of drought. Southampton have created the 6^th fewest Big Chances as defined by Opta, suggesting that their attack does tend to produce a lot of shots that aren’t particularly threatening. Now that they’re no longer converting headers at such a fine rate, that deficiency becomes much more glaring.

Southampton certainly seem to be aware that finishing is a problem, bringing in Manolo Gabbiadini in at the striker position in January. His first goal seems to suggest that he might help improve conversion rates, with a vicious left-footed strike into the roof of the net from a tight angle. However, his three missed headers in the same game tell another side to the story: how Southampton’s pattern of attack contributes to poor conversion. The Saints really need to rely less on the cross if they’re to march back into the top half of the table.