Premier League 2016/17 review: Team statistics

Date published: Tuesday 30th May 2017 11:44

In our first team stats roundup, we looked at defensive stats and started on the attacking side. Today we’ll finish the team section with more unusual attacking numbers…

A stat that for some reason always fascinates me is percentage of attacks sorted by field position: left, right, or centre. The most left-sided attacking team this year was Watford, at 40%, who passed Tottenham and Manchester United in the latter part of the season because of M’Baye Niang. The most right-sided attacking team was Swansea City, at 42%. Neither of these numbers is exceptional, but it’s notable that Swansea have been among the leaders in right-sided attacking in five of their six years in the Premier League.

But we do have something exceptional in shots position, as opposed to attacks position. If you think Manchester United didn’t shoot from the right side much this year, you’re (pun intended) right. Only 9% of their shots came from the right, which is not only 5% less than anyone else this year, but the only time any team has been below 10% from either side in the eight years these stats have been kept.

Another fairly remarkable position-related stat comes from Middlesbrough. In the second game of the season, Cristhian Stuani scored a nifty strike from outside the area. It was one to cherish, because it would be Boro’s one and only goal from outside the box, matching Stoke City’s low total from 2012/13. (And since it was against Sunderland, maybe it should only count half.)

So far we’ve looked at stats on where shots start; now for where they finish. Squawka, taking its stats from Opta, divides the goal into six zones: upper left, centre, right, and lower left, centre, right. It makes for some interesting number crunching. Overall, teams scored 78.2% into the bottom half of the net, and 21.8% into the top half. This needs no explanation.

But what does need explanation is the way these goals are skewed to particular sides, and opposite sides at that. Fifteen of the 20 teams scored more bottom-half goals to the left.  Conversely, 14 of the 20 teams scored more top-half goals to the right. The same biases existed last season as well, although the effect was a little less pronounced. What do we make of this?

Without a close analysis of tapes and shot charts, we can’t be sure. My instinct is that it has something to do with the preponderance of right-footed shooters. We see more right-footed curlers, which usually go into the top right, and more right-footed low cross/pulled shots, which naturally go into the bottom left. On the other hand, Everton and Middlesbrough, the two sides to have dominant left-footed strikers, finish more in the bottom left, Everton dramatically so. So maybe my theory is wrong.

In any case, the team splits for high/low offer some food for thought. Only one team in the top seven, Manchester United, scored more than the league average of high goals. That’s probably because the better sides get more shots closer in, which tend to be low shots. They also have players better at guiding the ball, and guided shots are more likely to be low.

Still, it’s not one of the top seven but West Ham who had the greatest percentage of lower-half goals, with 85.1%. Middlesbrough and Swansea were tied for the lead in high goals with 33.3%, and I don’t think it’s a coincidence that Sunderland, another side in the relegation race, were significantly over average in high goals as well. Hull City finished about average.

Team dribbles usually track closely with the top of the table. This year the top seven teams in attempted dribbles include six of the top seven teams in the league, with Everton replaced by Bournemouth in fifth.

At the other end, though, you can have some fun, because the teams that dribble fewest generally have distinctive ways of going about the game. That means Burnley, of course, who never met a long aerial ball they didn’t like and were last in dribbles with 306. This isn’t the lowest recorded (Norwich City had 274 back in 2011/12), but  it was a stunning 113 behind 19th place Hull City, and 182 behind 18th place West Bromwich Albion. The Hull case is interesting too, because they weren’t really a long ball team. Under both Mike Phelan and Marco Silva, they preferred the short pass, and preferred passes of any kind to dribbling.

Dribble percentages provide a few interesting notes this year as well. While Arsenal and Chelsea were the best percentage dribblers, Middlesbrough, with Adama ‘Tornado’ Traoré, crashed the party at third.  Spurs, top of the league in dribbles attempted, were only 18th in dribble percentage.

While we’re talking about the ball at the feet, we can go to stats on possession loss. Which side had the most unsuccessful touches, meaning failures to control the ball? Stoke City, which is quite a surprise, since they’re a decent technical side. So who was the best side for control? Hull City, another surprise, although under Phelan they made so many simple short passes that control was probably fairly easy.

The companion stat is times dispossessed by opponents. The best teams in this category are teams like Burnley and Leicester that get rid of the ball quickly, and so can’t get dispossessed. The ones who get dispossessed most tend to be top teams with lots of possession and lots of attacks into danger areas. Tottenham were the most dispossessed team the previous year and the year before. But number one on the list this year, dispossessed the most times, is Crystal Palace. And now you know why Big Sam left.

Speaking of world-class managers, Pep Guardiola didn’t have the happiest of first seasons in England. But his effect can be seen in passing stats. Manchester City made 22,709 passes last year, which is not only a boatload of passes, but the most ever in the league since the stat has been recorded. Jürgen Klopp, in his first full year at Liverpool, registered 22,282 passes, the second highest total ever.

Neither one is Barcelona yet, though:  In 2012/13, a Pep-led side recorded 28,591 passes, more than ten thousand (!) more than second-place Real Madrid. To give you an idea how ridiculous that is, if you subtract 10,888 from Manchester City’s total this year, you get Tony Pulis’ West Bromwich Albion, who were dead last in that category.

Crosses, particularly of the aerial variety, are a fairly inefficient method of attack, but that doesn’t stop teams from using them. Southampton are the biggest abusers here, with 947 crosses (which includes crosses both high and low), fully 74 more than second-place Crystal Palace. At least the Saints managed an average completion rate, at 22.1%. Best at cross success were Burnley at 26.8%, followed by Swansea and Crystal Palace. All three had a big man to shoot for. At the bottom were Bournemouth at 18.7%, who didn’t.

On a related note, who were the most and least accurate at corners? Because the stat doesn’t differentiate between short corners and regular corners, it’s biased toward teams who take a fair amount of short corners. Chelsea and Bournemouth, two teams who fit that description, are first and second in this department. But most teams take relatively few short corners, so we can pin the tail on the worst corner-takers:  Southampton, who had only 32.8% of corners find a teammate. Manchester United were second worst at 35.0%, then Sunderland at 35.4%. (Remember too that Sunderland were last in set piece goals and headed goals.)

Let’s talk assists now. It’s a stat you hear a lot of where individual players are concerned, but almost never in a team context. Plenty of goals are scored without assists, and percentage of assisted goals tells you a lot about how a team’s attack functions, or doesn’t. For example, Sunderland scored two more goals than Middlesbrough this year, but had nine fewer assists. In fact, the Black Cats had only 41.4% of their goals assisted, by several miles the lowest in the league. But Boro had a very high 77.8% of their goals assisted, nosing out Manchester United with 75.9%. I’m getting lazy in my late middle age, and haven’t checked whether these are records, but they’re probably close.

Assists direct from corners are a nice clean stat, and you know West Brom will be near the top—in fact, ridiculously so. The Baggies registered 10 assists direct from corner kicks, the highest ever recorded. Middlesbrough and West Ham were in another time zone, tied for second with four. Leicester, Sunderland, and Bournemouth all got blanked in this department, but remember this only covers assists credited to the corner-taker himself. A flick-on or complex play gives the assist to someone else.

Assists from free kicks are a comparable stat, and it will surprise nobody that Swansea and Spurs, with Gylfi Sigurdsson and Christian Eriksen delivering the goods, finished first and second in this category. Three sides wound up with zero: Liverpool, Manchester United, and Stoke City.

The flip side is the assist from the throughball, which requires a different set of skills. Chelsea lead the pack with 10, well ahead of Manchester United’s seven. Stoke find themselves at the bottom here too, with zero.

One last assist stat. We’ve talked about West Brom’s proclivities several times, and here’s a stat that sums it up. The Baggies registered 34 assists, 26 of which came from crosses or corners, at 76.5%. That’s the third straight year West Brom have been over 70% in that category. The record, incidentally, is held by Cardiff City, who went down in 2013/14 with 82.6% of their assists from crosses and corners.

We’ll end with a team stat which says a lot about the league as a whole. All season we’ve seen a huge gap between the haves and the have-nots, and this is most clearly reflected in goal differences. Take a look at the final table, and you’ll see the chasm between the worst plus and best minus, Everton’s +18 and Southampton’s -7. Everton’s +18 is an all-time top flight record high for worst plus, breaking the old record of +15. And the 25-point gap between plus and minus is by far the largest in top flight history, in fact the only one above 20. The trend in goal differences has been more or less in this direction, and we’re likely to see decent-sized gaps in the years to come. But 2016/17 smashed the records.

Peter Goldstein

More Related Articles