At Stats Perform, we wondered if the long lay off due to the coronavirus might see an impact on some teams as they struggle to get back into the swing of things. Will the Premier League throw up some surprise results like the Bundesliga or will it be business as usual?
Empty stadiums and the unexpected hiatus have seen the German league restart with a significantly lower level of home wins, just 20%, compared to the usual rate of 45% over the last decade. Looking at the table below, the start of the Premier League seems to throw up more unusual results than other competitions normally do.
The start to the football season brings anticipation and hope, but also it brings the unknown, nerves and often surprise results. Integrating new signings, players trying to get up to full fitness, plus encountering new opponents on the pitch and in the dugout often contribute to some unusual outcomes.
Of course, a mid-season break is a different situation to the start of a campaign, with no new players to embed, while teams are already fighting to avoid relegation or qualify for Europe with fewer games to offset bad results. But the long layoff will have had an impact on fitness and cohesion. With that said, we wondered which teams might hit the ground running or struggle out of the gates as the Premier League gets underway again.
When it comes to games played in August over the last 10 seasons in the Premier League, most of the Big Six seem to get out of the blocks smoothly. But it’s a different story in north London where Tottenham and Arsenal perform far worse than they do during the seasons as whole, with their win percentage dropping significantly.
When you look at those figures in terms of uplift, then Aston Villa move to the top of the class as far as being quick off the mark, but again, Chelsea, Liverpool, Manchester United and Manchester City definitely seem prepared.
At the other end of the scale, Burnley and Southampton are particularly slow starters, but in recent seasons have had to pick up their form later in the campaign. Again, though, both Arsenal and Tottenham have struggled.
It’s a similar story with points per game.
The start of the season has been fruitful for Wolves and Aston Villa, while Burnley and West Ham join Arsenal as those teams who see the biggest decreases proportionally in the points per game accrued in the early throes of competitive action.
In Arsenal’s case, Arsene Wenger’s team used to do all their preparation close to home, with a training camp in Austria and local fixtures for their friendlies, but in 2011 they followed most other clubs in heading off to far flung locations to play high-profile friendlies all over the world. Since then, their opening few weeks have been sluggish to say the least. In the last decade, the Gunners have won just 38% of games in August, while their win percentage in the 2000s decade for the same month was 55%.
Perhaps with lockdown keeping the team close to home, Mikel Arteta’s team might find it easier to get his team up to speed, but their first match away at Manchester City is about as tough as they come.
With Tottenham, their fate seems closely tied to the form of Harry Kane and the England captain was a notoriously slow starter with no Premier League goals in August until 2018. He has now netted four thanks to a brace against Aston Villa earlier this season but that tally is somewhat short of his expected goals of almost seven (6.8). It’s a far cry from his 129 goals from an total xG of 94.8 during the rest of the year, where his match sharpness sees him score far more than the average striker would normally convert.
Since 2014, Kane seems to take a while to get up to full speed – if you take the first three fixtures after any break of more than three weeks due to injury or off season, he’s only netted a goal every 2.6 games compared to once every 1.3 games when match fit.
To see how the Premier League regulars perform each month, check out the graphic below from @petermckeever.
Enjoy this? Subscribe to The Analyst to receive five stories each Friday from Stats Perform.