Chelsea 2019-20 Premier League Season Statistical Review

After nearly 12 months of ups and downs, with our very own Frank Lampard managing a team broadly comprised of our very own academy players, Chelsea have secured Champions League football for next season.

Going into the season, the “experts” didn’t give Chelsea much of a chance of finishing in the top four. Only 3 out of 21 pundits surveyed, of whom 2 were former Chelsea players, expected Chelsea to finish in the top four — shout out to former player/manager Ruud Gullit for his second place prediction; Proper Chels!

(Ed.note: and none of that is as amusing as most predicting sixth-place Spurs to finish third and eighth-place Arsenal to finish fourth!)

That said, thinking that Chelsea would struggle was not such a bold prediction. This author expected Chelsea (no pun intended) to finish sixth or thereabouts. About half of the WAGNH writers surveyed last season also expected Chelsea to finish outside the top four.

So how did the team pull this off? And are Chelsea better than last season? Let’s take a look.

---

Basic Numbers & Facts

Chelsea lost Eden Hazard, one of the five best footballers on the planet in June 2019, and replaced him with a bunch of 19-, 20-, and 21-year-olds.

The result? Chelsea became even better going forward. An xG of 2 per game across a season is an elite number by all accounts. While it is not as good as Manchester City’s, it is a thoroughly good tally regardless.

The total xG tally of 76.23 was good enough to rank second in the league behind Man City’s 102.21 (not a typo) and ahead of Liverpool’s 75.19. Interestingly, this is also the highest xG Chelsea have ever recorded in the available data-set and is the first time Chelsea have crossed 70 xG in a single season. The next best tally is from 2014-15, with Mourinho’s men putting up 68.64.

Photo by Chelsea Football Club/Chelsea FC via Getty Images

Defensively, Chelsea evidently have not been solid, but also have not been as bad as 54 goals conceded suggests. The truth is somewhere in between.

The next question then is, why did Chelsea struggle so much to get top four despite their excellent expected numbers? Bad finishing, an historically bad goalkeeper and some horrible injury luck all had their parts to play.

Realistically, the chances of a team of Chelsea’s resources scoring 7 goals fewer than expected while conceding 12 goals (conventional xG, not post-shot xG) more than expected are close to zero. Yet here we are.

In most cases, a team underperforming their expected numbers to this extent would not have had a chance of making the top four, but the crisis at Leicester offered a chance for Chelsea and Manchester United (another young team slightly underperforming their expected numbers) to sneak in.

Chelsea are unlikely to be this unlucky next season — [*touch wood*] — and there should be a bounce back towards the mean, with things improving dramatically at both ends.

---

Advanced Metrics

The first inference is that Chelsea were better at pressing this season than we were last season. The team allowed fewer opponent passes while also effectively bringing down the number and different types of shots conceded.

Maurizio Sarri’s principal strategy was to use the ball as the team’s best defender. While this meant Chelsea could be predicable and (supposedly) boring in quite a few games, there was a clear benefit in terms of defensive solidity.

Under Lampard, Chelsea seek to use the ball more directly. While this means the team has been more unpredictable going forward, it also meant being susceptible to getting exploited at the other end. This is reflected in the fact that the average quality of both shots taken and shots conceded has increased.

Sarri made it clear that Hazard was going to be the team’s main player from the first day. This resulted in Hazard receiving more of the ball than others and being expected to come up with more of the goods. To his credit, he did just that, directly contributing to exactly half of Chelsea’s league goals last season (16 goals, 15 assists).

Photo by Darren Walsh/Chelsea FC via Getty Images

Without Hazard, Lampard had to get more creative and eventually settled on a more adventurous style incorporating all of the team’s attackers. The absence of Hazard also meant that Lampard did not have to “carry” a player while setting up his team to press. As with everything in football and life, Hazard’s departure had its pros and cons.

However, like with most pressing teams at the early stage of their growth curve, Chelsea are not perfect. The average quality of shots conceded is pretty high for a team of Chelsea’s ambitions and this is not helped by the fact that opponents have been on an extremely hot finishing streak against Chelsea.

It is on Lampard to improve Chelsea’s solidity and ability to cut down opponent transitions. While the signs are highly positive thus far, plenty of work is still required — such as in, but certainly not limited to, defending set pieces.

(Please note that the counter-attack metric may not account for all fast-break transitions.)

---

Goalkeeping Metrics

No analysis of Chelsea can be complete without paying extra attention to the goalkeeping scenario. While Chelsea’s goalkeepers were merely poor last season, as evidenced by the 2.7% under-performance, they were truly terrible this season, under-performing expected averages by almost 10%. That’s shockingly bad.

Chelsea’s goalkeepers saved 9.5 goals fewer than the average goalkeeper would have based on fbref’s post-shot xG. For perspective, the next worst team were Everton, who conceded 4.8 goals more than expected. Chelsea almost doubled that number!

Photo by Clive Howes - Chelsea FC/Chelsea FC via Getty Images

Chelsea have now finished in the top four twice with a negative post-shot xG (-4.2 in 2018-19), joining Liverpool (-2.2) in 2017-18, though that’s as far back as we have this data available.

In terms of conventional xG, Chelsea conceded 12.91 goals more than expected according to Understat. The next worst? Claudio Bravo’s Manchester City, who conceded 9.39 goals more than expected. No other top-four team has conceded more than 5 goals above expected since 2014-15.

While Chelsea have conceded more, and better, chances this season than last, it is important to note that the team benefited from some subpar finishing from opponents last season too. This can be seen in how post-shot xG is lower than actual xG conceded for 2018-19.

One would think that with a more commanding and a more agile goalkeeper, Chelsea will improve even with the current set of defenders. Be it Onana, Pope or whoever else has been linked to Chelsea, most goalkeepers would represent a strong upgrade on Kepa and Caballero.

---

Next Season Expectations

Photo by MICHAEL REGAN/POOL/AFP via Getty Images

Next season, Chelsea can realistically aim for 80 points with the following additions:

• Kai Havertz
• 1 new goalkeeper
• 1 new left back
• Possibly a new centre-back
• While this is not an urgent requirement, possibly a new defensive-minded midfielder to play alongside or instead of Kanté when required would help too.

While the table did not reflect this, Chelsea played more like a 70-75 point team this season. Bad luck and variance meant Chelsea did not get the points commensurate with the underlying metrics on multiple occasions.

While talk of 90 points or more next season are premature due to the age profile and inexperience of the squad, there is no reason to believe Chelsea’s couldn’t finish at least third. And if Liverpool and/or Manchester City slip up, Chelsea must be ready and capable of capitalizing.

A few prayers to the Football Gods wouldn’t go amiss either. Chelsea’s underperformance at both ends was nothing short of absurd and the fact we still finished in the top-four was a mini miracle by itself.

---

Thank you for your reading. Please feel free to provide feedback or ask questions, however kind or harsh, in the comments or on Twitter.