Chelsea’s back three passes the eye test, struggles on paper

LONDON, ENGLAND - DECEMBER 22: (L-R) Antonio Rudiger, Kurt Zouma and Fikayo Tomori of Chelsea celebrate during the Premier League match between Tottenham Hotspur and Chelsea FC at Tottenham Hotspur Stadium on December 22, 2019 in London, United Kingdom. (Photo by Michael Regan/Getty Images)
LONDON, ENGLAND - DECEMBER 22: (L-R) Antonio Rudiger, Kurt Zouma and Fikayo Tomori of Chelsea celebrate during the Premier League match between Tottenham Hotspur and Chelsea FC at Tottenham Hotspur Stadium on December 22, 2019 in London, United Kingdom. (Photo by Michael Regan/Getty Images)

Chelsea almost always seems to look more solid in a back three coming and going. The stats on paper do not necessarily agree with that statement.

When Jose Mourinho said Chelsea goes into a 3-4-3 anytime they are in trouble, he was mostly on the money. Frank Lampard has, a few times this season, gone to three at the back when four at the back went stagnant.

While the 3-4-3 does sacrifice somewhat on the front end, it almost always seems to make Chelsea look more solid on the back end. Perhaps because of that defensive solidarity, the Blues have also looked more comfortable overall even if the attack was stifled a bit.

That is the eye test the formation passes. On paper the numbers do not quite match up when looking at the Premier League and Champions League. The xG and xG against stats are not as stark as expected when comparing the back three and the back four, though there are caveats.

In a back three, Chelsea is averaging an xG of about 1.5 goals per game while the xG against is about 0.9. In a back four, those numbers become 1.9 and 1.

So yes, a back three is more defensively solid than the back four, but barely. The Blues lose 0.4 xG for that 0.1 they gain on xG against. That does not particularly seem like a great trade off overall. When looking at the actual goals scored, however, back three does stand out more.

The Blues have not conceded more than two goals when using a back three. The three games where the Blues conceded three or more this season all came when using a back four.

There is a major asterisk surrounding this however. Chelsea has used a back three a total of seven times this season. Those seven matches included Tottenham twice, Valencia, Arsenal, Lille, Wolves, and Southampton. So the majority of those matches could be considered against difficult teams. Chelsea lost just twice in those seven.

For comparisons sake, when playing similarly difficult opponents in a back four, Chelsea has struggled much more. A draw to Arsenal and Valencia coupled with losses to Liverpool, Manchester City, and three to Manchester United is not great reading.

So while the 3-4-3 looks good to the eye and worse on paper, that may be in part due to the type of opponents it is used against. And given that Chelsea’s next set of opponents includes Bayern Munich (twice), Liverpool, a resurgent Everton, and Manchester City all before April, it may be wise to run with the back three a little bit longer. Especially with N’Golo Kante out injured.

Chelsea has played around with a few formations this season and there is still plenty to play for yet. With this difficult run coming up until April, it may be preferred to stick with what is passing the eye test until it passes on paper too.

It may not take too long, and it should bring out the best of the squad in the meanwhile.