Against Chelsea this weekend, Tottenham posted one of the lowest expected goals totals in Opta’s Premier League record books. Is there any excuse for this performance?
On Saturday evening in north London, Tottenham Hotspur put in a historically bad attacking performance.
Playing at home against Chelsea, one of their biggest rivals and for many of their fans, in one of the biggest games of the season, Spurs put in one of the most blunt, inept, and ineffective displays in their modern history.
It was the lowest xG total by any team in a Premier League game this season. It was also the second lowest by Tottenham in any Premier League match in Opta’s record books, which for this particular metric stretch back to 2012-13.
The only game in which Spurs have posted a lower xG was a defeat to Crystal Palace in September 2021 (0.06 xG) during Nuno Espírito Santo’s short and, perhaps ominously, ill-fated reign. That day, Spurs played away from home and did so with 10 men for more than a third of the match, making this weekend’s xG tally their lowest on record in a home game or without having a man sent off.
“That hurts massively,” manager Thomas Frank said after Saturday’s defeat. “I have never been in charge of a team that has created that little in one game.”
xG isn’t for everyone. Interim Celtic manager Martin O’Neill recently called it “total nonsense.” The main criticism of xG appears to be that there is nothing to learn from it. “Some people just use these words to try to sound clever,” O’Neill continued.
Being data people, we’re never going to agree with the Celtic boss, and what’s more, on this occasion as much as ever, we believe there is an awful lot to glean from the numbers. So, with the aim of both sounding (very) clever and analysing just how bad Tottenham were on Saturday, here we are asking what Tottenham’s xG of 0.10 tells us.
The headline fact is that if Spurs finished their chances in line with xG – at the average rate at which goals are scored in the competitions on which the xG model is based – Spurs could have played out this exact performance against Chelsea 10 times before they created enough chances to score a goal. In other words, if they’d carried on playing exactly as they were, they would have needed 900 minutes to find the net. That’s 15 hours of football. Or just under two-thirds of a day.
Chelsea, meanwhile, racked up chances worth 3.68 xG, meaning they needed just 24 minutes and 27 seconds to create chances worth 1.00 xG – or a goal’s worth of chances. In reality, they needed just over 33 minutes to actually score what turned out to be the winning goal, and Spurs had goalkeeper Guglielmo Vicario, who made eight saves, to thank for the final score being even the slightest bit respectable.
No team have managed to create as little in a Premier League game this season as Spurs did on Saturday. Their 0.1 xG is lower than newly promoted teams Leeds and Burnley created against Arsenal, who could break the Premier League record for fewest goals conceded in a season. It’s lower than Sunderland managed in either half against Aston Villa and they were reduced to 10 men after half an hour.
Looking further back in the history books, it doesn’t get much better for Spurs or Frank.
There have been 5,039 Premier League matches played since the start of the 2012-13 season (the time for which xG data is available). That gives 10,078 team totals in games (each match has two teams) in that 13-year period. Only 37 – or 0.004% – have produced less than 0.1 xG.
Of those 37, 27 were away from home, and the vast majority were teams who had either just been promoted or would go on to be relegated that season. Many of the others were facing peak Pep-Guardiola-era Manchester City.
It’s true that Tottenham haven’t created freely under Frank all season, particularly in open play, and so far, other than creating a bit of disgruntlement among the fans, it hasn’t really hurt them very much. Even after this dreadful performance, they are still fifth in the table, just two points off City in second.
They have overperformed compared to their xG more than any other team in the Premier League this season, scoring 6.8 more goals (17) than their xG (10.2).
But given they can no longer rely on the world-class finishing of players like Harry Kane and Son Heung-min, and are instead creating chances for the likes of Mohammed Kudus, Richarlison and João Palhinha – the Spurs players with the most shots this season. Doing so was never going to be sustainable. Against Chelsea, that lack of creativity was exposed brutally, and with grim results.
There are some counterpoints that are worth considering, including some of the limitations of xG models. xG doesn’t take into account periods of sustained possession deep in opposition territory or good moves that get close to goal but don’t actually end in a shot. A ball fizzed across the face of goal that a striker misses by an inch gets no xG value, for example. It doesn’t always tell the full story.
But anyone who watched Tottenham play Chelsea could tell you that the stats didn’t lie. Spurs were awful. Even at set-pieces, from which Spurs have been pretty reliable under Frank, they carried almost no threat. They couldn’t make any of their six corners count. Not one of them led to a single opening or even a moment where a goal looked at all possible.
However, it is worth mentioning the mitigating factors that Spurs and Frank can point to.
He is new to the club, and everyone there is still getting used to each other. Since becoming manager, Frank has – quite understandably – prioritised fixing Spurs’ leaky defence, which was such a huge problem under predecessor Ange Postecoglou, and the result has been some level of compromise at the other end of the pitch.
Spurs aren’t scoring goals or creating chances anything like as reliably as last season, but they are also much, much more solid at the back. They have the third best defensive record in the Premier League this season, and have lost 30% of their Premier League games, compared to 57.9% last season.
There is also a lengthy injury list. Many of those who are fit and available are either new to the club or to regular first-team football there, while many of the injured players would walk straight into the team. Nine players are confirmed as absent for Tuesday’s Champions League tie with FC Copenhagen, and two more are doubtful with knocks.
All that said, there is no excusing how bad Spurs were. Three shots, all from Kudus, all from positions where he was extremely unlikely to score, is a damning state of affairs, whatever the circumstances.
It’s not quite, as we stated in this week’s knee-jerk reactions column, time to rip everything up and get rid of Frank, but this terrible display needs to be left in the past. Spurs must do much, much better than this, or there’ll be little defending Frank.