Overview of the Match
Nobody of sound mind would suggest that Liverpool were the better side in their recent 3-1 defeat to Arsenal. It is interesting to revisit the match in the cold light of day, though.
Between half time and the calamitous blunder which ultimately decided the match, the Gunners generated three shots collectively worth 0.16 expected goals in open play. They knew where the door was but they were a long way from kicking it down.
It’s always interesting – and difficult – to determine where the praise and blame should lie after a match like that. Were Liverpool particularly poor? How great were Arsenal? Most importantly of all, did the Reds struggle because the Gunners played well or due to their own limitations?
Managerial Decisions
Managers are paid the big bucks to get to the bottom of these issues, and also to prevent them happening. Jürgen Klopp spoke after the game about the problems his team had with their right-sided triangle, though as he had selected Cody Gakpo and Ryan Gravenberch for that side – when neither has much experience there – it was perhaps inevitable.
In pulling together some immediate thoughts in the aftermath of the game for The Tomkins Times’ post-match analysis, one fact about Liverpool’s starting XI stood out a mile: none of the Reds’ top seven players for chances created per 90 minutes in the league this season got as much as an hour on the pitch and four of them didn’t play at all.
Creativity on the Pitch
As startling as that might sound, all seven wouldn’t have started even if they were available. The Reds’ reliance upon full-backs for creativity means that the leading septet for key passes includes Trent Alexander-Arnold, Conor Bradley, Andy Robertson and Kostas Tsimikas. It would be unlikely to see Harvey Elliott and Dominik Szoboszlai in the same midfield too, as it has only happened once so far, though the former is a potential option to fill-in for Mohamed Salah.
Anyway, the notion of Liverpool being without their most creative players for much of the match started a process of thought. It must be possible to calculate the theoretical level of creativity a team has on the pitch by summing each player’s expected assists per 90 minutes for the season (in the league only, to keep things simple).
Individual Creativity Analysis
It’s obviously not a flawless method. Darwin Núñez has created 10 clear-cut chances in all competitions this season but seven of them have been for Mohamed Salah, and the king has sent 12 to the Uruguayan in return. Their records suggest both would be less creative if the other part of the duo were missing.
But this simple system isn’t a bad starting point for a little insight. Adding together the xA stats for Liverpool’s starting XI at the Emirates Stadium yields a total of 0.99 per 90.
Some accounts on social media would post that without context and wait for the interactions to roll in. But it’s utterly meaningless in isolation so let’s put a little meat on the bone. A collective expected assist tally of 0.99 is the lowest Liverpool have fielded in their (coincidentally) 99 league games since the summer of 2021.
The average is 1.60, and the highest is 1.97, meaning the scale from best to worst is basically one. The next lowest behind the Emirates XI is 1.16, meaning the side which took on the Gunners wasn’t just the least creative but utterly adrift, 17.3% weaker than the next most blunt line up.
Creativity in Different Contexts
Klopp has fielded more creative sides in early round League Cup matches than that which took to the pitch to start what most people would likely consider the second hardest fixture of the league campaign. Is it any wonder they failed to create anything worth putting on a highlight reel? This side may have struggled to fashion high value opportunities against much weaker opposition by this logic.
We can apply the same principle to non-penalty expected goals. The team at Arsenal wasn’t quite so bad on this front – it was the 15th worst since the summer of 2021 – but obviously chances need to come from somewhere for this statistic to be as relevant. Plus, in the 15 lowest matches for npxG in the starting XI Liverpool scored 32 goals, compared to 25 in the same number of poorest xA games.
It’s important to note that there is only a weak correlation between this xAXI (as we can clumsily call it) and the expected goals which the team generates in the match; it’s 0.257, where 1 is perfectly in sync and 0 is not correlated at all.
However, there is one thing which stands out. If you group the games into batches of 10, the worst group for the expected assists per 90 in the side saw Liverpool earn 13 points with a total of 16.7 expected goals; the next worst figures were 20 and 18.6 respectively. The overall correlation might be minimal but the bottom of the barrel sides did tend to struggle, and in matches which included facing the likes of Nottingham Forest, Sheffield United, Southampton and Wolves.
Having only just thought of this method of assessing a side’s creativity, it only provides a snapshot. A player’s expected assist average will fluctuate throughout a campaign and they will have been more or less creative at different times, so using their season-wide figure to assess past matches isn’t that accurate. The findings are thought provoking, though, and let’s hope Klopp’s remaining teams are more creative than the side he was able to pick at the Emirates.
