In Search of Salah: Is Liverpool’s Star Out of Form or Out of Position?

By: Kevin Chroust | October 5, 2018

Liverpool’s Mohamed Salah hasn’t been the winger he was a season ago, but his performance might not be entirely to blame. It could be that he literally hasn’t been a winger. STATS Edge and Tier 6+ data show his touches aren’t happening as wide as they did in 2017/18, which combined with slight changes in team style has led to less involvement.

 

Liverpool are unbeaten and trail runaway 2017/18 Premier League winners Manchester City only on goal difference heading into Sunday’s fixture at Anfield, yet Mohamed Salah can’t catch a break. The Reds are joint second in goals scored with an exciting Chelsea side, yet their star supposedly isn’t doing enough.

When judging their attack through the lens of style and scoring, Liverpool are quantifiably similar to what they were last season, yet the PFA Player of the Year is getting fried in the media for being out of sorts. This warrants a deeper dive into data and video.

We’ll first look into whether this is a team thing, and that starts with comparing Liverpool 2017/18’s attacking summary and playing styles to this season before digging into the player specifics. First, 17/18’s summary:

This is comparing Liverpool last season to all 38 of their opponents with the dotted lines representing the league averages. The Reds outshot their opponents, and they also outperformed their own goal expectancy by 0.22 goals per match.

Now, 2018/19 through seven matches:

Liverpool’s shot differential is still drastically in their favour, but is has leveled off some. Their possession is down slightly, as are their crosses in play and corners won. And note that while they’re scoring nearly the same as they were last season, their finishing hasn’t been quite as clinical. We determine that from their xG being higher than their actual goals, meaning the chances they’ve created haven’t resulted in quite the level of finishing expected of the average side. Before we relate this back to Salah, we need to assess quantifiable team style across seasons.

First, 2017/18:

The bold values represent the per-match average of team possessions for that style. The percentage below represents the team’s departure from the league average of 0 percent, which is shown as the dotted line. It comes as no surprise that we’re talking about a fast-paced side that tends to counter and press.

Now, 2018/19, where we’ll see plenty of the same with subtle differences:

We’re still talking about an up-tempo side that tends to counter and press, but some of the possession-based styles have changed. For example, there’s less sustained threat, which we’ll tie back to Salah in a moment.

What’s the difference in Salah, and to an extent current leading scorer Sadio Mane, from this season to last? Very basically, Salah had 67 shots on target last season to Mane’s 26. This year, Salah has 12 to Mane’s 11.

Specific event data allows us to go deeper in various ways. One is it tells us the average touch location for a player over a match or season. Both Salah and Mane are shading into the middle more, and Roberto Firmino’s average touch has moved left toward Mane. Salah’s average touch is occurring 3.0 metres left of where it happened last season, while Mane’s is happening 2.4 to the right. One might think that’d get them more involved, but we’ll show in a moment how it’s had the opposite effect. Firmino’s is 2.2 left of last season, so Salah remains similarly isolated from his centre forward but in more of a central location, while Mane and Firmino have been in closer quarters.

It’s more than just positioning. For Salah, Liverpool’s changes have resulted in a dip in involvement. Note below that his build up, sustained threat and fast tempo workload has all fallen off this season:

Salah’s in fact seen a dip in just about every category, which may ultimately relate to Liverpool falling a bit short of that team expected goal mark we noted above. Salah’s 32 league goals last season came with an xG +/- of +13.1, so his value above the league average player probably can’t be overstated. This season, he’s at -1.6.

Within STATS Edge, analysts can go to the video to assess for themselves what’s going on with Salah by filtering any of his ball events within a specific style and search spatially by simply dragging a box over the area of the pitch they want to assess. Below, we’re showing the location of all of his shots for the season within all contexts, then choosing locations within or just outside of the penalty area:

From there we can view the video of each. Three of his four shots last weekend against Chelsea originated or occurred centrally, and the other came from the left after a corner. The results were less than ideal:

It seems necessary to reiterate that Liverpool are even on points with Manchester City. They haven’t lost. They have an improved defence and goalkeeping from this time last year, and there are signs of their attack evolving. Mane has also seen a dip in involvement but thus far seems to be benefitting from being closer to his teammates up top with a +1.4 xG +/- in league matches after all of 2017/18 yielded a +1.1. The player up top who’s actually been more involved is Firmino, who’s on pace for 10.8 assists after totaling 14 the past two seasons.

But there’s no denying that, so far, Salah has been involved less and he hasn’t been finishing quite as well. Could it be that if he were getting the touches he’s accustomed to over the course of a match in a position he’s used to he might be sharper? Maybe yes, maybe no. But it’s not even a question we’d think to ask without the watchful, objective eye of data and video.