Liverpool vs. THE MOST DANGEROUS LEAD IN FOOTBALL™

Liverpool haven't scored when 2-0 up this season, conceding four goals along the way. Is this a serious problem for Arne Slot or a fuss about nothing?

It has long been said that 2-0 is the most dangerous lead in football. You think you’ve won, then ease off only to come unstuck.

There are many clichés in football - in England, especially - but how many of them have their own Wikipedia page? The idea that teams regularly fail to win after going 2-0 up persists, despite being disproved numerous times.

In the Premier League, the win percentage from this standpoint was 89.6 per cent up to April 2020. It won’t have shifted much in the half decade since.

Liverpool will shift the figure singlehandedly this season if they’re not careful. Their goal difference from their time spent at 2-0 is -4, thanks to failing to score while conceding to Bournemouth, Newcastle, Atlético Madrid and Everton. The first three also scored again to equalise, only to later lose. No points have been lost after having a two-goal lead, more through luck than judgment.

Some of the luck could be argued to have been bad for Liverpool. Atléti scored twice with deflected shots, with a potential offside also in play for their first. Their second goal came via a 0.09 expected goals chance, which was 0.01 above Idrissa Gana Gueye’s strike last weekend.

The other two goals which made games 2-1 were at least Opta-defined big chances, albeit Bruno Guimarães’s header was only worth 0.17 xG. The Reds’ ‘first team’ have allowed a single clear-cut chance in the four matches since William Osula equalised at St. James Park. That’s a huge positive, not least as they didn’t have more than one clean sheet of this nature in any four game sequence last term.

The narrative that Liverpool can’t defend 2-0 leads will rumble on until they do. They’d almost be better off coming from behind to give the media something different to discuss. In the meantime, it’s worth digging into their expected goals by game state as there is a clear kernel of concern.

The Reds’ record in the early weeks of 2025/26 demands further scrutiny for a simple reason. Their total of four goals conceded at 2-0 tops the three they allowed in the previous two seasons combined.

Liverpool allowed 2-0 to become 2-1 a shade over a fifth of the time between 2015/16 and 2024/25, a long way from it happening in every match in which it was a possibility. This suggests the current run will end, the law of averages implying that the Reds might protect or increase their next 16 such leads. It’s hard to regress when the xG numbers are so ugly though.

Understat’s expected goal model is not revered as one of the greats. It does have some advantages over other publicly available data, in that it goes back as far as 2014/15 and it provides figures by game state. If we want to compare the Reds of Brendan Rodgers with Arne Slot’s side, we can.

As it stretches back that far, we have 11 full campaigns of Premier League data, plus the early weeks of 2025/26. The game state data is split into five types: tied, one goal up or down, then two-or-more in each direction. As Liverpool have yet to be behind this season and never trailed by more than one in 2018/19, we have 57 figures.

It’s reasonable to assume that a team that is leading by two goals will concede more shots than they take. Their opponent has to go for broke in an attempt to save the game. Yet Liverpool often lead the shot count regardless, with their solid defence freeing lightning fast attackers to maintain their dominance when better than a goal up.

Until now.

The Reds have a negative shot difference across the 60 minutes that they have spend leading by two in the Premier League this term. It’s only minus one (due to six for, seven against) and would stand at +1 if we included the Champions League match. But against their average of +6.2 per 90 minutes in that game state across the preceding 11 league seasons, it’s still a drop.

The expected goal difference of -1.5 per 90 (albeit from only a one hour sample) is well below the previous +1.3 average too. Only one of the other 56 sets of data has a worse rate. It occurred in the 84 minutes that Liverpool spent trailing by two goals in 2019/20, with almost two-thirds of that time spent hungover at a deserted Etihad after the title was secured.

The Reds’ 60 league minutes when winning 2-0 look historically bad in the data but why is that? How have they found themselves taking six shots for 0.36 xG while conceding seven that were worth 1.33 expected goals?

Answers on a postcard are appreciated if you have a theory. Having extricated the pass numbers for ‘2-0 time’ from the rest of the league matches on Stats Zone, it seems fairly random in two of the three cases. The exception was Everton.

Liverpool did not dominate the ball to the same extent for the ~29 minutes in which they led 2-0. While their pass accuracy was less than three per cent lower during this time, the percentage of their completed passes that were in the final third plummeted from 33 to 18.9. They were no longer able to keep the ball away from their goal as often, with Everton able to complete a much higher proportion of their successful passes in the Reds’ defensive third.

As they had to push to the finish line twice in the six days preceding this lunchtime kick-off, we shouldn’t be too critical. The period after the Blues’ goal was more in line with the match as a whole, so credit Liverpool for that.

There was little in the percentages if we look at the Bournemouth match. The home side were hit with an expertly delivered counter attack, with the Cherries completing just 10 final third passes during our period of interest.

The Newcastle match makes this whole exercise a sham. The Reds could have done little more during the 11 minutes in which they held a 2-0 lead.

Eddie Howe’s men completed one pass between the 47th and 57th minutes. One. One pass is all it takes. It was a masterclass in playing for territory. Rugby fans must have loved it.

A Nick Pope punt forward was headed into touch. The same happened to the subsequent throw-in, which moved the Magpies further up field. You know what happened after the long throw into the box which came next.

If there’s no doubt Liverpool could have defended better at key moments, it’s also true they were punished beyond what feels reasonable on the balance of play. People repeatedly say they can’t keep getting away with turning draws into wins so late on. Perhaps, but people should be saying they won’t continue to concede goals at 2-0 up too.