Liverpool have not made a flying start in 2014/15, yet I don’t believe that things have been quite as bad as they have been portrayed.
One of the key things that I monitor is expected assists, which as you may know by now is based upon where on the pitch chances are created, and the likelihood that those opportunities will be scored. Using this metric, the Reds appear to be in fine fettle.
I wrote about my expected assist findings at length here, but here is a summary of the findings. When I calculated the expected assists difference for the Premier League last season and ranked the teams, sixteen of them were within three places of their actual league position, with the average rank difference being 2.3 places. In other words, it certainly won’t ever predict the table exactly, but it can give you a pretty good indication of where most teams deserve to finish.
Here’s the table for the season so far; this shows the teams’ total chance difference (chances created minus chances allowed), and their expected assists difference, which the teams are sorted by.
I assume it’s probably better than you thought it would be, from a Reds perspective? It’s worth noting that Liverpool are the only team that are in the top three for both expected assists for and (least) expected assists against. For the record, Manchester City and Chelsea lead the way on the attacking table, with Southampton and Arsenal the best teams defensively.
Optimistic Kopites are currently clinging to the fact that Brendan Rodgers’ Swansea and Liverpool teams have always done better in the second halves of their league campaigns, and also that the Reds started last season slowly before bursting into life a few months into the campaign. With this in mind, how do Liverpool’s expected assist figures compare to the opening seven games of 2013/14?
It’s encouraging to see that progress has been made at both ends of the pitch, with a 20% increase in expected assists up front, and a 16% reduction at the back.
The overall picture is essentially the same if we compare 2014/15 with the corresponding fixtures from last season too, albeit the far greater gain has been made in defence (28%) than in attack (6%). That there have been gains at all though is clearly the main thing, as we can see here:
So if we extrapolate Liverpool’s form so far to a full season, thirty-eight game total (which isn’t hugely valid when based on only seven games, but will give an indication), we find the following:
There may be a dip in the figure up front (which you would expect having seen Liverpool play so far this season) but also a significant improvement at the back (which has been less obvious) which means that the overall difference figure is better. Based on last season’s Premier League figures, an expected assists difference of 14.17 would be the third best in the division.
It’s very important to remember that the system is very simplistic though, so for instance a cross into the centre of the box is rated the same as a throughball, when in reality the latter has a far greater chance of being converted.
Liverpool are crossing more (by four per game) and playing three fewer through balls per game than they were last season, which will explain why the improved figures here don’t exactly tally with performances, but the expected assist numbers suggest that the Reds are on the right track at least.
As I mentioned above, the expected assists system can’t predict a team’s form perfectly, but why are Liverpool three points and eleven goals in goal difference down on the corresponding fixtures from last season?
My initial assumption was that this season’s Reds have an issue with shots in the penalty box at both ends of the pitch. The table below shows Liverpool’s figures for the last six seasons, and we can see that the balance this season is not particularly bad.
Liverpool have had fewer shots in the box than in any of the past five full seasons, but equally they’ve allowed their opponents the fewest this season too. This may in part explain why the Reds currently have a goal difference of zero, but to my mind the reason that Liverpool’s results have not been better is down to the quality of chance that they have had and allowed in the penalty area.
Brendan Rodgers’ team have had fifty-nine shots in the box this season, but only seven of these have been classified by Opta as big chances (lest we forget, a big chance is “a situation where a player should reasonably be expected to score usually in a one-on-one scenario or from very close range”). This means that only 12% of their box shots have been big chances. Liverpool’s opponents may have only had thirty-six shots in the Reds’ box, but twelve of these have been big chances, which is 33% of the total.
One of these big chances may have been a wrongly awarded penalty against West Brom, but the point remains that whilst Liverpool are doing well at restricting opposition shots in their box this season, when they do allow a chance it’s often a sitter that their opponents are likely to score.
I haven’t looked at this metric before, so I have no idea if Liverpool’s figures of 12% and 33% are particularly good or bad (though that is now on my to-do list!) but if nothing else the balance is currently heavily tipped in their opponents favour, and I can’t see the Reds massively improving their results until they tip the balance back towards them.
To conclude, whilst it initially seems that Liverpool’s expected assist figures are encouraging, there needs to be a large shift in quality of chance in the box at both ends if the Reds are to kick on towards the top four and beyond.