A nerdy look at Leicester's poor start

Has Brendan Rodgers blown it, or is there more to the story?

The story at Leicester under Brendan Rodgers is supposed to take a predictable route.

9th November 2019, twelve games into the season, Leicester beat Arsenal 2-0 at home. This puts them a very impressive second in the table on 26 points. Considering how the side looked under Claude Puel, it seems that Rodgers is working miracles.

13th December 2020, twelve games into the season, Leicester beat Brighton 3-0 at home. This puts them a very impressive third in the table on 24 points, just one off the top. Considering the way the previous season unravelled for them, courting predictions of doom through a broken mentality, Rodgers seems to have done terrifically well to get the train back on the tracks.

This time of year is when Leicester have previously been brilliant. And yet here we are. Leicester played their twelfth game of the season this weekend: a 3-0 loss to Chelsea that put them in the bottom half of the table. If it’s this bad in November, just imagine how rough things can get when they implode in the spring! Or maybe football doesn’t work like that. We’ll see. But what has gone wrong here? Something huge and disastrous? Nothing much? Let’s dig into it.

---

A note on stats: all numbers are from FBRef using data from StatsBomb, apart from PPDA, which is from The Analyst using data from Stats Perform.

---

The broad strokes

Leicester are twelfth in the table with a negative expected goal difference. A lot of teams have a negative xG difference this season, so that’s not disastrous in and of itself, even if it’s obviously Not Great. What does look disastrous is the trend in the data over the last few years. Take their xG difference per game over the last several seasons:

18/19: +0.14

19/20: +0.45

20/21: +0.22

21/22: -0.14

The league table didn’t show it, but Leicester’s performances declined quite a bit last season against the year before. As StatsBomb previously noted, “last season only Sheffield United dropped off more in this regard.”

So we were already looking at a team that fell off significantly in the numbers, and now they seem to have done the same thing again. Ouch.

This was already a team when the old xG reversion might rear its ugly head. They scored twelve goals more than expected last campaign. The previous year, they outperformed a little bit at both ends, which adds up nicely. Rodgers didn’t show signs of breaking xG models at Liverpool, so it always felt like this would come to an end. It might be doing so at the worst possible time.

In the plus column, the fixture list hasn’t been kind so far. Eight of their twelve fixtures so far have been against teams currently in the top half of the table. The next four league games – against Watford, Southampton, Aston Villa and Newcastle – represent good opportunities to pick up points, and we might be having a different conversation in mid-December.

Having said that, Leicester picked up 23 points from the equivalent fixtures to their opening twelve games last season. This doesn’t look like a case of a particularly tricky run that should sort itself out soon. If Leicester are to improve results, they’re going to have to start playing better football. Let’s take a look at how.

Defence

Leicester’s 21 goals conceded put them joint third-worst in the division on this side of things. The xG is a little bit better, but not a lot. We’re still looking at a mediocre defensive team.

They’re not pressing as high as they did last time around. They’re allowing the opposition to make about two more passes before attempting to win the ball back. That may not sound like a lot, but it’s the difference between being the fourth most aggressive side without the ball last season and about average now. Similarly, the percentage of their pressures that take place in the final third has dropped from eleventh best in the league to 16th. This Leicester don’t seem to be quite so quick in the press as last year’s vintage.

The outcome is predictable: the Foxes are conceding more shots than they used to. Last season they gave up a very impressive 9.71 shots per game. This year, that figure has rocketed up to 15.7, the third-highest in the Premier League. This is to be expected. When you press high, your aim is to prevent shots from happening by stopping the opponent from even getting into situations to pull the trigger. However, when you sit a little deeper, you accept conceding more shots in order to limit the quality of chances faced. Picture a team taking 30-yard potshots against an opponent with everyone behind the ball and hopefully you get what I mean. So that raises the question: have Leicester reduced the quality of each chance they face?

Actually, yes, they have! But not by enough to benefit. Their xG per shot conceded has improved from a league-worst 0.13 to an above-average 0.09. The problem is, putting my B-grade maths GCSE to work, they can only give up about four more shots per game to be better off overall. And they’re giving up six more. There are positives to the new approach, but on balance it’s a negative.

Set pieces have been a talking point, but they only look like part of the problem to me. The Foxes have conceded five from set plays, which isn’t great, but I don’t think everything would be fine if they had let in a league-average three instead. Leicester were poor at defending set-pieces last season and that hasn’t changed.

Attack

Leicester, as we established, beat their xG quite handily in attack last season. The best finishers here were Kelechi Iheanacho, Harvey Barnes and James Maddison, netting 29 goals between them from 17.6 expected. One man who wasn’t banging them in with every half-decent chance last year was Jamie Vardy. The striker scored just seven non-penalty goals from an xG of 12.8 last time out. A smattering of penalties early in the season saved his blushes.

This year, Vardy’s got the golden touch back,¹ scoring seven non-penalty goals and sitting behind only Mohamed Salah in the top scorers table. He’s running a little over his xG, but the bigger picture is that Vardy is still Vardy. That’s good, as they really do need him right now.

Leicester’s attack is broadly the same under the hood. The xG per game has dropped by 0.07, which would average out to maybe 2-3 goals a season. They’re taking one shot fewer per match. But, again, these are slightly tougher fixtures, so it would make sense for them to just decline a touch. The problem is that we weren’t looking at an incredible baseline in the numbers last year. The Foxes put up the sixth most xG last season, which is nothing to sniff at, but it didn’t look like an attack that could push them into the top four.

Youri Tielemans has become an excellent creative midfielder, and he’s someone who Leicester really rely on to move the ball into dangerous areas. His 7.5 progressive passes per 90 are bettered only by Paul Pogba in terms of midfielders. It’s been important for him to become the creative heartbeat of the side as the person once expected to fill that role, Maddison, just doesn’t seem to be pushing on. He got a lot of credit last season for scoring some screamers, but the all-round contribution just wasn’t what it should be for someone of his ability. When Rodgers took over, I wondered if he would mould Maddison in a similar way he did Philippe Coutinho. It looks like we’re seeing the downside of that, as the attacking midfielder would often rather shoot from range than pick out an exquisite pass.

The trio of Iheanacho, Barnes and Maddison who smashed xG last season have this year scored just three times from an expected total of 4.6. That’s well within range of “shit happens”, but it also shows the pitfalls of relying on great finishing to power your attack. The trio scored 43% of Leicester’s goals last season. This time, they can’t even put up half the total of Vardy alone.

It sounds like I’m being harsh, but this is broadly the same attack as last season. They just ran hot in front of goal, and they’re not now.

Rodgers is, if nothing else, an improviser. When something isn’t working, he’ll more than happily change it, shamelessly ripping off whatever tactical trend he likes. We’re already seeing it this season with the recent move to a back three (that doesn’t seem to be making a huge difference). I wouldn’t bet my life on it, but I think the odds are pretty good that he finds a tactical solution to get them back to the standards of last season.

The problem is that might not be good enough. Leicester came very close to Champions League football last season, but they never looked like a team of that quality in the data. They were seventh-best by xG difference that year, which feels about the standard the Foxes should be playing at. Sooner or later, Rodgers should get them back to that level, but it will probably be too late to matter for the season. Leicester’s trajectory over recent years has been very positive and they shouldn’t throw the baby out with the bathwater, but this year will likely go down as one step back after taking so many forward.

1. In writing this sentence, I realised I had completely forgotten that the band Razorlight even existed, let alone were very popular in the 2000s.