Big questions for the new season

New season, new hopes and new dreams – and maybe new worries, if you’re the glass half-empty type. I tend to stay more upbeat, so here are the big questions that I’m looking forward to seeing answered in 2018/19.

Can we play a different way?

It was a huge, very pleasant surprise last season that United were essentially able to roll with the system that had been so successful in League One in 2016/17, with tweaks rather than any overhaul. And if anything, it seemed to work even better as the Blades stormed to the top of the league, demolishing more-fancied opposition along the way and drawing plenty of praise for the way our wide centre-backs would get forward and overlap the wingbacks to overload opponents.

At times, however, particularly in the second half of the season, that system became less effective. The results dried up, which to an extent was a regression to mean (we overperformed our underlying numbers in the start of the season – we were still good, but perhaps not that good), but it’s undeniable that our chance creation nosedived as we got into the back end of the campaign. You can put that down to a number of factors, all of which are fair comment and probably all contributed: Coutts’ injury, teams playing more conservatively against us, fatigue, and an infuriating inability to hold onto the lead late in games.

All of that I feel can be overcome with recruitment, planning, and better luck with injuries and the manner of goals conceded (more on that in a moment). What I’m more interested in seeing is: can we move away from our standard, possession-heavy 3-5-2 when the situation demands it?

Fulham created chances consistently through the 90 minutes, whereas United offered barely any threat until they were already 2-0 down and the game was already gone.

Fulham created chances consistently through the 90 minutes, whereas United offered barely any threat until they were already 2-0 down and the game was already gone.

At times last season – and it wasn’t often, but we’re talking fine margins of making the playoffs here – we were simply outplayed. I’m talking specifically about the away games with Wolves (47% possession, outshot 13 to 6) and Fulham (36% possession, outshot 17 to 7). The Expected Goals timeline for that Fulham game, via Experimental 361, show how easily we were held at bay (right).

Games against the two eventual promoted teams are always going to be tough, especially against ones of the quality that Wolves and Fulham possess. But these were two games where it was clear we had no obvious answer to the high standard of opposition. Games where, honestly, we looked out of our depth. And that’s fine! Players like Ruben Neves, Ryan Sessegnon, Ivan Cavaleiro, Aleksandr Mitrovic and Tom Cairney are (literally) in a different league to anyone United currently have on their books.

Other, more mediocre teams did run these two a lot closer than we were able to, by largely playing deep and then playing on the counter by getting the ball forward quickly. This is an area United struggled last season – largely due to a lack of pace across the attack, but also our patient approach. I would hope that this year we’re able – on occasion – to play in a more tight, counter-punching fashion when the situation demands it. And also in a fashion that doesn’t lead to us blowing late leads in 4 out of 5 games during the season’s crunch-time. Speaking of which...

Will our defence continue to underperform the advanced stats?

Back in April I wrote a piece making the case for our defence being better than the amount of goals that it had conceded. Simply put, United had one of the best defences in the Championship by pretty much every metric you could use – apart from the one that ultimately matters: goals conceded.

A quick glance at the end-of-season stats shows that this was broadly still the case: only Wolves and Preston faced fewer total shots, and only five teams faced fewer shots on target. However, we ranked 11th for goals conceded, with 55 shipped in 46 games – more than any team who finished above us in the table. In terms of Expected Goals – the quality and frequency of chances conceded – we actually improved over the final few weeks of the season, ending 2017/18 with the best xGA out of all the teams, and by almost two clear goals at that.

This table is ordered by Expected Points based on xG for and against, via @EFLStats on Twitter.

This table is ordered by Expected Points based on xG for and against, via @EFLStats on Twitter.

You won’t need me to tell you that this is frustrating. By advanced metrics we had one of if not the best defence in the league, and “should” have conceded around 17 goals fewer than we actually did. That’s a lot of goals – over 30% of our overall total – and potentially a lot of points down the pan.

The big question of course, is why. I had a few suggestions at the time, including the idea that we were just plain unlucky to run into some fantastic finishing on occasion, but after further consultation with our resident expert, there’s a couple of other things that may have contributed:

  1. While I like both Jamal Blackman and Simon Moore, their save percentages weren’t all that impressive (ranking in the bottom half of the league) – possibly why we presumably didn’t try to re-sign Blackman, and have instead brought in Dean Henderson from Manchester United.

  2. Our defensive shape meant that teams were usually limited to low-quality shots (as in, generally from distance) – but we struggled to pressure those shots, meaning they were actually “better” quality than the xG stats may show. Think of that period when we conceded goal after goal from long range, as teams either found space for a shot or were allowed to advance on our goal and take a shot relatively-unchallenged.

The good news is that both of the above could and should be improved this season with acquisition (Henderson, John Egan) and tactical adjustments.



Will Coutts slot back in – and will we be as good?

Stop me if you’ve heard this one before, but United’s results dropped off after Paul Coutts suffered a broken leg against Burton in November 2017. By all accounts, Coutts’ recovery has been successful and he’s involved in preseason training, although at time of writing he is yet to play any of the friendlies.

That lack of involvement so far – along with a lengthy period without any competitive football – would suggest that we may not see Coutts in the starting XI against Swansea on August 4th. That would be a real shame, as it was a genuine pleasure last season witnessing a midfielder in red and white at the absolute peak of his powers.

Sadly, we’re dealing with two unknowns here – will Coutts be as good as last season? And will we be as good as we were during those first few months? Coutts only turns 30 later this July, so there is likely plenty left in the tank yet – and his game hasn’t been based on pure athleticism or speed anyway. But a leg-break is a serious injury. The truth is, we just don’t know if he’ll be able to come back as strong as he was.

The second part of this is where I’d urge a note of caution among United fans. It is pretty unlikely that, even if Coutts comes back the exact-same midfield maestro he was last season, we will be picking up results at the same rate as we did to open 2017/18.

Jay broke this down in an article on Bramall Lane HQ towards the end of last season, which showed two things: 1) United were indeed a better team with Coutts in the side, but 2) our great form at the start of the season wasn’t necessarily down to Coutts’ presence. Breaking the season into two parts (this comparison was done on April 13) shows that we took a comparable amount of shots post-Coutts, but that our efficiency was way up in the first part of the season – almost absurdly so when you look at shots-per-goal:

pasted image 0 (2).png

Needing fewer than 7 shots to score a goal? Turns out, that was unsustainably high – over the full season, Wolves averaged 7.3 shots per goal (ranked 1st in the league). It subsequently regressed back to mean.

Now, you could also make a case that we regressed “too much” in the back half of the season – the league average for shots-per-goal was 10.1, and as you can see above we were averaging a goal every 11 shots in the 21-game sample. Not good. An easy conclusion to draw is that without Coutts and his incisive passing, we didn’t create the same level of quality chances in the rest of the season – but it turns out that’s not accurate either. Our Expected Goals in Aug-Nov was 27.8 (19 games) compared to 27 in Dec-Apr (21 games). Essentially the same.

This is a roundabout way of saying: don’t expect us to go blazing up the league table if Coutts has returned to full fitness. I hope it happens of course, but if it does, there are probably some other factors at play as well – namely those discussed in this article.

Will Clarke continue to perform at the same level?

Leon Clarke racked up 19 goals in the league last season, a fact I sort of feel we all took for granted in the end.

He was the joint-third scorer in the Championship, tied with Bobby Reid (who just signed for Cardiff for £10m) and just behind Lewis Grabban (signed for Forest for £6m) and Matej Vydra (rumoured to be joining Leeds for £11m). Unlike those three, Clarke didn’t take a single penalty last season (Vydra scored 6, Grabban 5, Reid 3): take pens out of the equation and Leon Clarke, ladies and gentlemen, was the top-scorer in the Championship.

While I’m firmly subscribed to the train of thought that there was much more to Clarke’s game last season than just goals – witness the fact that he was one of the few players who could keep pressing for 90 minutes, and the way he held the ball up and chased direct passes – I have to wonder how he’ll fare this season.

He’s 33 years old now, and his injury history is not great (even in his incredibly-successful 2017/18, he still missed time with injury on multiple occasions). It’s been a nomadic career spanning 444 games for 16 different clubs, from Crawley to Wolves, across fifteen years. Last season he played 3,332 minutes in the Championship, fifth-most in the squad. There are a lot of miles in those legs. Last season was also the first time he’d played 36+ league games since 2009/10, as well as the first time he’d scored more than 8 goals at this level in his whole career. Those stats suggest that last season could be a bit of an anomaly.


Those usage stats above don’t tell the whole story. Anyone who watched us more than a couple of times last season will know that Clarke’s goals weren’t due to a freakish run of finishes or shots deflecting in off his backside. We created plenty of tap-ins for him as well as the brilliant strikes he stuck away against the likes of Birmingham, Bristol City and Fulham. With United’s style focussed on creating high-quality chances, there’s a good likelihood that Leon will still be banging them in this season. Here’s hoping for another two at Hillsborough, anyway.

Ben MeakinComment