Premier League attackers in 2019-20 – how did we do?

One of the fundamental purposes of smarterscout is to simulate how a player from one league might perform in another. It's the big question when a player switches leagues as part of a transfer or loan. It's something that every club should be thinking about. And yet, many get it wrong. Did we get it right?

During the summer 2019 and winter 2020 transfer windows, we commented here and on social media about several of the forwards who came to the Premier League. In a couple of cases, we used statistics to predict how many goals and assists they might accrue – a particularly difficult task given how idiosyncratic both can be. We think it's important to look back and asses how well we did, even given the inherent randomness of league football. So here's a summary of just about everything we could find online:

Nicolas Pepe (€79m, Lille to Arsenal). Pepe was the marquee transfer of the summer, but we thought his numbers for attacking output in Ligue 1 were underwhelming at a Premier League standard. He was clearly an outstanding finisher, yet that still didn't suggest that he'd be a worldbeater at Arsenal. After going through our reasoning and statistical analysis on Twitter, we finally tweeted this summary:

In the end, Pepe had four goals (plus one penalty) and six assists. We'll take it.

Bruno Fernandes (€55m plus up to €25m in add-ons, Sporting Lisbon to Manchester United). We had been impressed with Fernandes's attacking since his time at Udinese, but ball retention had been his bugbear. He improved somewhat, and after two seasons of stunning attacking output we expected big things:

Some followers pressed us to predict goals and assists for Bruno, too, so we did:

Fernandes ended up playing about 1,200' – 20% more than our estimate – so the midpoint of our prediction would have been closer to four goals. That's exactly what he scored, plus four more penalties.

Fernandes also had seven assists, which would have been at the very top of our range for 1,200'. But three of those were from set pieces, which we probably should have predicted separately. Partial credit?

Wesley (€25m, Club Brugge to Aston Villa). The Villans' big signing of the summer was a smarterscout young prospect, but our metrics hinted that he hadn't yet fulfilled his potential. In fact, we were concerned that Aston Villa might have repeated its mistake from the summer 2015 transfer window: signing a load of talented youngsters who weren't quite ready for the Premier League. It would have to be a steep learning curve for Wesley:

A knee injury stopped Wesley's season midway through, at which point he had five goals and one assist through about 1,800'. That was tolerable for a young striker who was still developing, but it was a risky proposition for the focal point of the entire attack.

Mbwana Samatta (€10.5m, Genk to Aston Villa). After Wesley's injury, Samatta was the stopgap. It was a small fee for a Premier League forward, and maybe Aston Villa got what they paid for. We were a tad worried after looking at Samatta's numbers:

A few Aston Villa supporters were incredulous that such a prolific scorer could be anything less than a success at their club. But that's the whole point of our league adjustments – to acknowledge that the Belgian First Division A isn't the same as the Premier League. In the end, Samatta scored just once in about 1,000'.

Joelinton (€44m, Hoffenheim to Newcastle). The Toon had coveted this Brazilian – another smarterscout young prospect – for some time, but we felt his difficulties after leaving the Austrian Bundesliga were grounds for reconsideration:

We hoped Joelinton would grow into the role, but the odds were against him; in many respects, the Premier League is a shade tougher than the Bundesliga. He finished with two goals and two assists in about 2,800', and our metrics indicated he was back at his levels of three seasons ago. Still, some scribes found time to extol his other virtues on the pitch.

Steven Bergwijn (€35m, PSV to Tottenham). Bergwijn was another smarterscout young prospect whom we had followed for several seasons, and we were pretty optimistic about his chances in the Premier League:

The bonus with Bergwijn was that his style was so similar to that of Son Heung-Min, so it seemed like he'd fit right into the Spurs setup:

Fortunately for Bergwijn, his ankle injury coincided almost exactly with the break in the season for the pandemic. He finished with three goals and an assist in just over 700', a very good haul indeed.

Moise Kean (€27.5m, Juventus to Everton). Kean's move came when he was just 19 years old, so he had to be seen as a long-term investment. But Everton had been trying to find a competent Premier League striker ever since Romelu Lukaku left, so there was some pressure even on Kean to produce:

We thought Kean's game was fairly raw, and that he might have been something of a flat-track bully in Serie A, but his attacking output was already above average at a Premier League standard. And so it was in the Premier League, too, where he had two goals and two assists in about 900'. He was probably a little unlucky not to get more.

Ollie McBurnie (£17.5m plus up to £2.5m in add-ons, Swansea to Sheffield United). During McBurnie's loan at Barnsley, he excelled as a second striker or left wing coming inside on his way to the box, and we flagged him as a smarterscout young prospect:

But as a central striker, he sometimes had trouble getting shots off:

We also thought his excellent finishing in the Championship would disappear in the Premier League, where the goalkeepers were much more skillful.

And that's exactly what happened. McBurnie was among the strikers with the lowest shooting rate in the Premier League. Where he had scored 24% of his shots in the Championship, he only scored 13% in the Premier League – and the quality of his shots barely changed. He had six goals and zero assists in about 2,200'.

Apart from some comments about how good Sebastien Haller was in the air, these tweets were about all we could find in terms of our analytics views on the forwards who came to the Premier League on big transfers. We've continued to tweak our metrics since 2019 to improve their precision, and we hope to do the same exercise again next summer. Thanks for reading and following along with us!

Recent articles

How we spot smarterscout young prospects
Private membership – there can be only one!
Should clubs focus on shot quality using expected goals?
Introducing xGAR: expected goals above replacement