Missing out on European competition for the first time in years is disappointing. Especially in Atalanta’s case in which it seemed like the 2022 felt like one of those dreams where it is impossible to move. In Atalanta’s case it was on a runaway train that had the tracks cut out, but it couldn’t do anything to move out of the way once it ran out of rail. Essentially, everyone knew what was coming.
Earning 8th place while simultaneously only earning three victories against relegation sides should deservedly have Atalantini and management disappointed in not having reached the season’s likely goal of at least European football (if not another crack at the Champion’s League). On the other hand, everyone knows that the season was freakish in many ways. From the two COVID matches with half the team missing, to some dubious refereeing decisions, to Josip Ilicic’s unfortunate extended leave, to Duvan Zapata’s team-derailing injury – not a whole lot went right.
On top of that, the cruel mistress that is luck seemed to be opposite of Atalanta all year. And this is why it would be premature to raise the alarms too quickly. Its easy to say La Dea got unlucky this year – but it is more comforting to say it when there is available data to actually back up the claim. Let’s dig into some of that data first, and recap how things may have gone better if Atalanta could’ve just regressed to the mean.
Comparing Different xG Metrics
On the season Atalanta faired relatively well in the expected goals department. Now this is not to say to take the models at face value (because goal differential was a better predictor of points this year than expected metrics), but different metrics can still help us explain some things. If you want a deeper dive into expected metric comparisons, go over to my Serie Analyst POD and listen to Episode 8.
More often than not Atalanta finishes around 5th or 6th place in most expected goal differential metrics, and our Dea even outpaced both Roma and Fiorentina to finish 6th in goal differential. However, one specific metric stood out that really begins to explain the gap in Atalanta’s game strategy and the results that actually came to fruition: non-shot xGD. FiveThirtyEight.com evaluates matches in two formats: the classic shot based formula, as well as a non-shot based model that strips away the final shot and determines how a team should score based on field positioning, turnover creation/prevention, and other metrics.
By this metric, Atalanta fairs considerably better than every team in the league asides from Inter. And while it is not a perfect metric, it suggests that Atalanta goes out with a game plan geared towards winning and pushing the envelope, with the final touch just letting the boys down.
As you can see on the above chart, no team had a greater differential in its non-shot expected goal differential and shot based expected goal differential. And in some alternate universe where matches aren’t played and we use computers to determine who will become victorious, Atalanta would have finished 2nd on the season through non-shot expected goal differential.
From a possession based standpoint Atalanta dominated this year. An expected 36 goal differential is nothing to sneeze at – but when the shots don’t fall, it doesn’t matter how good the foundation is. And that’s when things can begin to rear their ugly head far too viciously.
Performance in Lopsided Matches
To further illustrate the point of how there is such a large gap between Atalanta’s non-shot expected goals and shot-based expected goals, let’s look at a few games a bit more closely. There were 380 Serie A matches played this year, and Atalanta had more matches than any other team in which it dominated the pitch; thus translated into an excellent non-shot goal differential. In fact there were 14 matches in which Atalanta’s performance should have translated into defeating an opponent by at least 1.32 goals (Top 75th percentile of matches based on non-shot expected goal differential). The Empoli game, fresh on everyone’s mind, was the most lopsided contest of the Serie A season – and we all know how that one ended. So how did Atalanta fare in the rest of these matches?
Six victories, five draws, and three defeats. Realistically Atalanta should be winning most if not all of the matches they dominate in such a fashion, instead La Dea is looking at a maximum of 19 dropped points. Even if the boys didn’t convert every victory, there are still double digit points being left on the table.
How does this happen? Underperformance in non-victories of nine goals cannot all be attributed to luck. There is definitely a latent issue underneath the hood that needs to be investigated. Spoiler alert – as everyone knows Atalanta had supreme difficulty in finishing this year.
Atalanta’s Finishing Woes
Atalanta has good strikers. Duvan Zapata and Luis Muriel consistently outpace their expected goal metrics every year, but 2021-2022 was a perfect storm for underperformance. On a macro level the entire team was below average, and maybe that’s the problem. A new stat for the vault is expected goals on target (xGOT), which essentially rates the probability a shot on target has of hitting the back of the net after weighing a few variables like goalkeeper positioning, velocity, body part of shot. So a shot that is right at the keeper may only have an xGOT of 0.10, but a shot that is headed for the top right corner may be closer to 0.75.
The stat is not all encompassing, as it does take into account shots off target or shots that hit the post, but is still a good indicator of finishing as a whole. Based on where this article is heading, it is probably unsurprising to most that Atalanta underperformed in xGOT more than any other team in the league this year. There’s several ways to dissect this stat.
First, Atalanta has the greatest differential between non-penalty XG and xGOT- turning 65.6 xG into only 52 xGOT. This 13 goal difference is by far the worst in the league. Even the other bottom feeding teams in the league like Empoli, Genoa, and Salernitana weren’t that poor.
Second, after adjusting for quantity (it can be unfair to compare teams with 60 xG to 30 xG through raw numbers), Atalanta still has the worst ratio between xG and xGOT. Lazio on the other hand, had elite finishing (outperforming its xG by 11%) singlehandedly making up for its shaky business, and now the Eagles are back in Europa League as a result.
And the strange thing is its not like Duvan Zapata, Luis Muriel, and Mario Pasalic had such woeful seasons it crated Atalanta’s chances of putting away goals. Zapata was below average this year (not egregiously), and Pasalic and Muriel both outperformed their xG targets (albeit just by a little bit). The problem lies in all the other players that Atalanta relied on to score goals this year. Over a full season pretty much everyone else was so poor at finishing that chances strikers would normally finish off simply ended in frustration.
Just look at all that additional underperformance. And before you realize it, it adds up to almost double digit goals down the drain, because what is sometimes Atalanta’s biggest strength can admittedly be its biggest weakness: interchangeability.
Looking to Next Year
Maybe that’s what La Dea has to change going forward. As addressed above the squad clearly has a good plan in place, and can oftentimes suffocate opponents. Thus Gasperini and company put up such a strong non-shot expected goal differential. However, Atalanta failed miserably in converting its great chances it into actual goals. This is where the tweak would have to be for Atalanta to improve next year. Without Robin Gosens, Atalanta isn’t blessed with lethal finishing in as many spots, but the squad is still set up to be more of an ensemble effort rather than having one or two focal point scorers.
Unless Joakim Maehle was hiding some finishing ability, this doesn’t seem to be a strategy Atalanta can continue to use – and the staff must go back to the drawing board to figure out how it can get its strikers into more favorable finishing positions rather than relying on goals to come from anywhere. Atalanta may have had 20 different scorers this year, but if it was at the expense of more overall goals (and more overall points) who cares! There’s still great parts underneath the hood, it may just take a few small tweaks to re-unleash the guys who are actually paid to score. Time will tell, but don’t sound the alarm bells yet! As always, Forza Dea!!!
