Why can’t Miguel Almiron finish chances?

Attacking Engine

Let’s start with something positive. Miguel Almiron has created more offense for Atlanta United than any other player in MLS for any other team in MLS. He’s the best. This reads like an opinion or a take, and that’s fine. It is, however, supportable with some objective if imperfect metrics. Almiron has scored 8 goals and 9 assists. But if we’re just talking about driving offense, and not necessarily results, let’s look at the underlying numbers. If you start with expected goals as a measure of the combined quantity and quality of goal scoring chances taken by a player, and add in expected assists which essentially links back the expected goals created from a shot to the player who played the assist, you get a sum that serves as a decent measure for the direct chance creation of a player. American Soccer Analysis calculates these metrics based on on-the-ball event data from MLS matches. If you sort ASA’s table on this metric (xG+xA) for past seasons you get names like Giovinco, Keane, BWP, Villa, and Wondolowski but also Valeri and Higuain. In 2018 Josef Martinez and Miguel Almiron sit atop this list at 20 and 19 “expected goals and assists” worth of chance creation, respectively. A portion of of the xA underlying each of these scores is attached to the other player: this is what some might call Fusion.

Recognizing however, that there is more to an attacking move than the shot and the final ball, as Tata Martino so eloquently and slightly angrily discussed here, it might behoove us to cast a wider net and credit anyone involved in a sequence that led to a chance being created. This is what ASA has done with what they call “xG chain” whereby every fraction of xG created from a given shot is attributed evenly to each player involved in the buildup (including the shooter and final passer). If you sort ASA’s tables on xG chain, you’ll see Miguel Almiron standing tall, atop the league at 24.5 xGC. This might be why your pulse quickens when he receives the ball and turns.

End Product

But I get it — this xStuff is not a great comfort when you’re tired of walking out of Mercedes Benz Stadium feeling deflated after the team dropped points (though Cookout is). In truth, Almiron’s 8 xAs have appeared in the corporeal form of real life assists 9 times, but his 12.8 xGs (or whatever they’re called) have only birthed 8 goals. And underneath the covers, it doesn’t look any better. If we remove shots from dead ball situations (notably 4 PKs), we’re staring at a borderline NSFW 4 goals on 8.2 xG worth of open play shots. He’s put the ball in the net from open play 4 times in 87 attempts for an infuriating 4.6% conversion rate. The average open play shot is converted 11% of the time in MLS.

To be fair, he’s a midfielder right? So they take deeper shots. So, maybe, maybe he’s — NO! Of that 11% open play average, shots from outside the box are converted in MLS 3.5% of the time and shots taken inside the penalty area score at a rate of 16.5%. How is Miggy faring in these buckets? He’s 4.4% from downtown - that’s nice! - and from inside the box, let’s see — these are the important shots — let me just pull this up, and it’s:

Miguel Almiron has scored 4.8% of open play shots taken from inside the penalty area compared to a league average 16.5%

2 goals on 42 shots inside the box. In fact across his entire shot distribution, he’s doing about average (not great) in all situations except for open play shots inside the penalty area:

This probably isn’t a surprise if you watch all of the Atlanta games. He’s been wasteful with really good chances around the goal many times this season. But you can’t watch all the MLS games all the time, nor all the soccer games in the world, and so to get some perspective, it’s helpful to turn further to some statistics.

To put this into perspective

If the season ended today, Almiron’s uderperformance as measured by goals minus expected goals would be the 6th worst in MLS since 2011 (the furthest back the data set goes). It would also put him in good company with recent 40M pound Everton signing Richarlison, who drastically underperformed his expected goals last season. So those of you saying that he better start scoring goals if he wants to make it to Europe, stop it. Stop it Joe.

Does he have a history?

What I’m more interested in is whether this poor form in front of goal is illusory — something that just happens from time to time in soccer — or if it will persist, an actual inherent weakness for the player. This is a tough nut to crack, and I’m not sure we can. We could start by looking at Almiron’s limited history. In 2017, he put up 9 goals, 8 from open play, and of those 8, 6 from inside the box. Compared to his expected goals, it looks much more normal than that of his form this year:

As it turns out, this isn’t all that surprising either. Finishing metrics rarely persist from season to season, at least as measured by the difference between goals and expected goals. But I don’t want to discount it completely. After all, the guy has taken 42 shots inside the box this year and many of them have missed, and those misses are seared into our collective memory. I could imagine there is a mental aspect to this after a player starts missing some big chances and has to answer questions to the media every week about a future in Europe. He’s also playing alongside a very goal hungry and record chasing Josef Martinez. There’s likely an element of this and an element of chance and elements of other things.

I don’t have access to expected goals numbers for his time at Lanus, but I reached out to someone that does: Howard Hamilton (@soccermetrics on twitter). He took a quick look at the 2016/2017 Primera season where Almiron played just over a thousand minutes, and took 14 open play shots (1.3 shots per 90) generating about 1.4 worth of expected goals. It’s such a small sample to work with, but he scored zero of these shots. Combining the 16/17 season with the 2016 Primera Season (a completely different season, i know), Almiron converted something like 6% of shots taken in the box while at Lanus. I guess this is kind of low.

It does feel like cherry picking to some extent however. If instead, we took all of the data we had prior to 2018 and stood at the beginning of 2018 and tried to make some guesses about the year to come, we’d probably sit there and go “OK, in 2016 and 2017 across 3 seasons of data in two different countries, he averaged 1.4 non-PK shots in the box per 90 minutes and converted them at a rate of 10%. Sure, he’s a more natural midfield creator but he played further forward for Atlanta in 2017 and finished his chances quite well.”

OK, let’s do this

So, should we really be declaring him a poor finisher - like, as a feature of his game? Did he forget how to do it in the offseason? Over in England, Christian Benteke underperformed his expected goals by something like 8+ last year. Don’t get me wrong, Benteke is trash, but I’ve never thought of him as a terrible finisher. Like, is it the case that Almiron is both a bad finisher and also better than Christian Benteke?

I think Miguel Almiron is absolutely going through a finishing slump, and to the eyes it’s one that is not simply bad luck. What I mean is, it’s not that he has run into a buzz saw of amazing keeper saves and clearances off the line and shots off the woodwork (he has had 2 though). There’s some of that, but there’s also just some accuracy issues.

Let’s try something. Some expected goal models (like ASA’s goalkeeper model) use the shot placement as a factor in determining a value for every shot. This is not the standard expected goals you hear about because in using shot placement data as an input, it excludes all off target shots which, in a sport somewhat starved of discrete data points drives the predictiveness of the model down. But that aside, it might offer us a rough glimpse at how Almiron’s doing with his shots. The data shows that while Almiron has created 6 xG worth of open play shots in the box using the traditional xG model, based on where he has shot the ball (often missing the goal), it’s more likely for the goal keeper to have allowed 3 into the net. As an aside, this is flipped for his outside the box shots where accounting for his placement boosts his expected goal tally from 2 to 3. Anyhow, his shooting in the box obviously leaves something to be desired, and there’s a difference between the misfortune of Bono saving Almiron’s late breakaway chance, and the “better luck next time” of Almiron shooting wide when 1v1 in the box in the first half of the same game. Both below:

Soccer is hard

Beyond these things, there’s a certain amount of volatility that you expect in shooting success simply because most things in soccer are failures. Players go through hot and cold finishing streaks because there are something like 25-30 shots in a game but only a couple of goals. Keep in mind that in many games where Josef Martinez scores multiple goals he also misses a sitter. Josef’s record chasing goal run is mostly about making creative and incisive runs, the kind of runs that demand service, and racking up these chances. And it’s to a lesser extent about converting the chances that come.

I think Almiron will turn it around, to at least average creative midfielder levels. In the mean time, I will delight in the elite rates at which he a) creates for his teammates, b) finds good shooting opportunities, c) tests the back line with sprint after sprint, creating space for others and d) pressures the opposition to win the ball back. I would prefer that he take fewer shots from outside the penalty area (though he is admittedly a decent long ranger shooter). Ultimately, I doubt that #clinicalfinisher will ever be tagged to him. He would be a bona fide cheat code if that were the case. An argument that Miguel Almiron is the best player in MLS is still a reasonable one. If a team in Europe were to somehow negotiate a cut rate transfer fee because of his 2018 finishing woes, they would be absolutely stealing from Atlanta United.

As a fan though, a fan that’s looking forward to the playoffs and not backwards to the Houston game, you want the players that are creating panic for the opposition because they’ll likely continue to. I want him to do that in the playoffs. I wouldn’t trade this guy for anyone else in MLS.