Productivity and Gabriel Martinelli

Do Arsenal get enough output from the left wing?

I’ve lost count at this point of how many pieces of mine have started through my stumbling across a Twitter conversation, but here’s another one to add to the pile.

Gabriel Martinelli, as you may be aware, was relatively credibly linked to Bayern Munich over the weekend. That should, as I ended up stating directly at one point, tell you all you really need to know about whether Gabriel Martinelli is actually a good player. But we’ll get into that more in a bit.

The idea of selling Martinelli naturally stirred up some feelings. Some people lamented the idea, which led to some backbiting, which leads me to what I saw.

That’s right, I’m here to talk about the idea that Gabriel Martinelli is not productive enough off the wing for Arsenal! Particularly over the two seasons following his 15-goal outburst in 2022-2023 season, this discussion has persisted. Not everyone believes it, but a lot of people do.

“This aint good enough.” That’s a question worth digging into, isn’t it? Martinelli’s got 14 Premier League goals over two seasons (no pens counted), while Saka’s got 15. Trossard, it should be noted, has 20.

And that number doesn’t exactly blow you away at first blush. Mo Salah scored 20 not including penalties all on his own this season. Bryan Mbeumo scored 15, Luis Díaz 13 and Jarrod Bowen 12. So, are Arsenal’s wingers…unproductive?

One of the most obvious points to be made here is that cumulative totals are incredibly reliant on total playing time. This framework has been weaponized, intentional or not, against the likes of Šeško, Havertz, Eddie Nketiah, and it’s been done so again here.

Sure, Saka’s 15 and Martinelli’s 14 pale in comparison to Salah’s 32 or Bowen’s 25, but Bowen has played 14 full PL 90s more over the past two seasons (about 1,250 more minutes) than either Arsenal winger. Bryan Mbeumo has played about 30 more 90s! Salah is about 13 ahead.

When you break it down per minute, the list of more productive wingers gets short. Among those who’ve played at least 40 total PL 90s in the past two seasons, here’s the ranking for np goals per 90:

1. Foden

2. Salah

3. Trossard

4. Bowen

5. Son

6. Diaz

7. Brennan Johnson

8. Mbeumo

9. Semenyo

10. Saka

11. Callum Hudson-Odoi

12. Martinelli

FYI - The gap between Mbeumo and Martinelli here is that between taking 283 minutes per goal and about 309 minutes per goal.

So, Arsenal have three of the top-12 most goal-scoringest (counting this as a word) in the league, according to math.

This should probably also include assists, which are key to the whole scoring thing, so for the record:

• Saka jumps into fourth in the PL in goals + assists per 90 (behind Salah, Son, Foden) over the past two seasons

• Trossard falls down into eighth

• Martinelli falls one spot to 13th

On underlying expected goals numbers, things look pretty similar.

• Bukayo Saka is 2nd in the PL in npxG+xAG per 90 minutes, again behind only Salah

• Trossard is 9th

• Martinelli is 13th

The good news is that after a slower start (likely influenced by Arsenal’s overall injury issues), Martinelli closed out 2024-2025 averaging 0.61 npxG+xAG per 90 minutes from his post-injury return in March through the final game of the season in May. Those numbers put you into top-five conversations instead of top-10, and would be exceptional to see continue.

That should give some idea of where Arsenal’s wide attackers rank in terms of recent output, now just how easy would it be to upgrade on that? I put together a couple of quick graphs that I think help paint a solid picture.

In the first, you can see where Martinelli and Saka line up against some of the more popular shouts and links from recent memory on a per-90 basis.

Saka is Saka, while players like Nico Williams and Jamie Gittens have matched or exceeded real goal contributions per 90, but not the underlying. In those cases, there’s some reliance on the player himself or his teammates remaining clinical to keep the numbers up, meaning that could be something to keep a close eye on.

Another thing I’ve been looking at more and more lately is output through the framework of usage, which at least in theory should measure efficiency of a player’s involvement.

Adjusting these same numbers to be per touch in the attacking third does some interesting things…

That’s weird, right?! Because I know a lot of Arsenal fans, and not even unreasonable ones, who’d call Martinelli wasteful. And here he is scoring or assisting more per touch, as well as generating more underlying xG, than Bukayo Saka.

That’s very likely influenced by Saka being a high-touch player and facing heavy numbers near the box, of course, and that’s the kind of context that watching games can add to metrics like this. But seeing Martinelli on par with someone like Kvaratskhelia, whom many Arsenal fans swooned over in the spring, is a very interesting thing.

Rodrygo, as I’ve said in my winger ranking piece, is another who is very high-usage when he plays, so it’s no surprise that his per-touch numbers are not as significant, particularly when you factor in the fact that this past season was a down one for him.

Perhaps all of this is a lot of hot air to you, but it feels worth consideration to me. And Bayern Munich’s directors certainly seem to think Martinelli’s is a name worth considering among the likes of Diaz, Barcola, Mitoma and Leão. So perhaps upgrading on this player, after all, isn’t quite so easy.