Elyounoussi Evaluated

The stock of Elyounoussi continues to rise within the Celtic support on the back of a run of 6 goals in 5 matches, and a sparky centre forward performance in the famous Lazio triumph.

I will declare I have been hot and cold on the Norwegian. My personal bias tends to be against players I perceive give the ball away carelessly or make bad decisions consistently.

So, I have been anticipating him reaching reasonable sample data size with some interest.

For this analysis I am going to use two comparators. Firstly Forrest. He is a player widely regarded as playing the most productive football of his career, and he plays a similar role in the team albeit on the right flank. Secondly, I will take Sinclair’s data from 2018/19. After all, he is the player Elyounoussi has effectively replaced.

The Norwegian Stokes?

Let me initially deal with my own confirmation bias.

Elyounoussi does indeed complete less passes per 90m than the other two (25.27) and less accurately (71% – Sinclair is 85% and Forrest 81%).

So, there is something in my perception. But, of course, creative players who put risk into their passes will be less accurate. Indeed, Elyounoussi’s assist to Forrest for the opening goal against Lazio was a highly improbably pass but ultimately hugely rewarding. As I will show you, that assist and the actions leading up to it, perfectly encapsulate the player’s style.

Elyounoussi’s assist rate is a fraction better than the others:

Elyounoussi 0.54 assists per 90m

Forrest 0.52 assists per 90m

Sinclair 0.21 assists per 90m

However, his Expected Assist rate lags behind what he is achieving, as does Forrest’s. Forrest is expected 0.39 assists per 90m and Elyounoussi 0.331. Sinclair also overperformed but to lesser degree (0.174 per 90m).

The Norwegian lags behind Forrest in chances created, managing 1.79 per 90m whilst the Scot creates 2.15. Sinclair’s creativity at 1.29 is lower still.

None of this is to bash his efforts and I would encourage you to see this through the lens of appreciating the wonderful form Forrest is in. Elyounoussi is not far behind but because he cost Southampton £16m and Forrest was a Celtic kid, there may be some underappreciation of what Celtic already have!

Goals

There is little doubt Elyounoussi has increased Celtic’s goal threat.

With 0.63 goals per 90m he is averaging more than a goal every other game. Forrest, amidst the scoring form of his life is maintaining a superb 0.47 per 90m average for a winger. Sinclair managed 0.34 last season which is perfectly respectable if not downright good.

I will caution that Elyounoussi may be in the middle of a “hot streak”. Which is to say, his expected goals numbers don’t quite live up to his actual productivity. Recall his first Celtic goal against Cluj deflected into the corner. Not a repeatable strategy.

Forrest is perfectly hitting his xG value of 0.47. That is to say it is reasonable to suggest he may continue to score at the current rate if he continues to get the same type of chances.

Elyounoussi is exceeding his xG value of 0.528. But that is a very high xG for wide attacker! So, whilst he may see a regression is scoring, it won’t be by much. One note of caution though. Elyounoussi is scoring from 70% of his shots on target – that is highly unlikely to be sustainable. Forrest’s 33% is a more “normal” rate.

Elyounoussi is posting high xG numbers despite having less touches in the opposition box than the other two (4.64 – Forrest is 6.19). And despite taking less shots (Elyounoussi 2.59 per 90m; Sinclair 2.68; Forrest 3.09) and being the least accurate (only 34% of his shots hit the target whilst the other two manage 46%).

Finally, Sinclair had an xG of 0.517 last season – had he actually scored at that rate would we even have Elyounoussi? Which is kind of why xG is important and makes it more of a mystery why the Englishman hasn’t featured more.

Scoring Contribution

All this means that Elyounoussi is contributing 1.16 goals and assists per 90m. Striker numbers. But hearteningly, his xSC of 0.859 is on a par with Forrest’s 0.861. With Sinclair posting 0.69 xSC per 90 last season, it is clear that the current wing attackers are the most likely to be the most productive.

Ok, What Else?

There is one more striking (pardon the pun) statistical indicator Elyounoussi is posting.

Defensive Action Success Rate. This is the metric I primarily judge centre backs on. It is proving to be a reliable indicator of defensive quality (Van Dijk scored 84% and this hasn’t been bettered).

Forrest (22%) and Sinclair (26%) have the numbers you’d expect from attacking wingers. They generally lose challenges as opposed to winning them.

Our erstwhile Southampton reject is, meanwhile, scoring 41% by this metric. Quite simply I have rarely seen such a number for an attacking player (unless they are called Christie).

The Norwegian is a grafter. In microcosm, his effort to dispossess Milinkovic-Savic before setting up the Forrest goal illustrates his desire, speed and work rate.

Summary

The data serves to suggest Lennon is right to have Forrest and Elyounoussi as his first choice attacking width. At the same time, completely discarding Sinclair makes little sense.

Whilst the Norwegian is shiny and new (and expensive), we need to appreciate that by keeping up with Forrest, it should highlight just how productive our home-grown hero is.

Finally, Elyounoussi brings a work ethic and desire to the team in keeping with the more “all action” style Lennon favours. In short, it’s great (and thoughtful) recruitment.

Let’s enjoy him whilst we can.