The most decisive football players of 2022 season

Hello !

Welcome to this 8th football post. For better visibility, I advise you to look at my posts on a computer (for the size but also because the colors differ on a phone that has a dark theme like me.
If the theme is light, only the size makes it difficult to read).
Especially this one because it will be an interactive graphic, and on the phone, it will be very complicated or even impossible to visualize correctly.

Today, a dataviz on the most decisive players of the 5 major leagues during the 2022-2023 season.
In the usual dataviz, either we are used to seeing the raw statistics, that is to say this player was decisive n times, or this player was decisive on average n times by 90 minutes.
Except that sometimes this last statistic is misleading. Indeed, let's take an example : a player plays 3 matches (with a playing time of 90 minutes by match) and scores 3 goals in total, with a hat-trick (3 goals) in a single match.
This therefore means that he did not score on the other 2. However, if we take the statistic to /90 min, we will say that the player was decisive on average once per match, but in reality he was only decisive once !
So, if we are very strict with the mathematical definition, that does not mean that he was decisive once by match, but in the collective imagination, that is what many people will think.

So, to try to visualize who is really decisive over a large number of matches, and to see the difference with the statistic / 90 min, I decided to make an interactive dataviz integrating these 2 statistics.
You can move your mouse cursor over a point on one of the two graphs, and see where the same player's point is on the other graph.
Some player informations will be displayed through a tooltip.
I recommend a 90% zoom, because at 100% I sometimes have a small display bug on the tooltips.

We can observe that some players have much higher values ​​on the variable / 90 min, such as Lois Openda :

This is also visible on the previous dataviz visible via this link.
His performances reduced to 90 minutes are truly impressive, making him one of the most decisive players in Europe, even though he is only 23 years old.
This is also partly why he was recruited after his season in Lens by RB Leipzig for the price of 38.5 million euros (great financial added value for RC Lens who subsequently bet on Elye Wahi, purchased for 30 million euros).

You still have to pay attention to this "new" decisive variable by match, because sometimes it can be interesting, but sometimes the interpretation can be bad. For example, a player who plays 5 matches completely and scores in each of them, and plays 5 more matches with a playing duration of 5 minutes in each of them, the value for the new variable will be medium then that it's going to be huge for / 90 minutes variable.
However, it is rather logical that the player has difficulty scoring in 5 minutes of playing time. This will have an impact on the value of this variable when in fact, this player will have been very decisive in his matches given his playing time in each of them.

Finally, this variable is more suitable for players who have regular playing time in each match played (and big value...because regular time of 5 minutes, the variable is no longer really of interest). To know which player satisfies this condition, we can refer to the value of the relative standard deviation (this is a statistical indicator showing the dispersion of the data relative to the average thereof. In other words, it shows the extent of variability relative to the mean).
The smaller it is, the more regular playing time the player has had in the matches in which he has participated.

Let's take two examples: Bruno Fernandes (16 decisive actions) and Edin Dzeko (13 decisive actions), with playing times by match.

We immediately see that Fernandes very often plays the same duration by match, unlike Dzeko. This is why when we compare the values of the rsd indicator, we have respectively : 0.01 and 0.61 !
And when we put this in relation to their statistics /90 min (0.43 vs 0.68) and /match (0.41 vs 0.27), we come across the warnings mentioned earlier.

Thanks for reading! I hope it was clear and you liked it.

You will find the code below by clicking the github link button.

If you have any questions or remarks, I invite you to create an account (it's free) to write a comment, or simply to be notified of a new post in the future !

See you soon for new content 👋