How much does each stat actually contribute to the efficiency of a striker? It’s a common belief that pace has a lot to say, but what about physicality and shooting? In this article, we will examine that by applying statistical methods.
The data we will use is goals-per-match data collected from cards currently up for sale. As described in my previous article about the performance of in-form players, there are a few statistical issues that need to be considered for comparison reasons. To work around these issues, I decided to use EA’s recently released top scorer tables from all the top leagues. However, I removed a number of players where other factors may affect the results:
- Wingers, midfielders and CF’s to reduce the effect of position bias
- Players where less than 300 cards are up for sale to reduce statistical inaccuracy in the performance data
This leaves 32 players, below ordered after the recorded goals-per-match average when I took out the sample.
|M goals||Goals / match||Price (PS)|
What we want to know here is basically how the stats affect the player’s performance. In statistical terms, we want to know how performance correlates with the stats. Correlation is a measure of whether one set of data may be statistically related to another set of data. A correlation of 0 indicates that there isn’t a relationship, whereas a correlation of 1 means that there may be a direct, statistical relationship. In this case, we expect to see a direct relationship, meaning that the higher the correlation coefficient, the better.
Which stats should we consider as likely causes of these scoring rates?
Although the players in question are strikers, I started out by including all six aggregated stats and workrates. Following that, I removed the stats one stat at a time in order to see whether it affected the correlation coefficient. Stats that had a positive impact on the correlation coefficient were left in.
Not surprising, pace, shooting, dribbling and physicality turned out to have a positive impact, whereas defense and passing wasn’t relevant. With regards to workrates, I have assumed that high attacking and low defensive workrates are preferable. I have assigned the value 90 to the best workrate, 60 to the second best and 30 to the worst possible workrate. This is definitely a bit arbitrary, but it’s the best I can do.
In the chart below, I have plotted each player, using the sum of his aggregated stats + work rates and goals-per-match ratio as coordinates:
I end up with a correlation coefficient of .70, which indicates a strong uphill relationship.
By adding all the stats together, it is assumed that all stats contribute equally. This is not likely, and the real purpose of my analysis is to determine the individual contribution / importance of each stat.
Hence, I have assigned a weight factor to each stat based on it’s assumed importance. Following that, I have adjusted the weight factors until I wasn’t able to increase the correlation coefficient any further. In this case, adding the weight factors increases the correlation coefficient to .76.
Below I have inserted the weight factors, which produced the highest correlation coefficient:
- Pace .28
- Shooting .23
- Physicality .12
- Dribbling .24
- Attacking WR .05
- Defensive WR .08
The numbers above are interesting, because they show the relative importance of each stat. As expected, pace is the most important attribute, although shooting isn’t far behind.
The reason why it is relevant knowledge to every FUT-player is pretty obvious: Building squads involves compromises, and the numbers above provides solid knowledge regarding to what extent it makes sense to accept a reduction in one stat in order to achieve an improvement on another stat. As an example, I would definitely go for 1 extra pace point instead of 1 point better dribbling, but I would definitely pick +2 dribbling over +1 pace.
As for chemistry styles, the consequences are obvious: Hunter is by far the preferable stat boost for a striker, closely followed by Hawk. This is no surprise, but still – now we know for sure.
Having identified the relative importance of each stat, it’s possible to calculate a weighted stat total (multiplying each stat by it’s weight) per player in order to identify the best strikers in the game. Below, I have inserted the top 20 among the strikers in consideration here, but obviously not including players like Messi, Zlatan, IF Ronaldo and other informs:
|Giovani Dos Santos||1,400||77,4|
A note from UltimateTeamUK: Based on Christian’s calculation of the most important stats in FUT, you should try to cherry pick bargain players! By incorporating a team full of important statistics you should notice your team performs better relative to the coins you’ve invested!
Also with the Team of the Year (TOTY) market crash just around the corner it may be an idea to invest in some of the popular strikers listed above!