Hy, I was studying this data set on Kaggle (https://www.kaggle.com/karangadiya/fifa19) for my machine learning problem. Idea is to from features of player to count overall rating. Ok, first we need a little bit of feature engineering to see what feature will be significant. One of this is country of player such as Croatia or Brasil.
Main idea is that country of player wouldn't have much impact on overall rating, but what if developer with intention give Brasil player much higher rating than Croatia player.The same idea can be used in club(Maybe equal player in Real Madrid and Dinamo Zagreb have higher rating if he play in RM).
So the problem is to show that country of player has no connection to overall. To try to solve this problem I use all Croatia and Brasil player and use T-test on their overall rating. The test show that Brasil player have higher rating, but this can be also reason that Brasil have better players. Another idea was also use T-test again but overall / value for Croatia and Brasil but I didn't get any clear answer.
Can you give idea how to approach the problem?