As you may have encountered in our platform already, we do not show Contribution Ratings for every player. Broadly speaking, we provide Contribution Rating metrics when the following two base conditions are met.
- Data Quality - The player plays or played in a league for which we have a sufficiently large and reliable dataset.
- Sample Size - The player played a sufficient number of matches to draw reliable conclusions on his contribution to his team’s performances.
The Contribution Rating metrics in the platform are season based. Consequently, we can only calculate Contribution Ratings for a specific season in a given league, if the corresponding dataset is complete and reliable enough.
For example, when we miss a large number of matches for a particular season within a league, we do not offer Contribution Ratings for this league-season on the platform. The metrics would be misleading, because we would not capture the full picture.
For the league-season combinations that are considered to be complete, the Contribution Ratings model (just as all other SciSports models) makes use of some inhouse data cleansing procedures. In essence, the rare anomalies in the input dataset are discarded.
In total we offer Contribution Rating metrics for 319 seasons across 99 different leagues in 61 countries. For a complete overview of leagues and seasons, we refer to the attached spreadsheet, at the bottom of this pahge. (last updated on: February 6, 2020)
Of course, we are continuously checking opportunities for increasing the league coverage for the Contribution Rating metrics as well as other metrics. If you miss a certain league-season combination in the attached overview that would be worth an investigation according to you, please let us know via firstname.lastname@example.org!
Besides some checks and requirements on the completeness and reliability of the dataset for a league-season combination (see previous paragraph), the decision to (not) provide Contribution Ratings for a specific player is also affected by the number of matches (or more accurate yet: the number of minutes) this player has played during the same season in that league. We do not show Contribution Ratings for players who haven't collected enough minutes, because their impact is likely not sustainable or repeatable. As a scout, you should not draw any conclusions from metrics that aren't robust.
Let’s take a look at the situation of Divock Origi in the 2018/2019 Premier League season. With Liverpool's first-choice attack being almost undroppable and Daniel Sturridge rejuvenated, Divock Origi spent much of the opening few months of the 2018/19 season not even being part of the matchday squad. However, his situation changed when he came off the bench in the Merseyside derby in early December to make his first Premier League appearance of the season. Origi scored the winning goal in injury time and from that moment on he became a structural member of the matchday squad, mostly being used as a pinch hitter or first replacement of Firmino in the less important matches. In the minutes Origi played in the Premier League, he mostly faced the tired opponents (dying minutes of the game) or “weaker opponents” (when Firmino got some rest).
When we look at the players with the highest offensive contribution in the 2018/2019 season, our example player Divock Origi appears on top. Looking at the other players in the top 10 and the minutes they have played, it would not be fair to compare Origi to the others. Origi’s contribution is measured in matches against worn out or weaker opponents, whilst the contribution of the others is based on a wide range of matches. Of course, Origi’s contribution to Liverpool’s offensive play was definitely important in those matches, but it would not be correct if we would conclude that Origi’s offensive added value is higher than that of Sterling, Hazard and all the others in the 2018/2019 season.
To prevent these types of comparisons that could lead to the wrong conclusions, we examined what the lower boundary of minutes played should be, in order to enable a fair comparison. This led to a lower boundary of 700 minutes per league per season before we provide insights in a player’s contribution to his team’s offensive and defensive performance.
Manchester City FC
Manchester United FC
Tottenham Hotspur FC
Manchester City FC
Manchester City FC