The Opta Index is a prestigious performance measure used to assess
English Premier League (EPL) football players. Although the Opta model
is proprietary, the general structure uses a multiattribute collection
of subjectively weighted pitch measures that either rewards or penalizes
a player with a potential range of points based on the quality of his
game performance. In addition, the specific set of measures used depends
upon player position: forwards, midfielders, defenders, and goalkeepers
each have their own unique set of measures even though there might be
some overlap. Although the player's Opta Index is calculated for each
game, it is the cumulative "grade card"—the final Opta Index calculated
at the end of the thirty-eight game EPL season in May—that is of
particular importance. The index, along with the large array of player
pitch data, is commercially distributed to the EPL clubs and appears in a
wide variety of television and print media outlets. This paper proposes
an alternative to using the full set of Opta data by identifying those
specific pitch actions that form a statistically significant
retrodictive linear regression model for the 2007-2008 EPL season.
Additionally, the importance of evaluating pitch actions historically
assumed to be clearly pertinent measures—such as goals allowed per game
for the goalkeeper—will be not only be appraised from a statistical
viewpoint, but also from a practical perspective. The Opta Index is a
prestigious performance measure used to assess English Premier League
(EPL) football players. Although the Opta model is proprietary, the
general structure uses a multiattribute collection of subjectively
weighted pitch measures that either rewards or penalizes a player with a
potential range of points based on the quality of his game performance.
In addition, the specific set of measures used depends upon player
position: forwards, midfielders, defenders, and goalkeepers each have
their own unique set of measures even though there might be some
overlap. Although the player's Opta Index is calculated for each game,
it is the cumulative "grade card"—the final Opta Index calculated at the
end of the thirty-eight game EPL season in May—that is of particular
importance. The index, along with the large array of player pitch data,
is commercially distributed to the EPL clubs and appears in a wide
variety of television and print media outlets. This paper proposes an
alternative to using the full set of Opta data by identifying those
specific pitch actions that form a statistically significant
retrodictive linear regression model for the 2007-2008 EPL season.
Additionally, the importance of evaluating pitch actions historically
assumed to be clearly pertinent measures—such as goals allowed per game
for the goalkeeper—will be not only be appraised from a statistical
viewpoint, but also from a practical perspective.
