The new Audi Performance Index promises fans “new insights to debate and discuss the sport they love in a new way.” What is there to debate, though? The index rates Ignacio Piatti and Dom Dwyer as the top players in Weeks 1 and 2, with scores of 2102 and 1547 respectively. These numbers mean little in isolation, and the metric’s Web site is intentionally vague on additional context. Let’s find a way to make this better.
According to the FAQ, the Audi Performance Index utilizes near-ubiquitous OPTA data, and then calculates actions in the subcategories of Technique, Dynamics, and Efficiency (hey, like a luxury car!) There’s a desire to make this a thing, as Audi promoted the index heavily during last weekend’s game. Without understanding how the index influences on-field success, though, fans are left with a number that provides no unique insight on player performance. MLS fans have been here before with the defunct Castrol index, another black-box figure that generated confusion.
Rather than have the Audi Performance Index befall the same fate, here are 10 ways to improve the presentation of the information. I appreciate any quantitative pursuit of increased understanding of the game, and these topics – bucketed in the same brand attributes – would popularize or introduce ways to assess player and team performance.
1) Passing accuracy for through balls.
While resources like whoscored.com tally through balls and short pass/long pass accuracy, I’d love to see more information on who executes through balls well. We know that some through balls are a result of a defensive break down, and others are the result of outstanding vision and execution. A combination of pass type, pass accuracy, and defensive position could further distinguish who does this well.
2) Accuracy on extremely long passes.
Typically 25-yards is the dividing line between a sort pass and long pass. Why not take this a step further and evaluate those who consistently hit an excellent 40-yard pass? I’d like to better understand which defenders can consistently bypass a line of defenders, or which midfielders switch fields effortlessly.
3) Chance creation after beating a defender 1 v 1.
An evolution of the “dribbled past” stat. I see this as an opportunity to highlight “linked event” stats, and can help highlight useful creativity. Taking a defender out of the play is a positive, and following that with an attacking opportunity is even better.
4) Fastest successful counter attacks.
Statistics from major FIFA events already include a player’s top speed or distance covered. Audi can measure the speed with which teammates attack collectively. This could determine what teams are most dynamic in transition, and how many players they commit to a counter-attack.
5) How attackers manipulate defenders and space.
Are there attackers who make excellent runs that cause defenders to vacate space? Do teammates recognize this and exploit the newly created space? The location data at the beginning and ending of an attack should yield new insights into coordinated attacking sequences, as well as man-to-man or zonal marketing defensive tactics.
6) Measure point of attack variation.
Heat maps show a player’s actions throughout the match, but Audi should take this a step further. Measure which players can create from anywhere on the field, including distance from goal and width across the field. Do some players live for Zone 14? Who floats in to any available space?
7) Define pressing by field position and commitment.
Two years ago, Colin Trainor introduced a terrific metric to measure defensive intensity called Passes Allowed Per Defensive Action. Building on this work, Audi could determine how high a team presses, how long a team presses, how many players a team commits forward in the press, and how often the defensive pressure results in change in possession.
8) How well goalkeepers organize a wall on free kicks.
Location data should be easiest to collect on set-pieces, where most players start from a static position. Assessing free kick efficiency helps fans understand goals and saves better. Did a goal happen because a player made an excellent shot, or because the keeper failed minimize the shooting angle with his wall?
9) Defensive positioning efficiency.
This would be especially insightful with Patrick Vieira’s early NYCFC tactical plans. Audi could utilize its location data to answer key questions behind a team’s defensive shape. What teams have the best coordination with backline positioning and spacing between defenders? What teams keep things most compact between the midfield and defense? Does this positioning force the opposition to drift away from its attacking tendencies?
10) Possession efficiency and risk/reward ratio.
Jared Young recently evaluated MLS passing percentages and expected completion rates based on the type of pass. Why not overlay field position to this analysis? What players are indexing above league average when it comes to passing into the attacking third? In the box? Do some players, as Michael Bradley postulated, need to play riskier passes to help a team win?
There are countless other areas for exploration by quantifying on-field events. Any of the above would provide MLS fans with more understanding about the game than a single number from an unseen calculation.