xGenius: Expected Goals and the Science of Winning Football Matches

by James Tippett

Arno Michels

Rainer Schrey,

Ralph Hasenhüttl

Henry first learned how to exploit market inefficiencies by trading corn and soybean futures

Tom Werner,

Michael Edwards, the Director of Technical Performance.

Ian Graham came knocking. The team’s Director of Research

In that game against Mainz, Ian Graham’s printouts demonstrated that Dortmund took 19 shots compared to their opponent’s 10. They worked the ball into Mainz’s penalty area on an impressive 36 occasions; their opposition managed only 17. Dortmund missed a penalty in the 70th minute, only to score an own goal four minutes later. Overall, Dortmund created 2.13(xG) to Mainz’s 0.93(xG). Based on these xG figures, there was only a 10 per cent chance that Dortmund would lose this game. They dominated by every conceivable measure, except the final score.

Barry Hunter,

who was installed in Liverpool’s scouting department.

Another key cog in the Liverpool backroom staff was Michael Edwards,

analyst for Prozone, a budding football data company. The business operated

Deepak Ravindran and his wife Emilie Vanpoperinghe

Dr Ian Graham, former Director of Research at Liverpool

Michu fell victim to what is known as ‘regression to the mean,’ another statistical phenomenon which originated with Francis Galton.

The key is to find the misshapen fruit.

Starlizard, Smartodds and consultancies such as Twenty First Group

The professional poker player Caspar Berry argued, ‘Whenever someone innovates in business or in life, they almost inevitably do so by accepting a negative metric that other people are unwilling to accept.’

ShotQuality is a basketball analytics consultancy which helps teams gain a competitive advantage by providing cutting-edge insight into, well, shot quality.

Simon Gerszberg, the founder and CEO of ShotQuality, was first introduced to this xG way of thinking at Colgate University. Dave Klatsky

The nearer to goal, the better: This one is obvious, and we’ve already seen how it’s affected the shooting distance of teams over the decade following the advent of xG. The more central, the better: Shots from the central part of the penalty area are more valuable than those from tighter angles. Feet are better than heads: All other things being equal, shots taken with feet are much more likely to result in goals than headed shots. Crosses are hard: In general, crosses are more difficult to convert than ground passes, through-balls, and shots after dribbles. The clearer the sight of goal, the better: Attempts with fewer defenders between the shooter and the goal are of higher value.

The shot maps for Brentford strikers over the last decade or so generally show a cluster of shots in and around the six-yard box, with only a handful taking place from outside the box.

William the Conqueror.

it doesn’t tell us a great deal about how the ball arrived at these dangerous positions in the first place. This is what xThreat seeks to do.

A player making a pass from F4 (0.01) into G10 (0.03) has increased his team’s chances of scoring by 2 per cent - or 0.02(xT).

player successfully dribbling the ball from C10 (0.03) to D11 (0.14) has increased his team’s chances of scoring by 0.11(xT).