I wrote about the strengths and weaknesses of using expected goals (xG) analytics in handicapping soccer last summer. Since that time, Brighton and Hove Albion have become the poster child expressing the limitations of relying too heavily on these metrics. Bettors banking on the Regression Gods to finally help the Seagulls see more of their shots reach the back of the net likely find themselves in trouble now.
As I wrote last summer: “Expected goals is a metric that determines a statistical probability on every scoring chance a team generates in a match. In this adventure of quantitative analysis, similar scoring situations are logged to determine a scoring probability from a deep data set in a way similar to measurements that predict the accuracy of an NBA shooter attempting a 22-foot corner 3-pointer. Shot attempts that have an empirical success rate of 35% or higher have been categorized as Big Chances. By reassessing a soccer match from the expected goals (xG) and expected goals allowed (xGA) given the activity and nature of all the shot attempts in a match. If xG analysis offers a better evaluation regarding how a team is playing, then it could provide a more precise way to measure subsequent action.”
Brighton played at West Bromwich Albion yesterday (February 27th) as a -0.5 goal-line road favorite. For handicappers relying almost exclusively on xG, that match may have looked rather tasty. The Seagulls may have been only 4 points above relegation land 16th place in the EPL table, but their expected points generated from a dissection of their xG and xGA for the season projects them as the fifth-best team in the English Premier League. West Brom, on the other hand, was in 19th place in points and dead-last 20th place in xPTs. Easy win for Brighton, right? If those bettors then looked at the xG results after that match, they might have started shopping for their new beachfront property. The Seagulls generated 3.28 xG against the Baggies while surrendering just 0.73 xG. The most likely score given that activity is a comfortable 3-1 win for Brighton. The actual score? West Brom 1, Brighton 0.
Perhaps that was yet another statistical aberration. Just like last week, when Brighton dominated Crystal Palace by a 3.03-0.27 mark in xG but lost, 2-1. Just like two matchweeks ago when the Seagulls outclasses Aston Villa by a 2.44-0.44 mark in xG but settled for a 0-0 draw.
I like to refer to the gambler’s expectation of outlier numbers returning back to a normal a call to the Regression Gods. The Miami Dolphins’ defense was not going to continue to bail out Tua Tagovailoa’s meager passing days in his rookie season by forcing multiple turnovers week-after-week-after week. When called, the Regression Gods eventually arrive. But these Gods never promised to show up promptly — and we need to keep our bankroll for when they finally make their triumphant return in the pursuit of justice.
Sometimes these underlying numbers are not simply outliers due for regression. Sometimes these numbers are descriptive. To paraphrase former NFL head coach Dennis Green, sometimes the numbers “are we thought they were!” (“and we let ‘em off the hook!”). Perhaps Brighton has scored only 27 goals despite their xG projecting that the typical team typical players would score 37.85 goals precisely because the Seagulls are a roster consisting of below-average players! As I wrote in the summer: “Expected goals attempt to determine the most likely outcomes. But not all outcomes are created equal. Lionel Messi is going to score more goals than Glenn Murray dribbling up the left-wing and talking a shot from 30 yards out.” Well, Aaron Maupay may have replaced Murray as the Brighton striker this season — but he is still no Messi.
Don’t get me wrong, I love xG — and incorporating expected goals analysis has improved by handicapping in soccer and hockey (where similar principles apply). We just should not become zombies to these numbers — it will drive us to bankruptcy. You wanna be an analytics fundamentalist and exclusively following the betting advice at Football Outsiders when betting the NFL? Kiss your bankroll goodbye in about a month. The most successful handicapping incorporates a variety of tools in the proverbial toolbox.
A final tip regarding xG: use these numbers to illustrate the prospective floor and ceiling regarding a team’s potential. Brighton’s xG promise did pay off on February 3rd of this month when they upset Liverpool by a 1-0 score. They won the xG battle by a 1.32-0.97 margin — so this was not a fluky victory. Perhaps one lesson regarding the handicapping application of xG is this: underperforming teams in xG make dangerous underdogs but unreliability favorites.
Best of luck — Frank.