SV Elversberg have scored seven goals in their opening two Bundesliga matches, and they have done it from only 3.9 xG. That leaves them with a finishing efficiency of +3.1, the best in the Bundesliga after two matchdays, which is why this trip for Bayern München looks a lot less routine than most people would have expected.

Elversberg's finishing edge

The raw numbers are hard to ignore. Seven goals from 3.9 xG is a strong return over two games, and it tells you Elversberg are taking their chances at a rate above what the chance quality alone would suggest.

That does not mean the finish will last forever, and nobody sensible should pretend it will. It does mean Bayern are facing a team that has been ruthless in the opening phase of the season, with a +3.1 gap between goals and xG sitting at the top of the Bundesliga after two matchdays.

Vincent Wagner's side are not walking into this without belief either. As bundesliga.com put it, "Bayern will be doing all they can on Matchday 3 to prevent that from happening – but the new boys have a plan."

Bayern's early wobble

Bayern have not won their opening two matches of the season for the first time in five years, and that alone makes this a different kind of early trip. Their previous Bundesliga visit to Saarland came in October 1992, when they drew 1-1 with Saarbrücken.

There is also the broader early-season context. Bayern had scored in their previous 59 competitive fixtures before the Schalke draw ended that run, so the stumble is unusual by their standards.

Bundesliga.com's feature on the match summed up the main warning from Elversberg's side neatly: "Elversberg haven't needed an abundance of chances to score seven goals either, as they have an Expected Goals (xG) figure of only 3.9."

Bayern still carry the heavier squad and the bigger reputation, but Elversberg have already shown enough efficiency to make this a real test. The match on Matchday 3 will tell us whether that seven-goal start keeps travelling when Bayern come to Saarland.

Written by Jack Mercer with AI-assisted research, cross-checked against 1 outlet. How we work →