ABSTRACT
Previous research on the advantage experienced by soccer teams playing the second leg of a knock-out confrontation at home yielded ambiguous evidence. Some studies confirmed the well-established soccer myth that this advantage is substantial while others did not find any significant evidence. We contribute to this literature by analysing all ‘non-seeded’ two-leg confrontations in the UEFA Champions League and the UEFA Europa League between 2010 and 2017. We find that playing the second leg of a knock-out confrontation at home is not associated with a substantially higher chance of proceeding to the next stage of the tournament.
Disclosure statement
No potential conflict of interest was reported by the authors.
Notes
1 Beside professional soccer, scholars have investigated determinants of success in other professional sports (see e.g. Pic Citation2018a, Citation2018b).
2 We limited our summary to peer-reviewed articles published in English. However, we are aware of one non-English research article by Pic and Castellano (Citation2016) covering the second leg home advantage.
3 The round of 32 of the UEFA Europa League and the round of 16 of the UEFA Champions League are organised as two-legged confrontations as well. The drawing of these rounds, however, is not random but depends on the performance of the teams in the previous group stage. Including the information of these stages, which has been done in earlier contributions mentioned in our introduction, can only be done in a credible manner in case one is able to control for the relative strength of the teams. Therefore, we only included these data in the context of robustness checks.
4 The indicator of Baert and Amez (Citation2018) defines relative strength as the natural logarithm of the quotient of the home and away teams’ UEFA team coefficient for that season plus 1 (to avoid division by 0 for teams who did not participate in one of the two European competitions during the five previous seasons).
5 Including this extra indicator is possible since (i) the correlation between both indicators for relative strength is reasonable (r = 0.452) and (ii) the multicollinearity diagnostics yield variance inflation factors lower than 2, which is substantially below the threshold value of 10.