Ah, the Europa League: bimonthly trips to the freezing climes of Eastern Europe, small midweek crowds at home, meager prize money, the chance to sell a third jersey… these are the trappings of UEFA’s second-class continental competition. Fans of English clubs love their European nights, but the endless rounds of the Europa League can take their toll on squads already stretched by two domestic cups. Is it worth all the trouble?
I decided to do a simple exercise to find out. I looked at Premier League standings going back to the 1999-2000 season, which was the first one in which the UEFA Cup absorbed the European Cup Winner’s Cup. For all the non-promoted teams, I took the difference between their final league position and their position in the previous year. Then I used a regression to see whether playing in the Champions League, UEFA Cup/Europa League, and/or Intertoto Cup was correlated with the difference in final league position.
The regression included binary variables for qualifying for the Champions League, UEFA Cup/Europa League, and Intertoto Cup. A team could have values of 1 for the UEFA Cup/Europa League and Intertoto Cup if they qualified for the former through the latter. For teams that dropped from the Champions League into the UEFA Cup/Europa League, I only gave a value of 1 for the Champions League. This was because continuing in the Champions League would have implied roughly the same commitment of time and effort. By contrast, the Intertoto Cup took place at a totally different time of year.
The constant in the regression was slightly greater than zero (0.45), implying that non-promoted teams’ league positions worsened slightly, on average, from year to year. This made sense. In the Premier League seasons since 1999-2000, all three promoted teams were never relegated immediately. As a result, some of the non-promoted teams in my sample, whose league position in their previous seasons was never worse than 17th, had to drop down the table.
There was virtually no effect associated with the change in league position from qualifying for the Champions League. In other words, the top teams in the Premier League pretty much stayed there from year to year. The effect associated with playing in the Intertoto Cup implied a small improvement in league position, but it was far from being statistically significant. The effect associated with playing in the UEFA Cup/Europa League was very significant: a drop of 1.9 league places, on average, with a 95% confidence interval stretching from 0.5 to 3.3 places.
This result seemed to imply that playing in the UEFA Cup/Europa League was a bad idea for teams that cared a lot about their positions in the Premier League. That’s not to say it was a bad idea from a financial perspective. The difference between, say, finishing fifth and finishing seventh was £1.5 million in Premier League merit payments last year. By coincidence, the fee paid to teams for playing in the Europa League group stage was also about £1.5 million, with anywhere from several hundred thousand to a few million more for teams in the knockout rounds. That wasn’t Champions League money by any means, but it could pay a couple of weeks’ wages.
So was the Europa League just a chance to earn a little extra cash at the cost of Premier League places? Not necessarily. The regression explained very little of the variation in the difference between teams’ league positions from one year to the next – only about 3%. There was a strong chance that other variables were more important, and these initial effects could be misleading.
To take a stab at explaining more variation, I created another variable intended to represent luck. I used a new regression to find the effect on league position associated with goal difference. Then I calculated how much “luck” each team had in each season by looking at how much their league position appeared to outperform their goal difference. In general, we’d expect teams that were lucky in one season to have a worse league position in any other season. So in a second version of my original regression, I added luck in the previous season as an explanatory variable.
This made a big difference. The effect associated with luck was statistically very significant, at 0.82 with a 95% confidence interval between 0.59 and 1.04. In other words, teams that gained a place in the table because of luck lost about four fifths of a place, on average, the following year. The constant only dropped slightly, to 0.36, and the effect associated with qualifying for the Champions League was still small and statistically insignificant. But the effect associated with playing in the UEFA/Cup Europa League weakened quite a bit, to a drop of 0.79 league places, with a 95% confidence interval between -0.52 and 2.09. The chance that playing in the competition was associated with a worse finish was still about 76%, but there was a possibility, albeit small, that the opposite was true. Either way, from a financial perspective the Europa League was looking like more of a sure winner.
(Though it’s now a footnote to history, adding luck into the regression also strengthened the effect associated with playing in the Intertoto Cup. On average, the effect was an improvement of 1.34 league places, with a 74% chance that the effect was truly positive. Perhaps tuning up over the summer helped teams to perform better in the Premier League.)
Overall, this broader regression explained 21% of the variation in the change in teams’ league positions from year to year. Naturally, adding other variables could further alter the effects associated with playing in the European competitions. Yet even if the Europa League really did erode a team’s final position in the Premier League by 0.79 places, I doubt teams would opt out of the competition. Fans like to have more chances to watch their teams, and even the UEFA’s lesser tournament carries some prestige. Throw in a net gain in prize money and extra shirt sales, and it looks like a no-brainer for budget-conscious clubs. But then again, what if playing in the Europa League also worsened finishes in the FA and League Cups? I’ll have to get back to you on that….