Countdown to Champions League Final: How horrible is Saturday’s match up? Ridiculous numbers shed light.
It’s difficult to get too fired up about a Champions League final contested by a team that finished sixth in their domestic league and an opponent that just got blitzed in their domestic cup final. If the Champions League is, in fact, supposed to be about the best facing the best, this year’s final defies the definition. Few would argue that either Bayern Munich or Chelsea are amongst the world’s top five-or-so1 teams right now, but it would also be foolish to argue that these teams don’t deserve to be in Munich.2
It’s a bit unfair to compare the build up for this year’s final to last year’s Barcelona-Manchester United meeting, but there’s an obvious (and possibly refreshing) diminution in excitement. While it might help if Frank Lampard stepped up with a reputation-undermining scandal, the perceived quality of the matchup is tempering anticipation for a meeting of two of the world’s most popular clubs.
All that was a bit too nebulous for me, so I decided to combine a series of random numbers with spurious assessments, trying to determine just how bad this match up is. I looked at the 20 years of the Champions League and tried to assess, based on league finishes and league strength, whether we’d ever seen a “weaker” final. Trying to cover a number of angles, creating a number of different ways to rank the final match ups, looking at (click here to skip the boring stuff):
- combined table (ordinal) finish,3
- combined points, adjusted for the number of points it took to win the league,4 and
- combined points, adjusted for the maximum number of points possible.5
Then for each of these measurements, I also created rankings that attempted to adjust for league quality:
- combined table (ordinal) finish, with each value multiplied by league’s UEFA coefficient rank (ordinal),6
- combined points adjusted for league winner’s points, adjusted for how the league’s coefficient (points) compare to the top coefficient,7 and
- combined points adjusted for league winner’s points, adjusted for how the league’s coefficient (points) compare to a pseudo-maximum coefficient.8
Whether you want to look at raw, relative, or adjusted numbers, these quick-and-dirty assessments should cover the spectrum, but none of this should be taken too seriously. The goal here is to move from “Man, this matchup stinks” to something a closer to “OK, this is a slight more valid reason to think this matchup stinks.”
And by the six “measures,” Saturday’s matchup is the stinkiest in only one:
Table 1: Best, worst match ups by combined league rank
All we’re trying to see here is which finals have features the best league finishers. There’s no shortage of meetings of league winners (“Best”, where the score ends up being 2). This year’s final, however, is right there with 2000’s as the weakest, by this measure.
It bares noting La Liga in 2000 was amazing, with six teams within eight points at the top (Deportivo La Coruña took the title).
We account for this kind of clustering in the next measure, where we don’t look at league rank; rather, we take consider how close the teams came to winning the league. The numbers, below, are the combined percentages of the points each team earned divided by the league leader’s (multiplied by 100, to make pretty). 200 would be a the best possible score, one earned by those seven meetings of league winners.
Table 2: Best, worst match ups by percentage of league leader points
This measure tries to capture how far back of their league leaders each team was when they made the final. When Milan and Liverpool were competing against each other for Champions League, there weren’t meaningfully competing for their domestic titles. Interesting, the four finals rating “worst” by this measure have all occurred in the last six years.
The next chart is similar, but instead of looking at teams relative to their league leader, we look at total points available. There are a certain number of points out there at the beginning of the year. How many did the finalists grab? If you had two finalists with two perfect league records, the score would be 200.
Table 3: Best, worst match ups by percentage of maximum points
There’s that La Liga season again. In 2000, points were so evenly dispersed in Spain that Valencia (third, 64) and Real Madrid (fifth, 62) were still title contenders. To put that in perspective, this year’s Chelsea - never real title contenders - finished with 64 points.
To this point, we havent taking league strength into account. In these final three ... things ... I try to do so. First, I take the league finish rankings and multiply them by the league’s UEFA coefficient ranking (before combining the numbers). The best possible score here would be 3.9
Table 4: Best, worst match ups by combined league rank, UEFA coefficient adjusted
A meeting of teams from fifth and sixth-ranked leagues drives up the 2004 final. It didn’t help that Monaco was the third place team in Ligue 1 that season.
Staying with the idea of adjusting domestic results for league strength, we shift back to point totals but make our UEFA coefficient adjustment. That adjustment: take the league’s coefficient points and divide it by the leader’s coefficient points to create our “factor”. For example, if we’re talking about a team from the number one ranked league by UEFA, the factor will be 1; however, if we’re talking about a league that’s only accumulated half the points of the best league, the factor will be .5 and the team’s domestic point total will be downgraded accordingly:
Table 5: Best, worst match ups by combined league points (relative to leader), UEFA coefficient adjusted
In 1994, Barcelona finished well off the pace in a league that was miles behind Italy as UEFA’s best-rated. If you want strong teams from strong leagues in your Champions League final, 1994 Barcelona may be the worst finalist of the last 20 years.
And we’re finally at out final table. This one is like Table 5, but except using the best-rated UEFA league as the European standard, we shift to our theoretical maximum UEFA coefficient points.10 The goal here: assess teams’ absolute league results when adjusted for league’s absolute coefficient “quality”.
Table 6: Best, worst match ups by combined points earned from maximum, UEFA coefficient adjusted
Hold on a second: How did this year’s matchup make it into a “Best” list? For this ranking, it’s all about the UEFA coefficient. In historical terms, the rankings of the current top three leagues are very high. You get a matchup between the two of them, and it’s going to climb these charts. In 1994, even though Milan was from Europe’s top-rated league, the actual rating wasn’t that high, as far as coefficient-leading rankings are concerned.
So ... after all that anybody else ready for Saturday?!? Wasn’t this exciting? Yeah, I know. The numbers become a bit of a buzzkill after a while, but look at it this way: There was no buzz to kill for this weekend’s match, exactly the reason why we did this in the first place.
There is the assumption that this year’s matchup is a bit of a stinker, and it is. But there are have been a lot of stinkers in the past, no matter how you look at it. And as the prevalence of the Milan-Liverpool finals on these lists show, some the match ups the numbers see as horrible end up being among our most memorable finals.
Take a look at some of the raw data, if you’re into that kind of thing.
1 - Let’s go ahead and say Real Madrid, Barcelona, Juventus, Borussia Dortmund and Manchester City, with a good argument for AC Milan ahead of Bayern ... and most certainly Chelsea.
2 - It’s not that the argument’s wrong. It’s more that “deserve to be” there is just a weird way we discuss these things. It always leads nowhere. You’d be foolish the engage in that discussion. It’s too Baylessian.
3 - League finish of Team A added to League finish fo Team B (or for second place Bayern and sixth place Chelsea, the final value out be 8).
4 - League points for Team A divided by the league leader’s points, added to the same measure for Team B. I multiplied by 100 for aesthetics. If two league leaders meet, the score would be 200.
5 - League points for Team A divided by maximum league points, added to the same measure for Team B.
6 - This one’s just as it sounds. Chelsea’s sixth place finish and England’s first place coefficient ranking yield a value of six. Bayern’s value is also six (second and third ordinals). The match up’s final “score” is 12.
7 - This takes the league leader point adjustments we did for note 4 and multiply them by a similar coefficient leader adjustment based on league coefficient points and divided by coefficient leader points.
8 - Similar to what we did for note 5, but since it’s near impossible to come up with a maximum coefficient, the figure used as to adjust here is the highest coefficient seen yet (England 85.785 in 2011) divided by .897 (which is the closest any club has come to claiming full league points in the Champions League era). The goal is to scale the value in a manner similar to some of the non-coefficient adjusted measures.
9 - The first place team from Europe’s best league meeting the first place team from the second-best league or the second place team from the best league.
10 - Which comes out to 95.635, and is really not a theoretical maximum at all; rather, it’s a scaling agent.