Some differences between the two leg format: MLS vs. UEFA competitions

Before the Christmas/New Year's break, I published a post at my HexagonalBlog site that ran through the numbers for the two-leg format in the first round of the MLS playoffs.  The format has come under some criticism since its introduction in the 2003 season, especially for having the higher seed team host the second leg instead of open the playoff series at home (which happens in all other American sports).  It did appear to me that the lower seed advanced more often than not, but I wanted to look at the statistics and determine whether that was actually the case.  I also wanted to compare the statistics with two-leg formats in other competitions, so I selected the UEFA Champions League.

The two-leg playoff system is similar to both competitions with one exception: in the MLS playoffs, the away goals rule is not in effect.  That is, if the aggregate score is tied at the end of the second match, the winner is the team who has scored more goals away from home.  So if a team knows that it is trailing in the away goal tiebreaker it must score an outright win, which means that fewer series end up going to penalties.  This is reflected in the statistics for both playoff series, which I'll return to later.

Collecting the data was straightforward yet time-intensive: I searched the results of the MLS playoffs between 2003-2009 seasons and the Champions League rounds during the same period (preliminary rounds, round of 16, quarterfinal, semifinal) and tabulated the number of times the first leg ended in a win for the home team, a draw, or a win for the away team.  From those matches, I tabulated the number of times the playoff series was won by the first leg host, won by the second leg host, or decided on penalties.  (For UEFA Champions League, I included occurrences for series wins under the away goals rule.) 

There are some commonalities between the two-leg format in MLS and UEFA competitions.  The first-leg host's chances of winning the tie depend on winning the first match outright; a draw swings the advantage to the second-leg host and an away win even more so.  The difference is that wins by the second-leg host (the higher seeded team) in that first match occurs more often in the UEFA Champions League.  There are also more positive results in the first leg match, which the table below shows:


First Leg Outcome MLS Playoffs UEFA Champions League
Home win 50% 45%
Draw 32% 25%
Away win 18% 30%

I think the statistics reflect the relative differences in strength between the participants in the US and European competitions; the European sides are able to convert the kind of matches that would have ended in draws in the MLS playoffs into wins.  (I am sure that there is a difference in the team performance in the preliminary rounds and the knockout rounds after the group stage of the Champions League, but I haven't conducted that kind of study at this time.)  Also, the European sides are more adept at converting any advantage gained in the first leg into a series win.

First Leg Outcome Series Won by 1st Leg Host Series Won by 2nd Leg Host Series Decided on Penalties
Home win 73% 24% 3%
Draw 22% 72% 6%
Away win 3% 96% 1%

And here is the performance summary of the MLS teams under the two-leg format:

First Leg Outcome Series Won by 1st Leg Host Series Won by 2nd Leg Host Series Decided on Penalties
Home win 43% 50% 7%
Draw 22% 67% 11%
Away win 0% 80% 20%

Now, there is a tremendous disparity between the number of two-leg series between the two competitions: MLS has had 14 such series since 2003, while the UEFA Champions League has had 381 during the same period.  That being said, there are still two big differences between the two competitions.  The first is that MLS sides are not always able to press home any advantage they obtain in the first leg match: the chance that the winner of the first-leg match will go on to win the two-leg series is much lower than in the UEFA Champions League.  (If the first match ends in a draw, the probabilities between the MLS and UCL are identical.)  The second difference is that MLS playoff series are more likely to be decided on penalties than the UCL series.  This illustrates the effect of the away goals rule in forcing outright results over the 180 minutes (or 210 minutes in rare cases).  The away goals rule was applied in 8% of all Champions League two-leg series during 2002-03 and 2008-09 seasons — not negligible, but not a large percentage of matches either.  If MLS would like to reduce the number of series that go to penalties, it would behoove them to introduce this tiebreaker.

The result in the following table makes sense given what has been presented up to this point, but is still counter-intuitive to those in MLS:

Series Outcome MLS Playoffs UEFA Champions League
Win by 1st leg host 29% 39%
Win by 2nd leg host 61% 58%
Penalties 11% 3%

So while MLS playoff participants may not believe so, there is indeed an advantage to hosting the second leg of the playoff series.  It is more likely that a MLS playoff series will be decided from the penalty spot, which adds a lot more uncertainty to the final outcome.  I am impressed that the two-leg series exhibits the same trends between MLS and UEFA competitions, and it would be interesting to see if that holds in other competitions as well.