MLS, as many of you know, is trying (more or less) the traditional two-legged playoff system this year for the first time. This has led to much debate about who truly has the advantage in a two-legged scenario, whether you'd rather have the first game at home or the second game at home. A few years ago, I studied the Champions League and UEFA Cup competitions and found that whether a team hosted the first or second leg made virtually no difference to their chances of advancing in the tournament. I don't have those numbers on me as I type this, but it was so close to .500 either way, it was scary. Now that the A-League has had two legs in its playoffs for a few years, I re-ran the numbers and found that, in this very limited sample at least, there was a good-sized advantage to having the second game at home. The A-League went to two legs in its conference semifinals and conference finals in 2000, then went to two legs for its entire playoffs (except the one-off Final) in 2001. So we're looking at four years worth of data (which isn't much, admittedly). Anyway, there have been 32 two-legged ties in that period. The Game 1 host has advanced 12 times out of 32 (37.5%). The Game 2 host has advanced 20 times out of 32 (62.5%). This despite the fact the W-L-T percentages of home teams in either leg were fairly consistent (16-8-8 for First Leg home teams, 15-10-7 for Second Leg home teams). I believe (someone check me if I'm wrong), that in the A-League, the higher seed has the choice of hosting the first or second leg, while in MLS, the higher seed automatically gets the second leg. This could skew the results a bit, I would think. It may be that when you have two teams that haven't been "seeded" by virtue of having played a season's worth of games against basically the same opposition, and where you presume that they're closely matched by virtue of the fact that they both won their way into a very difficult competition (in the case of the CL or the UEFA Cup), that the two-legged tie reduces the impact of the advantage of hosting one leg versus the other. It may be that in a closed league, where one team has proven over the course of a season that it's superior to its opponent (because you usually see a highly-seeded team take on a low-seeded team, like a 1-4 or 1-8 scenario), that the quality team will win out more often than not, and therefore the advantage does lie with the team hosting one leg versus hosting the other (and by this very limited sample, it appears the leg to have is the second one). I have heard of (but not seen - I believe Phil had it) a study conducted that encompassed many years worth of results in Europe that showed a definite advantage to hosting the second leg. That would be more credible than my original analysis of a year's worth of CL and UEFA Cup results (as well as this limited study), as long as the unseen study didn't use results from before the War or something. If they're recent results, the more data, the better. So right now I'm leaning towards believing that there's some sort of advantage (maybe it's largely psychological) in having the second leg at your place. MLS obviously thinks there's an advantage, giving the higher-seeded teams the second leg this time. That doesn't mean there can't be or won't be upsets, but it looks as though history is with the Fire, the Revs, the Quakes and the Wizards in the first round. The numbers: 2000 (6 two-legged ties) Game 1 host went through 3 times Game 2 host went through 3 times Higher seed won 4 of the 6 ties 2001 (10 two-legged ties) Game 1 host went through 4 times Game 2 host went through 6 times Higher seed won 7 of the 10 ties 2002 (10 two-legged ties) Game 1 host went through 3 times Game 2 host went through 7 times Higher seed won 5 of the 10 ties 2003 (6 two-legged ties) Game 1 host went through 2 times Game 2 host went through 4 times Higher seed won 4 of the 6 ties Total (32 two-legged ties) Game 1 host went through 12 times Game 2 host went through 20 times Higher seed won 20 of the 32 ties
Hi Kenn, Believe it or not, I was finally able to locate the info: http://groups.google.com/groups?hl=en&lr=&ie=UTF-8&oe=UTF-8&selm=3bc1c87a.1488557@news.cis.dfn.de Just to add, I looked at the ten most recent Mexican League playoffs, where the higher seed plays the second leg at home and the tie-breaker is a bit different: if the teams are level on points and goals scored, then the higher seed advances. There, the breakdown was 48-19 in favor of the higher seed, 71.6%.
The other part of this, that I neglected to mention, was that there was a huge advantage in winning the first leg. In the original study, in the 1999-2000 UEFA Cup, the teams that won the first leg went through 30 times and failed to go through only twice (6 times the first leg was a draw...and in 5 of those 6 ties, the team that hosted game two eventually went through). In that same year's Champions League, teams that won the first leg went through 99 of 120 times (82.5%). Which had me questioning my logic: If it seems obvious that winning the first leg makes you, if not a lock, certainly a very strong favorite to advance, wouldn't you rather have that first leg somewhere you'd be more likely to win? Like at home? Wouldn't you want to maximize your chances? BTW, teams that won the second leg also went through 81% of the time. In the CL study, 17 teams lost the first leg at home. Only 1 went through. 26 teams won the first leg at home. 22 went through. It'll be interesting to see how this plays out over time. If anyone knows of anywhere to get more data, it would be interesting to study further.
Just to be a smartass: how about the teams winning both games? Adjust the calculation for that, and then maybe home advantage means more.
What would the percentages be if one of the teams just won the two games outright. In that case, in hindsight, home field advantage doesn't mean squat.
So take out the instances in which one team won both games, and then see what the home-field winning percentage is, or who advances out of the tie? I'm not sure that'll do what you think it'll do. To me, that sounds like taking out all the games in which the Galaxy lost this year and saying they went 9-0-9 in those games. That's nice, but I'm not sure that tells you anything about anything. But, then again, I'm not very good at math, so it may very well be that you have a point.
Update: Adding the results of the 2004 A-League playoffs (the two-legged part of which is just completed). Since the advent of the two-legged playoff system in 2000, there have been 38 two-legged ties. In only 14 of those ties did the team that hosted the first leg advance (37% - right about what it was before). Second-leg hosts have advanced 24 times (63%), including two this year that hosted the series overtime (Seattle) and penalty kicks (Syracuse). First-leg hosts are 19-11-8 overall (.605). Second-leg hosts are 17-12-9 overall (.566). Three of the six ties were won by the higher-seeded team in 2004, bringing the all-time total to 23 of 38 (.605). New numbers: 2000 (6 two-legged ties) Game 1 host went through 3 times Game 2 host went through 3 times Higher seed won 4 of the 6 ties 2001 (10 two-legged ties) Game 1 host went through 4 times Game 2 host went through 6 times Higher seed won 7 of the 10 ties 2002 (10 two-legged ties) Game 1 host went through 3 times Game 2 host went through 7 times Higher seed won 5 of the 10 ties 2003 (6 two-legged ties) Game 1 host went through 2 times Game 2 host went through 4 times Higher seed won 4 of the 6 ties 2004 (6 two-legged ties) Game 1 host went through 2 times Game 2 host went through 4 times Higher seed won 3 of the 6 ties Total (38 two-legged ties) Game 1 host went through 14 times Game 2 host went through 24 times Higher seed won 23 of the 32 ties
There's good evidence that home advantage in the second leg of a two legged tie bestows an added advatage on a team even after you've stripped out things like team strength/away goals rule etc. Take the English FA's playoff system which has been in force for 15+ seasons. First thing to deal with is the seeding.The strongest team by finishing position plays away first & at home subsequently. Looking at the league points gained & relating this to ability,the home sides in the second leg are on average the better side by around 0.15 of a goal.Not a huge difference,but a difference never the less. You can then relate this advantage to the expected win/draw/loss aggregate result after both legs.You'd expect the better teams(at home second) to win 40% of the aggregate games,draw 28% of them & lose 32% of them. Actual figures are 54% wins,20% draws & 26% loses. So they do considerably better. The reasons include.Tense,tight first legs.Having watched quite a few playoff games neither team wants to lose so early in the contest,goals are fewer(down to almost 2.0 goals per game),this inturn increases the chances of a first leg draw,which naturally favours the home team in the second leg. Playoffs by & large are littered with terrible football & horrible to watch if you've a vested interest in one of the teams.The finals are even worse. The same effect exists in European games,Champions league,uefa cup & cup winners cup. First point to make is that the random draw should eliminate any bias due to the better team being awarded the home leg in the second game. However as stated home teams in the second leg still win more often,qualifying around 54% of the time. This could be down to the home side in the second leg hosting extra time &/or penalties.So you need to look at the win/draw loss records instead. Still the advantage persists. Home teams in the second leg are ahead on aggregate 46% of the time,behind 38% of the time & level 16% of the time. T.