Followers of MLS know that LA's point total for the season was the second most dating back to 2005, when San Jose (now Houston Dynamo) racked up 64 points. Something jumped out at me when I looked at the historical data of team finish for seasons 2005-2010, and I think it suggests that perhaps LA's run this year was more impressive than Houston's.
When looking at LA's 59 point haul this year, what impressed me was the balance in where they earned their points - 29 at home and 30 on the road. This lead me to look at the historical split in teams' points totals, and I found that indeed LA is special.
Taking the 2005-2010 season data, I normalized it for the number of games played and teams in the league by calculating the per cent of available points earned by each team. I then divided the number of points earned at home by the number of points earned by the road. Plotting the data on a modified version of a production possibility frontier chart yields the plot below (click to enlarge).
The plot illustrates the balance between teams with a better balance in where they earn their points (vertical axis) versus their total points (horizontal axis). Each axis is plotted as a percentile - being higher on the vertical axis or further right on the x-axis indicates a higher percentile and thus a better score.
The two dots in the upper right corner are LA (2010) and San Jose (2005), with LA being the point on top. While San Jose did score more overall points, they did so by getting a greater proportion of them at home vs. away (in fact both Houston and LA scored the same amount on the road). One should also note that all other teams at the 80th percentile or greater for total points earned ranked 90th percentile or lower on the balance metric. LA, at the 98th percentile for the balance metric, is head-and-shoulders above the rest in the balanced approach in which they earned their points.
The above conclusions only deal with the observations made to date. Where would the Galaxy's performance fall against a long term average when future seasons are added?
We can get a good indication of this if the data set for home vs. away goals is normally distributed. It turns out that it is not, but a simple mathematical transformation of the original data does produce such a normal data set. Taking the square root of (Away Goals)/(Home Goals) produces a distribution with the following properties.
LA's score on the transformed metric is 1.017. Generating a Z-Score and converting it to a percentile shows that only 5% of the teams would have a point distribution more skewed towards away points than LA did this year. No one else even comes close to matching their balance (a score of 1.0 on this metric). Any that do are often in the lower half of the table, and the closest team in the upper half of a table had a score of 0.96 (biased towards home points and twice as far away from the balanced 1.0 score as LA). The only other teams to have a score greater than 1.0 - and thus biased more towards away points - on this metric were DC United (2010, last place) and Chicago (5th in 2005).
So LA took the more difficult road to their Supporter's Shield. Perhaps some would knock them for not defending the Home Depot Center more consistently, which certainly penalized them in the race to best San Jose's point total record. However, I feel LA deserves special commendation for their balanced approach (especially considering RSL's 2010 point total was biased nearly 2:1 in favor of home points).
Now if those pesky playoffs didn't penalize successful teams, LA might have pulled off the rare Supporters' Shield/MLS Cup double. More on that later...
- Posted using BlogPress from my iPad
Location:NE 38th Pl,Kirkland,United States
No comments:
Post a Comment