Showing posts with label Major League Soccer. Show all posts
Showing posts with label Major League Soccer. Show all posts

Saturday, June 18, 2011

Sporting KC and Home Field Advantage in Soccer: A Luchametric Quickie

By MJD

Last night Sporting KC got its first home goal and its first home win, beating a surging San Jose club at Livestrong Sporting Park 1-0. We had predicted a SJO win, what with their lead-leading 1.54 Goals For per Game average going into the match, and their excellent recent form: 4 wins and a draw in their last 5 matches.  But KC looked better than we've seen them all year and prevailed in what we thought was a tight match.  In thinking about what we had missed in our analysis of the match, we realized that we had forgotten the most fundamental truth about professional soccer and about wagering on professional soccer: the 50/25/25 rule.

What's that you say? The facts are that in top leagues across the globe, including the EPL and the MLS, all things being equal, the home team will win 50% of the time, the away team 25% of the time, and the teams will draw 25% of the time.  That this is so is just part of the deep fabric of the game, decreed by the soccer gods long ago.  There have been various reasons put forth for this phenomenon, to travel demands on players, familiarity with the pitch, to the energy of the home crowd supporting the home team.  But we'll save any in-depth discussion of home field advantage for another post.

Here's the quickie.  Going into last night's match, KC had played 12 games, 11 of them on the road. In a normal season, half of those 12 games would have been at home, where (again, all things being equal) they would have won a full 50% of those matches (let's call it 6 home games just to make our back-of-a-napkin analysis easy).

Applying the 50/25/25 rule, KC's point total from 6 home games and 6 road games would be 10.5 pts from the home matches (3 wins and 1.5 draws) plus 6 pts from the road matches (1.5 road wins, 1.5 road draws) for a total of 16.5 pts total, not the 10 they were sitting on going into last night's match.

The point is that playing on the road for 11 games to open the season has distorted KC's record considerably, by about 34% or more.  This means they are a considerably better club than they've been given credit for.

Playing on the road in any league is tough, resulting in fewer chances created, more chances allowed, more mental mistakes, etc. KC has more home games left in their schedule than any other club.  Their PTS/GM will inevitably go up.  They will beat clubs at home that are substantially higher than them in the standings.  The only caveat to the 50/25/25 rule is that if two teams are not equal in strength, then the percentages get adjusted to the point that with two very uneven teams, the numbers change to 65/17/17 or even higher.

Our claim is that KC probably belongs in the middle of the pack, not down at the bottom with the likes of New England, who have no excuses.

Friday, July 23, 2010

Luchametric MLS Power Rankings Updated: Dallas you say?

Current Luchametric MLS Power Rankings
By MJD
Dallas sits atop our Power Rankings.  We've spent the last hour or so trying to figure out if there is a mistake somewhere.  There isn't.  Before you blow a gasket in disbelief, note that the LM is based on production per gameAnd if you compare Dallas' PTS/GM to the Crew, RSL and LA, the four teams are much closer than might appear just from looking at the single table.  Plus you have to factor in that Dallas has the best recent form, which we're perhaps weighing too heavily.  We need more time to look at that possibility.

The R-Squared when we regress the LM to point production in the league currently stands at .76.  This means that our metric explains 76% of the differences among the teams in terms of point production.  That's pretty good, but not as good as the R-squared for 08 and 09 combined, which is .83.  And our R-squared for point production in the EPL is around .90.  So, we're still examining why the MLS metric is not as sharp as the one we developed for the EPL (both are virtually the same).

Further, we'd point out that Dallas has only lost 13% of its games this year.  This puts them second in the league behind LA at 11%, and makes them over twice as good as RSL (26%) and well below the league average of 38%.n  They currently lead the league in draws.  In short, they are very, very, very tough to beat. They have drawn or won 87% of the games they have played thus far.  Our metric treats draws and performance in draws as significant evidence of a team's power--what we call "production."  The American proclivity to ignore draws as nothing more than "kissing your sister," we think ignores what draws tell us about a team. As such, Dallas gets more credit for how well it plays in wins, losses AND draws, which probably accounts for their place in the standings.  If they can maintain their rate of production, they should start getting some clean wins.  But it could also be that they've peaked at mid-season, and by the end of the year will drop back down in our rankings.  We'll see.