By Jared Young (@jaredeyoung)
What do the LEGO Movie, HBO’s True Detective and the Bible’s Book of Ecclesiastes have in common? Everything. The world is obsessed with condensing this chaotic world into mid-afternoon snacks and we’ve reached the logical end of summarizing everything. There is the famous LEGO Movie song “Everything is Awesome” - which is of course wildly optimistic. There is True Detective’s infamous and terrible line in episode 5 where a prostitute claimed that “Everything is F*cking” - which is wildly disturbing and pessimistic. And there is the wisdom of King Solomon, who in the Bible’s Book of Ecclesiastes, said that “Everything is meaningless, a chasing after the wind.” According to Hollywood we now know that Stephen Hawking developed The Theory of Everything and there’s the classic sports diehard “winning is everything.” It sounds like everything is being reduced to well, everything.
It seems like soccer, or at least Major League Soccer, could use its own hyperbolic notion of itself. What exactly is everything in soccer? I’ve been calculating a PScore which attempts to calculate how proactive a team is with the ball. Do they try to possess the ball or are they more prone to a direct style of play? Success can be had with all style but it’s been important to commit to a style. But seeing success at the top and bottom of the PScore list got me thinking – what matters in soccer, at least in Major League Soccer in 2015?
First here is the updated PScore table for August. Orlando City and Montreal have maintained their position as the most proactive teams. Colorado took the lead at the bottom of the table and is now the league’s most reactive team.
|Rank||LastMonth||Team||PScore||Home Pscore||Away Pscore||H - A|
One characteristic of proactive teams is that they tend to pass the ball more than their opponent. Here is a chart of passes attempted by PScore. Not surprisingly teams that are more proactive have more possession and attempt more passes.
But what do passes get a team? Are they worth anything? Here’s a chart of passes attempted compared with shots attempted.
Um, nope. There’s definitely a positive relationship, but with an Rsquared of 6.4% there is too much variance to depend on passes earning a team shots. It’s safe to saying passing isn’t everything.
But do shot amounts even matter in a game?
Again it’s a positive trend as expected but a team actually can’t be certain they will score more goals simply because they attempt more shots. This chart also reveals just how unimportant passing attempts are for scoring goals.
But so far we’ve been looking at individual game statistics. What if we add up the season’s worth of statistics and compare those the bottom line: winning? Well, King Solomon mostly comes out looking the smartest. There’s still a lot of nothing.
Possession: meaningless, a chasing after the wind. It’s not easy getting a 0% Rsquared.
Ok, but what about a team that possesses the ball more in their offensive final third? Surely that will matter some? Nope.
Shots attempted will be meaningful now, right? I mean we’ve added up the whole season. Chase that wind.
Okay, but teams have to play defense too, so how about if we combine shots attempted with shots allowed?
Who here is betting on a 3% Rsquared? Not me. Shots are not everything.
It’s not until we actually look at finishing until a meaningful trend arrives.
Teams with a 10.8% finishing rate this year or higher all average 1.3 points per game or above. Only one team below that level is above 1.3 points per game. But even still a 1% increase in finishing rate is only worth 0.07 points in the table on average or 2.5 points for the entire season. It’s not huge but we’re on to something. What about defense?
The spread between a team’s offensive finishing rate and their opponents’ finishing rate is the strongest relationship. This metric is essentially the PDO metric which adds shots on goal finishing rate and a team’s save percentage together. Even still nearly half the variance in the table can’t be defined even by a thing as powerful as finishing.
We are faced with the reality that in this MLS season possession, passing and taking shots don’t matter in the least. We are left with the cold hard reality that finishing is everything. Defenses don’t even have to worry about limiting shot attempts, just keep the opponent’s finishing rate low. And don’t forget that lady luck continues to threaten to make all of this meaningless.
In prior MLS seasons possession has had a meaningful relationship with average points in the table. This year there is no such relationship and PScore can help us understand why. This year Montreal, NYCFC, and Orlando have performed near the bottom of the league but have been proactive teams. Meanwhile successful teams like Sporting Kansas City and FC Dallas have chosen to play more directly. Therefore, it’s reasonable to infer that it’s never really been the case that possession is important, at least in MLS. It’s simply been that a greater number of successful teams have chosen to win by gaining possession. This year there is a counter trend. Some good teams have chosen to play reactively and it’s worked for them.
The bottom line is finishing. It’s everything. Tactics don’t matter. Passes don’t matter. Even getting shots doesn’t matter. A coach must select a style that best suits the talent on the team, with the sole purpose of maximizing the finishing rate of the team and minimizing the finishing rate of the opponent.