Have the Philadelphia Union been unlucky this season?

It's hard to keep track of luck over the long term. Does it even out or are some teams just unlucky while others just seem to get all the breaks? The Union have no doubt had their share of bad breaks this season. It started with an early string of injuries, including two goalkeepers getting hurt within a 45 minute span, in practice no less. There have also been a series of questionable calls: the early foul in Salt Lake on Maurice Edu that resulted in a game-tying penalty kick, the "no handball" against Chicago, and then last week a series of no calls in the box that might have given the Union penalties. They were certainly the kind of fouls that were called against Barnetta in his opening match against New York.

Then there is the kind of luck that just happens in the normal flow of the game. Take that most recent New York Red Bulls game at home. The Union outshot the Red Bulls 17-9 yet walked away with a 3-1 loss. Bad luck or deserved outcome? If you plug those shot figures into a binomial distribution model and assume the two teams finish shots at the same rate they have all season, their expected goal distribution looks like this.

The Red Bulls had just a 7% chance of scoring three goals from nine shots. The penalty kick helped for sure. The Union had a 31% chance of scoring just one goal. Multiplied together that outcome will happen just 2% of the time when the Union win the shot battle 17-9. That seems pretty unlucky. Let's come back to that game later.

But has this example of bad luck been happening all season? It's hard to remember game after game if these types of outcomes are evening out. One way to look at this is to use expected goals models to simulate what should have happened in each game. Expected goals measure how many goals a team should have scored based on the shots taken and shot quality of each shot. If we assume that these expected goals outcomes were the "right" outcomes then we can see if the Union were lucky or unlucky.

There is one issue with expected goals models and that is they assume all teams are average, and we know that's not the case. So I've adjusted the expected goals outcomes from American Soccer Analysis based on season-long finishing rates to better assess the "right" outcome. So a team's season-long performance should be factored into the game-by-game outcome below. That column is labeled "Expected Points Adjusted." I've also assumed that expected goals outcomes with less than 0.5 goal difference between the two teams should end in a draw. Here are all of the results for the Union this season.

Note on how to read the table: In the opening match the Union earned 1 actual point against Colorado. American Soccer Analysis said the expected goal outcome was 1.35 to 0.13 in favor of the Union so the expected points are 3 points. Even after that outcome was adjusted by the Union's season-long shooting percentage and shooting percentage allowed, they still earned the 3 points.

When looking at the totals it appears the Union are actually performing just slightly better than expected. If they were performing at just an average level on offense and defense they'd actually have a 1.4 points per game average. As it stands they are just over one. If you look at that New York Red Bulls game that 17-9 difference in shots actually only resulted in an expected goals difference of 1.29 - 1.17 in favor of the Union. The Union just didn't take quality shots in that game, while the Red Bulls earned a valuable penalty kick, and the most fair result was a draw.

Looking at "luck" in this way doesn't take into account the injuries or bad calls I mentioned above. That data is hard to track and compare against other teams. Given the Union haven't had many players miss games for international duty I don't think the Union have had to deal with an unusual amount of missing players. For example, San Jose was missing four players for international duty in the last game. As for the bad calls, I can only leave that up to the fans to determine. Were there other calls that were in the Union's favor?

From this analysis it clearly doesn't seem to be a story of being unlucky. That might be disheartening for Union fans but at least they can resist the temptation to listen to the end of the year rhetoric suggesting that with better luck, things should get better.