The MLS Table hates Atlanta United

When soccer fans pull up a league table, their attention is almost always immediately drawn to their team’s place in the table. What happens next is normally some dubious mental math, something like “ok, looks like we’re 6 points out (of top 4 if you’re following Premier League, or from the 6^th place spot in the East if you’re an Atlanta United supporter). So, we need to win twice and have the other guys lose twice! That’s all!” The efficacy of that exercise is questionable, but it is what it is.

It’s absolutely what I do when I pull up the table, and I can’t imagine I’m alone. So, even though last Saturday was immensely enjoyable as an ATLUTD supporter, it’s still not fun to look at the Eastern Conference table right now. Here’s how it looks on whoscored.com in the standard format (I added the all-important “Red Line” that’s seared into my brain).

I think Brittany Arnold on Atlanta United Matchweek even said something like “Don’t panic. There’s still plenty of season left.” That is comforting I guess, but I think there are more compelling arguments for not panicking. My favorite one is that we’re looking at the wrong table. In this post, I’ll suggest a new table to use, and then we can all make our own decisions about whether to panic or not.

Because of the league’s notorious unbalanced schedule, the official MLS “standings” table is almost useless until the season draws near it’s end. There are a couple reasons for this, and both directly impact where Atlanta falls in the standings at this point.

First there’s the extremely obvious fact that the teams have all played different numbers of games. Everyone above Atlanta in the table has played more games than the expansion side – with 3 of the 8 having played 2 more. The smart folks over at AmericanSoccerAnalysis.com correct this by dividing the number of points by the number of games played to get a significantly better sorted table:

There’s a ton of other cool data they include in their tables, but the single biggest improvement is simply the “PPG” column, far right. Again, I’ve added the red line myself because I *need* it.

And, hey! Would you look at that? Atlanta are sitting pretty, above the red line now! One of the only downsides with this view is that if I’m only looking at PPG, it’s not immediately intuitive to me just how “far ahead” Atlanta is from Philadelphia (I know it’s a narrow lead, but how exactly should I think about 0.03ppg, aside from unwinding the math back into the standard format?). My point is that I still want to do that dubious mental math exercise I referenced above where I say “if only we won some games and they lost some games, than we’d...and on and on.” But, there are other ways to further balance the table.

“Win your home games” is something that I hear every week about MLS. Sometimes its “win your games at home, and try to pick up points on the road” or some other variant. The conventional wisdom in MLS for some time has been that because of the salary cap, there is a general level of parity which doesn’t exist in most soccer leagues around the world. And because of this parity, home field advantage is huge -- even bigger than it already is in sports. And this shows up in the results. If you look back through the last 5 MLS seasons, teams average 1.78 PPG at home, and only 0.96 PPG away.

So with this in mind, and selfishly, because Atlanta has played more away games than most teams (although Toronto and NYCFC have also played 7), I propose we attempt to further balance our table to normalize for disparities in the home/away schedule to date. I’ve recalculated the table against the baseline “win at home, draw away” mantra. It should be noted that if a team were to actually win all their home games and draw all their away games, they’d almost assuredly win the supporters shield (with 68 points), so our table is going to have teams at the top with values close to zero (meaning they’re tracking on this elite shield-claiming pace), and all other teams will show up as negative. Specifically, how it works is that I award a team zero points (the baseline) if they win at home or draw away. If they draw at home they get -2 (a 1 point draw at home is 2 off of the baseline of a 3-point win), and if they lose on the road, they get a -1 (a loss on the road is 1 real table point off the general pace of drawing your away games). You can also get positive points for exceeding the baseline (winning on the road). I’m calling this “points off the pace” (POP). But for the love of God, please help me come up with a better name. Here it is, my baby:

It’s not perfect but it has some advantages (aside from Atlanta moving up a couple spots). For one, because the baseline (winning at home, and drawing away) is zero, this naturally balances the table for differences in games played by the different teams -- there’s no need to divide the results per game or anything like that. If a team has played more games than everyone else and they’ve “taken care of business” by winning away and drawing at home, they won’t be given any advantage in the POP table (just the baseline of zero points). Conversely, if a team has played more games than the rest of the league but they’ve squandered the advantage by drawing at home or losing on the road, they’re penalized relative to the rest of the league. In Atlanta’s case, they’ve missed out on some points at home but overachieved on the road and the POP table suggests that Atlanta is 4^th best in the East relative to the general win at home, draw away truism.

The other advantage to this view is that while imperfectly, I can still sort of “do the math” just by looking at the POP column. I can say “ok, Atlanta is 1 excess point behind Chicago. If we keep the overall pace, we’re going to need them to lose on the road once to be level with them, or have them draw at home once to overtake.

I want to make clear that this calculation is in no way “predictive.” It has no direct information about the opponents each team has left on their schedule or how good those teams are, or their current form, or recent injuries, or anything like that. It’s purely backward looking. A metric with which to say “here’s where my team stands based on what’s happened so far.”

There are certainly drawbacks, and I considered adding in some other strength of schedule elements, for instance lowering the baselines when a team plays the heavy favorites in each conference (Toronto and Dallas), and I have that available if there is demand for it, but ultimately I think this one is the most logical. One drawback is that not every home game is created equal (e.g. David Villa is about to show up in the A), but this at least levels the table against that “win home, draw away” pundit speak you hear every week. Let me know your thoughts. I post tactical and numbers-based stuff twice weekly at atlutdinsight.wordpress.com.