Orlando are unquestionably better than they were in 2019. But just how good are they? Unfortunately, the numbers suggest the hot start probably isn’t sustainable.
If a team surprises everyone out of the gate, it’s normal to ask whether the performance is sustainable, or whether they're primed for a quick and angry regression to the mean.
Unfortunately, the obnoxious answer is generally the correct one: teams that wildly outperform expectations are usually getting lucky. Sometimes, everyone misses something, and the change is durable. But more often, the law of averages takes over and things pretty much revert to normal.
Right now the team that has everyone asking questions is the Orlando Pride, who surprisingly sit atop the NWSL table. The Pride finished dead last in 2019, and then missed most of 2020. While most prognosticators expected them to be better this year, I’m not sure anyone thought that meant topping the table.
Several folks have written pieces explaining this turn of events. I like those pieces, and I encourage folks to read them. Many of the specific points they make are persuasive. For one thing, the Pride have improved their roster. They’ve also settled into a system that works. To that point, North Carolina coach Paul Riley recently took a dig at them for not playing “any football.”
Lots of people called that sour grapes, and maybe it is. But it’s also basically correct. Orlando are no longer trying to do anything fancy, and that’s a big part of why they’ve found better results. And speaking as someone who loves watching all styles of football, from 'real' to 'so pragmatic it makes your teeth hurt,' I say more power to them.
All that said, I feel obliged to offer a bit of sour pessimism to balance the story. Not only do I not expect the Pride to spend much longer at the top of the table, I would actually be surprised if they maintain anything close to these kind of results. To my eyes, Orlando still look like a mid-table team at best, and maybe not even that.
The underlying numbers suggest Orlando have been lucky
Why am I so down on Orlando? It’s actually pretty simple: they haven’t actually played that well in 2021.
I know they currently have 10 points. Obviously, that’s great. But if you want to know how well a team will play over the next 20 games, the topline results over the first four games are by no means dispositive, and can actually be quite misleading.
The Pride have won three games out of four and drawn the other, but carry the smallest possible goal difference that these results could produce—just +3. That’s good, but actually behind both Portland and North Carolina.
Running up the score in a game that’s already effectively won isn’t as *important* as scoring the winning goal in a close game, but if you want to know a team’s true ability, there’s not all that much difference between the two types of goals. The reality is that a team’s record in close games includes a huge dose of luck, and there just isn’t very much evidence that some teams are genuinely better at grinding out close results. Basically, if you want to know how good a team really is, and could only look at the points total or the goal difference, I’d prefer GD every time.
That said, it’s not like the Pride have a bad goal difference. Their +3 is third in the league after all. But we aren’t limited to just goal difference. Over the past decade, advanced statistics have made a huge splash in soccer commentary. These tools help those watching attach more quantitative values to what they see on the pitch. Of course there are limits to what the stats can tell you, particularly in a game like soccer with so many fluid and interactive elements. So the stats should never simply be taken as gospel. But they ARE informative.
And in the case of the Pride, the stats are pretty negative.
Orlando's Expected Goals numbers are actually pretty bad
Specifically, you can look at Expected Goals (xG), which takes all the shots from both teams in a game and assigns a value based on how often that shot is expected to produce a goal on average. Anyone who has ever watched a team dominate the game but fail to convert is familiar with the idea of a gap between xG and actual goals. Alternatively, you’ll often see teams produce almost nothing but manage to score from their only vaguely promising attack.
It’s not right to attribute those results entirely to ‘luck.’ A team can certainly beat it’s xG with clinical finishing or great goalkeeping. In fact, the best teams generally DO beat their xG because their players in the key positions are better than average.
Even so, over time the luck does tend to balance out. And it’s not hard to understand why. Goals are very rare in soccer. They take a combination of skill and chance. And from game to game, those odds won’t always stack together. Because goals are rare, you end up with a very small sample size and plenty of unexpected results. But xG measures a lot more data points, and therefore suffers much less from the problem of statistical noise that you get from small samples of rare events.
And what xG says matches what my own eyes tell me: the Pride are still giving away a lot of good chances, while not creating many good chances of their own. They’re scoring at a prodigious rate when the chances appear, and doing a lot of last-ditch defending on the other end. That’s great. But it’s generally not very sustainable.
To quantify that point, they have played eight games in 2021 (four regular season matches and four in the Challenge Cup). They have achieved a higher xG (expected goals) than xGA (expected goals allowed) one time over those eight games. These numbers come from Statsbomb via fbref and the wonderful Arielle Dror (who you should definitely be following on Twitter).
Overall, the Pride have accumulated 9.24 expected goals, while conceding 11.71. That’s not awful, but it’s the profile of a team that’s hanging around the edges of the playoffs, not one pacing the field.
The case for optimism
Both my gut and my head tell me that the Pride will come falling back to earth pretty soon. But I don’t want to overstate the case. There are some reasons to think this is more sustainable than a blunt application of the stats would suggest.
First, the trend is good. The Pride just had (by far) their best game of the season, amassing an impressive 2.46 xG against Kansas City against only .66 xGA. And they recently held the very good North Carolina Courage basically even, an impressive feat. So there’s an argument based on trajectory.
That said, Kansas City are, well, pretty awful. Now, every game matters and you can only play who is in front of you. But I’d like to see the Pride outplay a decent team before drawing many conclusions.
Second, there’s a pretty clear profile of the kind of team that might outperform its xG. Great teams usually do, because xG assumes an average shooter and keeper, but great teams generally have great players in those positions. And even if Orlando aren’t a great team, they definitely have great strikers.
In simple terms: it’s not shocking that the team with Alex Morgan is putting away more than its fair share of chances. And on the other side, Ashlyn Harris has plenty of naysayers, but even the sharpest critic will acknowledge that her biggest strength is in shot-stopping. What she gives away in the buildup, she can often make up in saving the final shot. That’s precisely the sort of thing that might generate a disconnect between xG and actual goals.
It’s not the sort of thing you can really count on, especially when some of your key players to run this gap will be leaving for an extended portion of the summer to play in the Olympics. But it’s certainly not nothing.
Third, there's a plausible confidence narrative here. In this account, Orlando was significantly underperforming their talent. A general malaise had settled, and the new coach wasn’t able to get things to budge. But some admittedly lucky performances can go a long way toward rebuilding morale. So maybe the regression isn’t so much the results catching up with the performances but the other way around.
That certainly feels similar to the last time the Pride were good, when Morgan and Marta were firing on all cylinders and suddenly the whole supporting cast also clicked into gear. Maybe it will happen again.
Predicting the Pride
So those are the positives. I think they’re reasonable. And it's also true that (whether by luck or not) Orlando have already built a pretty good buffer for themselves. Basically, they don’t actually need to be great going forward. If the goal is the playoffs, they’ve already earned themselves a six point buffer over most of the other teams who will be fighting for those spots.
Given that, I think they've still got a great shot at the playoffs, even if serious mean regression sets in. But for the sake of clarity, I'll stop hedging my bets and offer a concrete prediction here.
My bet: after picking up 10 points from 4 games, they’ll only manage a point per game for the rest of the season. That would put them on 30 points, which would be an exceptional result compared to their 2019 campaign (just 16 points), but would definitely feel like a disappointment after the hot start.
One final note: despite everything here, by no means am I rooting against Orlando. This run has been a lot of fun to watch, and I would be absolutely thrilled to be proven wrong.