clock menu more-arrow no yes

Filed under:

Quality, quantity of Sounders' chances started improving before shakeup, still getting better

New, 28 comments

Using expected goals both for and against, we look at the Seattle Sounders' 2016 season to determine whether they've been bad or just unlucky, and if that trend has shifted.

Kim Klement-USA TODAY Sports

Through the first 23 games of the up and down 2016 MLS Regular Season, the Seattle Sounders have been victim to bad luck, poor finishing, and general apparent awfulness. For the first little bit of the season, we fans hung our hats on bad luck, claiming regression would see the Sounders bounce back to the middle of the pack. The promise of poor finishing turning wasted chances into future goals never really became more than momentary flashes of competency, all of which resulted in the midseason firing of coach Sigi Schmid and the promotion of long-time assistant Brian Schmetzer. This move also corresponded with the arrival of both Alvaro Fernandez and Nicolas Lodeiro.

But how much of the first part of the season was due to bad luck? How much of the sustained awfulness that saw the Sounders claim seven points from Sigi's final 11 games was due to poor finishing? Thankfully the beautiful minds at American Soccer Analysis (@analysisevolved on Twitter) put together a comprehensive, game-by-game list collating every team's Expected Goals for each individual game. From this list I pulled all the data regarding the Sounders, looking at Expected Goals Allowed (xGA) versus Expected Goals For (xGF) and figuring out what result should have actually happened.

From this, we're left with two numbers: what a team should have allowed to score against them and what a team should have scored themselves. For the purpose of this exercise, I did not do any rounding, so the results are based solely on total aggregate expected goals. It is also important to note that these numbers include Penalties, both for and against, but we'll talk about that later.

Seattle Sounders 2016 Season

Opponent xGA GA xGF GF Result xResult
SKC 1.08 1 0.55 0 0 0
RSL 1.14 2 0.98 1 0 0
VAN 2.08 2 1.35 1 0 0
MON 0.77 0 0.79 1 3 1
HOU 1.06 1 1.56 1 1 1
PHI 1.29 1 2.27 2 3 3
COL 2.82 3 0.60 1 0 0
CLB 1.22 0 2.43 1 3 3
SJE 1.67 0 1.30 2 3 1
DAL 1.57 2 0.78 0 0 0
COL 0.75 1 1.11 0 0 3
NER 1.81 2 1.12 1 0 1
DCU 1.62 0 1.10 2 3 1
NYRB 1.06 2 1.09 0 0 1
NYCFC 1.44 2 1.57 0 0 1
TOR 1.53 1 1.50 1 1 1
LAG 1.28 1 2.41 0 0 3
DAL 0.20 0 2.45 5 3 3
POR 1.47 3 1.19 1 0 1
SKC 1.83 3 0.04 0 0 0
LAG 1.41 1 3.17 1 1 3
OCSC 1.60 1 2.37 3 3 3
RSL 0.85 1 3.17 2 3 3
Totals 31.55 30 34.90 26 27 33

Based on what we've seen here, we can see with this rudimentary not-rounding analysis of these numbers, the Sounders should have around 33 points right now, compared to their 27. There are several games that pop out, games where the Sounders won where they probably should have lost or drawn (D.C. United, San Jose Earthquakes, and Montreal Impact); and games where the Sounders lost or drew but should have won instead (LA Galaxy twice and Colorado).

However, not rounding is a somewhat sloppy way to sift through and understand what this data is telling us. Take for example the first game against Real Salt Lake, where the Claret and Blue side had an xGF of 1.14 while the Sounders had an xGF of 0.98. By all accounts, this should be a 1-1 draw instead of the 2-1 loss or 1-0 not-rounding-analysis. To compensate for this, I feel a better way of understanding the what-should-have-happened process is to take the delta between the xGA and xGF, with any difference half a goal or greater denoting victory one way or the other. This next table takes that into account.

Seattle Sounders 2016 Δ Results

Opponent ΔxGF - xGA ΔResults
SKC -0.53 0
RSL -0.16 1
VAN -0.73 0
MON 0.02 1
HOU 0.50 3
PHI 0.98 3
COL -2.22 0
CLB 1.21 3
SJE -0.37 1
DAL -0.79 0
COL 0.36 1
NER -0.69 0
DCU -0.52 0
NYRB 0.03 1
NYCFC 0.13 1
TOR -0.03 1
LAG 1.13 3
DAL 2.25 3
POR -0.28 1
SKC -1.79 0
LAG 1.76 3
OCSC 0.77 3
RSL 2.32 3
Totals 3.35 32

There are some differences that jump out at us initially, instead of winning against Montreal and San Jose, we draw; instead of drawing against Houston, we win. In fact, most glaringly, 7 of our loses now become draws. It's not that the Sounders have been winning games that they should be losing or drawing, it's that they're losing games in which they should have earned a point. Based on this model, the Sounders should have a record of 8 wins, 8 draws, and 7 losses, good for 32 points.

Another system for estimating what should have happened based on goals is using a Pythagorean formula based on goals allowed and scored. The simple formula described here, would estimate the Sounders to have 31.5 points based on just their actual current goal differential. If you instead used the xGD, it jumps all the way up to 36.6 points.

But the issue of penalties still remain. Teams' xGF are being rewarded for taking a penalty kick, which is not entirely indicative of one team's offense versus another team's defense. It's been well established that MLS has some problems with consistency regarding their referees, with some teams and certain players also having a propensity for diving at first contact or attempting to elicit poor calls that can change games. So in order to fully understand the Sounders true xGA and xGF independent from penalties, we reached out to the fine folks at American Soccer Analysis to see if they could give us a blanket number that penalties account for in the total xG for a game.

Apparently it's not as easy as that, as each penalty is wildly different and based on the angle of the shot, the goalkeeper protecting the goal, and other minute factors that boggle my mind, the number is roughly 0.75 xG per penalty kick. So far this season, there have been 8 penalties taken in games involving the Sounders, with 5 of those being converted (one post, one crossbar, and a Stefan Frei save account for the three failed conversions). In order to truly understand luck vs unluck, I created a third table that used xG independent from Penalties.

Seattle Sounders 2016 xG without Penalties because they're fluky

Opponent xGA - Penalties xGF - Penalties Penalty Independent ΔxGF - xGA Penalty Independent ΔResults
SKC 1.08 0.55 -0.53 0
RSL 1.14 0.98 -0.16 1
VAN 0.58 1.35 0.77 3
MON 0.77 0.79 0.02 1
HOU 1.06 1.56 0.50 3
PHI 1.29 2.27 0.98 3
COL 2.07 0.60 -1.47 0
CLB 1.22 2.43 1.21 3
SJE 0.92 1.30 0.38 1
DAL 0.82 0.78 -0.04 1
COL 0.75 1.11 0.36 1
NER 1.06 1.12 0.06 1
DCU 1.62 1.10 -0.52 0
NYRB 1.06 1.09 0.03 1
NYCFC 1.44 1.57 0.13 1
TOR 1.53 1.50 -0.03 1
LAG 1.28 2.41 1.13 3
DAL 0.20 1.70 1.50 3
POR 1.47 1.19 -0.28 1
SKC 1.83 0.04 -1.79 0
LAG 1.41 3.17 1.76 3
OCSC 1.60 2.37 0.77 3
RSL 0.85 2.42 1.57 3
Totals 27.05 33.40 6.35 37

The biggest difference here lie within three games. Our first FC Dallas game is now a draw if they don't have their penalty. Our game against New England Revolution is now a draw if Lee Nguyen doesn't score from the spot. But most importantly, that game against Vancouver Whitecaps, where Mark Geiger put it upon himself to display his strong whistle-and-point game, that somehow turns into a victory for Seattle, much to the shock of us all. The other penalties against Seattle, both Colorado Rapids and San Jose Earthquakes, do not affect the final ΔxG enough to change the expected result, this also being true for both of Seattle's penalties, which were given in what would be victories with or without the penalty kicks.

Again, we can also use the Pythagorean method to analyze this. Using the Sounders xGD - penalties, the formula expects them to be at 38.8 points instead of their current 27 points.

So with all this being said, it's easy to sit here and say that we've been unlucky, "The Sounders should have 37 or 39 points, by golly!" But that's not entirely accurate or fair. Sure, the Sounders have been victims of their own bad luck, but they also have scored nearly 9 fewer goals than their xGF would indicate. It's important to look at trends too. Through the first 16 games of the season, the Sounders only had a ΔxG greater than 1 goal in one game, against Columbus. Every other game was close or a blowout by Colorado. In the seven games since, the Sounders had a ΔxG greater than 1 in four games, two of those being under the new coach Schmetzer. And the one game under Schmetzer that didn't see such a great difference was Seattle's first come from behind win in over a year.

The Seattle Sounders are trending up, both the points and xG are showing it. No matter how we separate the data, it's showing that Seattle should have earned 16 points in their past 7. With the newly instilled confidence under coach Schmetzer and the aggressive attacking since his appointment, the playoffs are in reach. The Sounders were never a bad team, merely an average team on the wrong side of fortune. But now they're a good team creating plenty of chances and scoring goals. As the players continue to grow together and trust in the abilities of their teammates, who knows what these next few months have in store.

* thanks go to Andrew Beck for the assist with the Pythagorean Method. And by assist I mean he wrote that part and I took his name off the byline.