2016 NWSL Season – A Statistical Review of all teams- The Good, Bad, and Unlucky– and Who is Lucky & Normal?

facebooktwitterreddit

NWSL Season Review- The good, bad, and unlucky. A statistical look at all of the teams. PDO was looked at to see which teams were lucky or unlucky.

Let us statistically determine what happened to your favorite NWSL team this year. Did they win the championship, come in last place, or somewhere in the middle? Did they deserve their overall place in the standings, or was “luck” a cause? I looked at the NWSL official statistics and came up with a few calculations.  I will review the teams in alphabetical order:

Boston Breakers- Unlucky?

Boston finished at the bottom of the table with 11 points. To put into perspective, the league average was 27.9 points, and Western New York Flash had 32 points and was the lowest ranked team in the playoffs. To put in terms of games, Boston needed 5 more victories to be almost average in the league.

So, what did they do wrong? They did not score, only having 14 goals, but also gave up a league worst 47 goals. I went through the NWSL statistics and came up with some numbers. PDO is a calculation that Brian King came up with years ago. It is shooting percent plus goalie save percent, and the number can be normalized to 1, 100, or 1000 depending on sport and personal preference. I will set to 1.000. Boston had a score of 0.841. A low number indicates “bad luck”, a score near 1.000 is normal, and a high score, such as 1.200 indicates “lucky”. Boston is clearly “unlucky”.

Why would Boston be “unlucky”? Is it moon phase? Conspiring gods? Bad pennies in their wallets? In soccer, it is a difficult question to answer. One answer is that they faced league champions Western New York Flash 4 times. Maybe that was their cause of their bad luck? Can you blame bad refereeing on “bad luck”? In one game against Orlando there was a controversial penalty kick awarded to Orlando. Was there other similar factors that did not go their way in other games?

We can also look at Total Shots Ratio (TSR). TSR is basically overall shots per game compared to both sides. Basically, “neutral” is 0.500, meaning that you and your opponents shoot equally. Therefore, the higher the number, the more you are shooting compared to your opponents. Boston had a horrible score of 0.354. This means that opponents clearly outshot them on goal. If all other things were equal , they could not overcome the barrage of goal-shots by their opponents.

Chicago Red Stars – Good?

Chicago went to the playoffs with 33 points. I checked their PDO score and they have a 0.961, which is close to 1.000. So, if anything they are slightly unlucky. My limited logic skills tell me that if you are in playoffs despite some bad luck, then you are good. A look at their fairly high TSR of 0.588 does indicate that they were good as they were outshooting their opponents.

FC Kansas City – Normal?

Kansas City had 26 points and missed the playoffs. They also have a PDO of 1.006. Therefore,  they were at league average for points, near 1.000 for PDO. This one might go to goalkeeper Nicole Barnhart. The league average for save percent is about 74%, while “Barney” had a league best at nearly 81%. She was the only goalkeeper to play all of her games. Therefore, she gets full credit for the team’s goalkeeping statistics. The other scores such as TSR, scoring conversion, etc. indicate that they were league average.

Houston Dash- Normal?

In my season review of the Houston Dash. I gave an overview of PDO, but also calculated it for both halves of the season. They ended up the season with 22 points, and with a PDO of about 0.976. Therefore, they were “normal”, but a little unlucky.  But in the first half of the season they had a poor shooting percentage, and at the midway point of the season Kealia Ohai began shooting the ball very well, thus bringing up their PDO in the process.

Orlando Pride- Unlucky?

More from NWSL

Orlando had a “mixed” year. They started off quite well with playoff aspirations, but something went wrong along the way and ended up with 19 points. Their PDO was 0.939. This implies some bad luck. One thing that happened was a season ending injury to Stephanie Catley around  August 22. They lost a good fullback, and the team was unable to replace her adequately. Was this a source of their bad luck?

Could another possibility be fitness? I have not studied this statistically, but one observation I have from their games is that Orlando is a “first half team”. They would score more goals in the first half. But in the second half, I have noticed a lot of “ball watching”. Watch the game below form around the 53:48 minute mark. You will notice that all Washington field players touch the ball once except the goalie. Or is this just poor overall Orlando defense?

Also, can playing in the Orlando heat and humidity all summer be a factor? Especially with Houston as their “rival”? That is 12 games that could be in southern U.S. humid conditions. The downside to that argument is that both teams are playing in the same conditions. But, did the humidity affect Orlando more? And can this be attributed as “luck”?

But a look at the TSR of 0.451 and their scoring conversion rate of 9.8% shows that they were below league average in both categories. What is scoring conversion rate? This statistic might be better analytically as it is total number of shots compared to goals scored. A “shot on goal” or “shot on target” implies the goalie did some work to save the goal. But a “shot” can miss by 50 miles. What if a team has a low score here? Then, there are a lot of wasted opportunities. The league average was 11.1% . So, some scoring opportunities were wasted, but we are looking at 1 goal per 100 shots.

Portland Thorns FC-  Lucky?

Portland won the NWSL Shield award with a 41 points.They also have a PDO of 1.111, which implies luck is involved. In my team review, it is noticed that they have 9 National Team members from all over the world. To have that many great players on one team compared to others will probably show up statistically as “luck”. In real life, that would be called “great scouting / recruiting / coaching /general managing /etc.”  Portland seems to have created their own “luck” (thanks to the original allocathttps://www.youtube.com/watch?v=ebR4jGW7_pUion process in 2013).

Seattle Reign FC- Normal?

Seattle just missed the playoffs with 30 points.  They had a decent TSR of 0.557, and above average scoring conversion rate of 12.1%. They also have a PDO score of 1.014. This implies that they got what they deserved. Maybe they deserved to go to the playoffs, but maybe other teams were “luckier” than they?

Sky Blue FC- Unlucky?

Live Feed

NWSL offseason: Ranking the 10 best free-agents available
NWSL offseason: Ranking the 10 best free-agents available /

FanSided

  • The legacy Ali Krieger will leave behind Skyscraper Blues
  • Gotham FC's NWSL Championship is a Cinderella storySkyscraper Blues
  • NWSL Championship: Takeaways as Margaret Purce sparks NJ/NY Gotham FC to title FanSided
  • Megan Rapinoe's soccer career ends in heartbreaking fashionFanSided
  • Gotham FC has a new title: NWSL ChampionsSkyscraper Blues
  • Sky Blue finished near the league average with 26 points. They also had a PDO of 0.926. This implies that they were “unlucky”. Further investigation showed that they had a shooting percent of 24.2%, which is nearly identical to the league average of 24.0%. Therefore, scoring goals was not the issue. Looking at goalkeeping save percentage, we see that they had about 68.4% , where the league average is about 74.3%. Sky Blue had 2 rookie goalkeepers in Caroline Stanley and Caroline Casey. Their combined goal against average was 1.50, which is slightly worse than the league average of 1.32. Perhaps, a couple of goals were scored against them due to inexperience, and that was the difference?

    Washington Spirit- Lucky?

    Washington finished second with 39 points, both in the standings and losing the championship. Did they deserve to be there? They had a PDO of 1.077, which implies a little luck was involved. Looking at shooting percentage, we see a clue. They had a scoring percent of 27.5% compared to the league average of 24.0%. Also, their goalkeeping save percentage was also high at 80.2%, compared to about 74% for average. This implies “luck” on both sides of the ball. But does being more accurate in scoring, and having good goalkeeping  count as “good luck”?

    Can some of this “luck” be attributed to playing Sky Blue 4 times, who were a little “unlucky”? Did something bizarre happen in those 4 games to indicate a lucky bounce, poor refereeing, or weird moonphase  in those games? I did not re-watch these games, but  I do not recall anything controversial happening in any of those games, so maybe it is “luck”?

    A look at TSR shows that they average at 0.507.  But Washington had a scoring conversion  of about 12%, which is over the league average of around 11%.. Again, 1 goal scored per 100 shots does not look like an exciting statistic. The TSR is by far the lowest of all of the playoff teams, but the score is normal. meaning that they and their opponents made an equal number of shots on goals.

    Western New York Flash- Normal?

    WNY Flash won the league championship on penalty kicks over the Washington Spirit, after getting into the playoffs barely with 32 points. They had a PDO of 0.982, which is near the league average of 0.983, in which 1.000 is expected. Why would there be a difference between the league average and a perfect PDO? Conspiracy theorists of the NWSL would likely point to the sometimes questionable calls of the referees. Or can this be normal math error due to the small sample size? Perhaps a little of both.

    But at any rate, WNY Flash is considered “normal” despite playing a very “unlucky” rival team of Boston 4 times.This is odd despite the fact there was some statistic agreement between Washington and Sky Blue that one  team was “lucky” and the other was not. What the PDO score attempts to show is that a team scored a relative number of goals and conceded a certain number of goals that was “expected” of them.

    WNY dominated in TSR scoring a 0.622. They simply overwhelmed their opponents by shooting more. Therefore, by creating more scoring chances, then they are expected to win.

    Next: 2016 NWSL Season in Review- Seattle Reign FC- The Lost Year?

    To summarize, some teams may have been lucky, but probably created their own luck. Some teams may have been unlucky, but other statistics showed that they probably deserved the losses. And some did what was statistically expected to happen.