Data Shows Impact of Departing Spirit Players


The Washington Spirit cannot seem to escape the headlines in this off-season. It has been announcement after announcement of players leaving, being traded, suffering injuries, etc. The Spirit is one of eight original teams introduced in the inaugural season of the National Women’s Soccer League. Since its poor first season, the team has improved tremendously. Some of its improvement has come from changes in coaching staff and player personnel, but much more of it is due to a framework of players whose chemistry provides consistency in passing and defensive action. This past season, especially, was a display of consistency on the field for the Spirit. So how will the loss of so many framework players affect the Spirit’s performance? The best answer that I can give you is to show you what kind of impact those players had on the data collected from this past season.

Compare some headlines for the Spirit with respect to the average shots and goals differential per year. The black line indicates a differential of zero.

As you can see, the majority of players were either part of the Spirit at the beginning or joined during the Mark Parsons era. In 2015, the average differential (goals for minus goals against) was just beginning to break even and most of these players remained on the Spirit and they finally saw a positive differential.

The 2016 line up illustrates the starters in the beginning and how it transitioned during the Olympics and afterwards.

In fact, this can be considered as a simple assist network, in which one player assists another to score a goal and this dictates the direction of these lines. However, we can also highlight the number of players who have left the Spirit to truly illustrate the dismantling of the team.

The 2016 assist network is literally a history book from the Mark Parsons Era. Highlight the players (in red) leaving the Spirit and a majority of the offense falls apart.

This can also help give an idea of how the lineup changes affected this network. For instance, during the month of July, when the Olympians were gone, the Spirit had more unassisted goals (6) than assisted goals (4). When the Olympics were over,  the team did not return to the pre-Olympics lineup, and, for many, this was surprising. I know I was baffled during the Reign vs.Spirit game at Memorial Stadium, because I had assumed that the starters from the beginning of the season would be involved to secure the Shield as the season came to a close.

The 2016 season month by month. The number of goals scored with respect to the shots on goal gives a good indication of efficiency going into the playoffs.

Because the team has endured so many changes (trades, injuries, unknown status, etc.), it is worth trying to understand the impact for the 2017 NWSL Season with greater detail ahead of the college draft. Before I begin to illustrate the following data, please note it does not represent the full season. In fact, while I illustrated the shift in the lineup towards the second half of the season, much of this data reflects the team prior to those changes. For a complete picture, WoSo Stats requires more fans to cover at least 30-40 more games. I will present soccer field heat maps for the current data. On the left is the 2016 original team and the right is the team without all the players who have made official announcements. Speculative news (Canadians) does not count, but if it turns out to be true, I will tweet out new visuals.

For the last quarter of 2016, I spent a good amount of time exploring defense in the NWSL. I defined two terms based on WoSo Stats terminology – defensive action and pressure, in which action can be defined as conservative or aggressive. Washington Spirit defenders, at lease for the preliminary data, are the absolute best in terms of conservative action. Specifically, Krieger is the best at interceptions in the NWSL and both her and Oyster recorded some phenomenal stats. The original backline in general (Dydasco/Kleiner, Oyster, Zadorsky, Krieger) maintained a fantastic balance between one another, with all four being largely conservative and the outside backs being a bit more aggressive. It should also be noted that their defenders tended to have a huge rise in defensive action in the last 15 minutes. It should also be noted that their midfielders did not contribute as much defensive action compared to the rest of the NWSL, with exception to Huster (conservative) and Lohman (aggressive). Overall team performance can been seen in the interactive linked earlier. Where does the Spirit have to make up ground defensively? Hopefully, this visual illustrates it well enough – the new players will have to preserve this original defensive line.

Total Defensive Action includes ball shields, blocks, clearances, interceptions, dispossessions, tackles, and challenges. The hotter regions indicate more action. The Spirit will need to rebuild in big portions of the field.

When it comes to the Spirit, I will never forget the one goal in which every single player touched the ball before they scored. Remember that? That kind of team chemistry shines on the field. While the team may be scattered in their pass completion under pressure, they were solid overall. Banini is probably the most astounding—in her 650 minutes recorded thus far, she had the most passes under pressure per 90 on the Spirit and a completion of 92%.

Many players of the Washington Spirit perform well under pressure. The average for NWSL is illustrated to aid comparison.

Completed passes can be illustrated on the field as well. Here, the thickness of the line indicates the pass frequency. For passes within the same region, there is a circle and the size of the circle gives a similar indication. With Krieger, Nairn, Dunn, Oyster, and Banini already gone, along with 3 ACL injuries, it is clear this network takes a hit.

The thickness of the line indicates frequency of completed passes. Dots in certain locations indicates passes within the same area. The Spirit will have to rebuild this network completely.

The Spirit will miss through balls from Nairn, take-ons from Dunn, interceptions from Krieger, the conservative defensive action of Oyster (on par with Krieger), and offensive individuality from Banini. But is not fair to list just one trait each for these players, because they were good at many of these things. It is clear, while some of these may be individual traits, the team result is one rooted in history. Will the incoming players be able to replicate this chemistry? The Canadians have not announced anything as of yet, and while I did not mention the bronze medal Olympians much, they also played key roles on the 2016 Spirit team, and you do not want to see what the field looks like if I take them out. The one thing Spirit fans can take solace in: Lohman and Huster still remain, and they provided some of the best recoveries for the team. Other than that, all we can do for now is wait—wait for the draft, wait for more data; wait for the 2017 season to begin!

Special thanks to WoSo Stats team for fantastic data collection. If you are interested in helping, please join! Special thanks to certain data hungry fans who helped me collect some of the basic data to make the assist network. Also, never trust NWSL Play-by-Play records (lesson learned). Additional graphics will be provided on Positives and Negatives

Follow us on Twitter