How DC United continues to beat the odds / by Coleman Larned

By Coleman Larned (@thesoccerswell)

MLS will not be a top league until its coaches find better ways to connect projection metrics to tactical changes. Ben Olsen is showing he may be on the right track. The more data that is collected on variables such as game variability, individual player contributions and spatial frequency, the more information there is to digest and to affect playing, teaching and coaching. Therefore the issue lies in the translation of such data into meaningful, impactful resolutions that can be measured by on-field progress.

There's an opportunity for both sides of the analytical coin to learn and communicate more effectively. Can large data aggregation along with sound, statistics application give insight into the complexities of soccer from a perspective lost from the naked eye? Can tangible trials and consistent experiences within the sport mold the direction in which data is aggregated?

In this piece we'll take a look at the recent seasons of DC United and how they've usually outperformed our xG projections in order to better understand this disconnect. In recent MLS memory DC United has surprised, frequently outpacing projection metrics and 'punching above their weight'. Let's talk look at what Olsen has done tactically and managerially from 2012 to the current 2015 season to remain stubbornly successful.

2012

In Olsen's 2nd full season in charge the personnel and tactical changes started to create drastic improvements in results. Coming off of a 2011 season that saw DC United finished 7th in the Eastern Conference and 13th in the overall MLS standings, the team rose to 2nd in the Eastern Conference and 3rd overall, ending their season by losing in the Eastern Conference finals.

Chris Pontius prominently featured to bust the 1.39 xGF by amassing his best career goal return (12) and 2nd best assist return (4). What the xG models do not do well is take into account, among variables, where defenders are positioned and the quality of the shooter. As a midfielder, these returns, especially in goals scored, are fantastically unexpected. But Pontius is a very good striker of the ball and tends to shed defenders easily, potentially explaining what the metric misses.

DCU v SKC 11/22/2012: Kitchen shielding back 4 and providing main distribution channel

DCU v SKC 11/22/2012: Kitchen shielding back 4 and providing main distribution channel

The 2012 season also began the beginning of the successful maturation of GK Bill Hamid and CDM Perry Kitchen. Hamid posted his best positional statistics of his career (SO: 8, SV: 88 GA/G: 1) and Kitchen took and kept a deep lying #6 role for DC United, and hasn't given it up since. Kitchen's trajectory from coveted draft pick, to being the first player Ben Olsen picks on the team sheet is emblematic of DC United's spatially aware, technically sound and bulldog tough playing style.

The scoring load was distributed among five players who scored seven goals or more. Although this might seem advantageous for a well balanced attack, having a midfielder (Pontius) contribute most of the goals and not identifying a true striker to be able to replicate top goal scoring form causes longevity  and consistency issues. This was seen in the following season.

2013

Coming off of a fairly successful season, DC United and Olsen made it clear that their intentions were to press on and qualify for a CONCACAF Champions League position. Although intentions were established, the club's offseason activity seemed to contradict instead of aid such goals. DC United shipped off arguably their most promising young talent and 3rd highest assist tally from the previous season, in Andy Najar to RSC Anderlecht, and declined their contract option on Maicon Santos, who accumulated the 3rd most goals for the side in 2012.

The season ended up being an utter disaster, with accumulating .29 less GF than their xGF and allowing .23 more GA than their xGA, but their side was young and the club was in a transition of identities. The departure of long time CEO and president of the club, Kevin Payne, opened the door for Ben Olsen to assume greater ownership in the club's market policy. During the 2013 campaign Olsen fielded the youngest starting XI in club history on August 10, 2013 with an average age of 23.21 years, with 4 players being 20 years old or younger.

Both extremities of the team shape were to blame for the 3-7-24 season. The club was tied for the worst GD in the league (-1.01), had the lowest GF total (22) and didn't have a player accumulate more than 3 goals throughout the campaign. Changes were needed and they came in droves before the 2014 season.

2014

The 2014 season was a transformative year for the modern iteration of DC United. Faith was shown in Ben Olsen to make changes and implement them appropriately, as he remained head coach. Three of the newly acquired players ended the season in the top 16 players with the greatest G - xG differential, using our xG metric (3. Silva: 5.52; 6. Espindola: 5.27; 16. Rolfe: 3.02).

DCU transfers out/in for 2014 season - click to expand.

One clear variable stands out as to how Fabian Espindola was an outlier: 5 of his 11 goals came from balls struck outside of the 18-yard box. This speaks to the inability of the metric to properly account for a player's ability and propensity to shoot from distance, as some players are better than others.

DCU v POR 5/3/2014: Kitchen tasked with most of distribution, but now in advanced spaces

DCU v POR 5/3/2014: Kitchen tasked with most of distribution, but now in advanced spaces

As defensive changes were made, Olsen started to find a back four that he trusted, which led to the stingiest defensive record in the league (tied with LA Galaxy for least GA: 37). Bill Hamid chipped in and usurped his previous season best in 2012, with career highs in shutouts (10) and saves (110) and led all Keepers in xGmG, earning him an MLS Best XI spot, along with fellow teammate Bobby Boswell. Between Hamid's performance and the consistency between defensive formation and personnel, DC United were able to put men behind the ball consistently and in an organized fashion. This could have accounted for their exceeding the xGA metric.

DC United not only defends compactly, with a rigid shape, but they defend a lot. Once again, Perry Kitchen is an integral cog in almost every phase of the side’s game: shielding the back four, being the main distributive outlet to transition into the middle third and now providing box-to-box energy to affect possession in more advanced spaces.

2015

It’s been all about the defensive solidity for DC United in 2015, scoring an impressive 1.00GA against a 1.35xGA. This is the number that’s keeping them afloat atop the Easter Conference in MLS. Although the team in conceding the second most shots per game (14.5) in the league, the defensive effectiveness can be measured in another weakness of the xG metric: defensive awareness and positioning. The side ranks 2nd in blocked shots (3.6) and 3rd in crosses blocked (2.8), along with spending the least amount of time in the opposition’s half in MLS.

DC United not only defends compactly, with a rigid shape, but they defend a lot. Once again, Perry Kitchen is an integral cog in almost every phase of the side’s game: shielding the back four, being the main distributive outlet to transition into the middle third and now providing box-to-box energy to affect possession in more advanced spaces. As Perry Kitchen matures and his role within the team evolves to be more influential in a variety of spaces on the field, the harder DC United is to beat.

DC United's progression in MLS over the past 3.5 seasons has embodied different weaknesses in the xG metric in different seasons.The underlying causes behind this are still grounded in the implementation of sound soccer tactics. As Ben Olsen has grown with this team he has created consistency in crucial areas: a strong, vertical spine through the team, defensive compactness, proper spacing and an unwavering commitment to on-field work ethic. It's hard to see it falter anytime soon, so expect more metric busting.