FanPost

Selected DC United Numbers vs Portland (Game 8)

Let us examine the DC United game using soccer statistics. I'll be using selective statistics that I find potentially very interesting largely taken from the chapters of The Number Game: why everything you know about soccer is wrong by Chris Anderson & David Sally and Soccernomics by Simon Kuper & Stefan Szymanski.


I know not all statistics are useful for soccer. But empirical evidence shows selected statistics are predictive of results, goals, and player value. I will entertain the notion of doing more complicated analyses or examining additional statistics provided the data are available. I am using data from WhoSocred.com and MLSSoccer.com.


Hopefully these numbers provide useful game insight and predictive utility. Explanations for why I selected these numbers can be found in my first post of this series.


The Weakest Link


season avg. WhoSocred

vs POR 5-3

vs DAL 4-26

Andrew Dykstra

6.90

DNP

DNP

Bill Hamid

6.72

5.7

7.1





Bobby Boswell

7.20

6.1

8

Cristian Fernandez

7.09

7.4

7.6

Sean Franklin

6.93

DNP

7.4

Chris Korb

6.10

6.2

6

Jeff Parke

7.09

6.6

7.1





Davy Arnaud

6.81

7.2

7.1

Alex Caskey

6.17

6.2

DNP

Nick DeLeon

6.87

6.4

6.7

Jared Jeffrey

6.00

DNP

DNP

Perry Kitchen

6.76

7.1

6.5

Lewis Neal

6.10

6

DNP

Kyle Porter

5.99

DNP

DNP

Chris Rolfe

6.88

6.1

7.7

Luis Silva

6.52

6.4

6.3





Conor Doyle

6.11

6.5

6.5

Fabian Espindola

7.32

6.5

10

Eddie Johnson

6.75

DNP

6.9

*Chris Rolfe's first three ratings (including the DNP) are from his performances with Chicago.

*An explanation for WhoScored’s ratings can be found here

Inferences

Hamid goes down as the lowest rated player. He had a 33% pass accuracy. Ok, distributions is not the biggest part of a goal keepers responsibility but it’s fairly important. Rickets had 50%. Julio Cesar had a 95% this week (seems extraordinary) he looks to average about 65%. Hamid has to do better with his distribution. But maybe Hamid did so poorly because Johnson was not there to win headers. Last week Hamid had a 53% passing accuracy. Which seems to be about Hamid’s season average.

Lowest rated field players were Neal, Boswell and Rolfe (6.0 to 6.1). Why them? I’m not sure. The black box that makes up the player ratings is a downside to using player ratings. If we consult the player rating description "There are over 200 raw statistics included in the calculation of a player’s/team’s rating, weighted according to their influence within the game. We take into account every event that was of importance, its area of the pitch and its outcome; whether it is positive or negative….. The scoring system for the ratings is out of 10, starting from 6.0 and 10 being the highest score. " Well, it appears Neal, Boswell and Rolfe had little to no positive impact on the game. That I can believe. Once again the substitutes had no real impact. Silva the highest rated of the three was a 6.4.

From the last game I wrote "DeLeon and Doyle really need to step it up". De Leon did worse this week. Doyle did the same. Both are consistently poorly rated. Doyle is usually a sub so that hurts but not as much as a locked in starter not producing. This is MLS and no teams have a perfect 11 but given DeLeon’s technical abilities you would expect more from him.

KEEP THE BALL!

Possession: 47% (black part of the pie, green being Portland 53%)

Total passes: 408 vs 452

LSR = .18

Turnovers: 53% (Season Avg. = 50% SD = 4%)

Through Balls = 0

DCU Final Third passing: 74 completions (red bar) of 116 attempts (black bar)

POR Final Third passing: 93 completions (light green) of 137 attempts (dark green)

Inferences

Once again conceding possession. Not surprising, especially away to Portland. Out passed. Again, not surprising. 53% of the turnovers. This is within DCU’s standard deviation (the average amount of variability for the season, including this game) but on the high end and not good. This is also the third consecutive game DCU has the worse percentage of turnovers. Giving the ball away, especially more than the other team is bad.

DCU had 0 through balls while Portland had 2. Also reflected in the final third passing numbers Portland had many more attempts and many more completions (64% vs 68% completion). DCU final third numbers are right at their average of 119 attempts and 77 completions. And Portland is right at DCU’s opponents average for the season of 142 attempt and 98 completions. DC United is rarely the team in the with more passes in the attacking third.

Click here for the first match numbers.

*I would also point out that the above statistics are correlated with league points earned but that is for an entire season (with European data too - may not apply in MLS) and not necessarily predictive of individual games.

SHOOT THE BALL!

DC United: 16 Shots (19 by Portland)

BIG CHANCES: 0

2 Goals

TSR = .46

Inferences

Portland had more shots. Ugh. And more goals. Double Ugh? Triple Ugh. DCU actually had more shots than their season average of 12 (but within their SD of 4.6). Sadly, Portland had more than the average 13 (SD = 5) shots given up by DCU. In fact Portland’s 19 shots ties NJ’s as the high.

Summary

Out played. Out possessed. Out passed. Out shot. Out scored. Out played. What’s interesting is DC United put up numbers entirely consistent with their season. This was a completely average game (by these selected numbers) for DC United. Now, DC did give up more shots than usual but they also took more. MLS parity? DC United’s fall to middle (below middle) of the pack begins? A pessimistic projection. But when a team gives up goals in the first and last minutes of a game it’s upsetting. Maybe they stop doing that? Maybe goals is the outlier statistic? Maybe the team will get a stadium? Oh.

Trending Discussions