Editor's note: Bumped to front page for general awesomeness. Though I've disagreed with the author on this subject our conclusions are the same. - Dave
On several recent occasions in SounderAtHeart comment threads I have found myself buried in discussions over the relative passing abilities of Brad Evans and Erik Friberg. Although one or the other player will commonly receive the favorable comparison on these pages, the debate remains largely subjective with only the most general passing statistics brought to bear. When Dave last assembled pass completion percentages, Evans held a 4.4% advantage over Friberg. If one were to simply take this measurement at face value and favor the former then one would similarly disapprove of the passing acumen of Mauro Rosales, Fredy Montero and Alvaro Fernandez, each having completion rates as weak or weaker than the Swede. An unsurprising conclusion: pass completion varies with player role, inviting the question whether Friberg and Evans truly share a role in the Sounders' 11...
Subdividing the passing statistics by pass type and directionality is our readiest means of assessing role apart from the position of the players on the field. A cross is more difficult to complete than average, while a lay-off is easier. A pass forward encounters more potential defenders and is far more prone to interception than a pass back. The chalkboard data at MLSSoccer.com provides us with a rough measure of pass direction. The chart below provides a summary of the last 4 regular season games in which Evans or Friberg served alone exclusively as the CM pairing with Osvaldo Alonso. Direction was split into 45 degree octants, merging the forwardmost 3 and the backwardmost 3 (e.g. a "forward" pass covers the area 66.5 degrees "east" and "west" of "north" where N designates a pass directly upfield).
These four games cover roughly 142 passes for Friberg and 151 for Evans. The analysis specifically excludes head passes and throw-ins in an effort to characterize distribution tendencies strictly with the feet. Putting a value to the graphical representation, Erik Friberg passes the ball forward more than 16% more often than Brad Evans. The difference exceeds the game-to-game variability of either player (1 standard deviation of forward pass percentage is about 6 for Evans and 12 for Friberg). A back-of-the-envelope calculation would suggest, in a 40 pass game, Friberg would miss a couple extra passes and successfully advance the offense forward with his feet about 5 times more.
So who is the superior passer? Leaving aside the problem that we need additional analysis to make the above statistics more robust (a more comprehensive data set for both Evans and Friberg, a discrete analysis of pass completion rate vs. direction, and some measure of the difference in positional role over the course of a game would all be helpful) we clearly need some measure of value for a completed forward pass (as well as the negative value of a failed pass) which largely depends upon the team in question. Adding in headers makes up a significant amount of the difference in distribution between the players (one role that Evans clearly plays is to contend long aerial midfield clearances by the opposing defense and send the ball back into the offensive third... or the offensive flick-on header), but this is not precisely a measure of passing skill when on the ball, Evans' forward headers have a tendency to be unsuccessful passes, and Evans' absence from the field simply means that other players would contend the same ball if possible (Fernandez, Neagle, Alonso).
Therein lies the rub. Distribution clearly shows that different styles of play are expected from the two CMs. One or the other would likely provide more to the overall play of the team dependent upon other personnel on the field (for both teams) and a level of "form" that exhibits a high degree of variability from game-to-game. The difference in passing success between the two players falls well within the variation one would tentatively expect from the difference in role. Referring to Evans' as the superior passer is clearly not strongly supported by the statistics we have available. We need further study to truly compare them.
If, instead, one would like to make a strictly subjective argument in favor of one or the other player with respect to passing skill, have at them.
Finally, the difference in role somewhat complicates the question of which player (if either) should be protected in the coming expansion draft. Evans and Friberg are not truly redundant with one another. However, both have styles of play which contribute to the team overall. It is difficult to state which play style is more important at the present time (harder still to project it for next season's roster). My commonly stated opinion is that Alvaro Fernandez (or Mauro Rosales) should shift over to the center midfield upon Steve Zakuani's return to allow our top 4 midfielders on the field at the same time. The difference in value between our (prospective) 2nd CMs is not great enough to justify a protected spot. I stand by that opinion.