Since MLS made available detailed match statistics provided by OPTA early in the 2012 season, PSP has published a number of posts in which we try to better understand what was happening on the field with the Philadelphia Union through the use of various statistics. These posts have been something of an evolution, ranging from a closer examination of what happened in a particular match in order to better understand what went right or wrong, to trying to understand larger trends, be they for the team as a whole, across positions, or for individual players. The usage of such statistics was at first largely descriptive but as more data became available over time, and also thanks to the contributions of Rolando De Aguiar, the posts also became more analytical.

One post we did early in the offseason looking at the passing numbers of each player led to much discussion about what the numbers revealed. It also resulted in another post from Rolando that used some impressive visualization software to better understand the numbers collected.

We have continued to try to better understand what is happening in each game by looking at individual match statistics. And even though we are only four games into the 2012 season, we turn again to the question of passing statistics. Hopefully, as the season goes on, the Union’s passing numbers will provide some insight into the team’s success.

First, below are the team passing stats. Over four games, the Union has a overall passing accuracy of 72 percent (71.98), 72.8 percent at home and 71.0 percent away. In 2011, the overall average was 72.6 percent, 73.1 percent at home and 72.1 percent on the road.

Turning to individual passing numbers, recall that I said above “even though we are only four games into the 2012 season.” This is because, due to the fact that five players have yet to play 90 minutes, calculating passes per 90 minutes enters into the realm of forecasting rather than adjusting for playing time.

For example, Roger Torres appeared twice in the first four games: he played 27 minutes against Colorado, recording 12 successful passes from 26 attempts for a passing accuracy percentage of 46 percent; he played 45 minutes against Chicago, recording 22 successful passes from 28 attempts for a passing accuracy of 77 percent. Yet when we calculate his passes per 90 minutes (total number of minutes ÷ 90 = x, total number of passes ÷ x = P per 90), we get 67.5 passes, more than he’s attempted in total, let alone in a single game. Nevertheless, it does make for some interesting speculation about what Torres might be able to do with more minutes.

In addition to overall passing numbers, I also collected the numbers for passes in the final third.

Again, some strange numbers turn up for those with less than 90 minutes played over four games.

With only four games worth of information to go on, it is difficult to reach any conclusions from the charts above. But here are a few observations for players that do have more than 90 minutes:

• Daniel’s overall accuracy of 76 percent combined with his 71.1 percent accuracy in the final third argue well for him to continue to see significant minutes.
• The players who have played at outside back all have significantly more P per 90 than the central defenders. Unsurprisingly, for Williams, Garfan and Lopez, this is particularly pronounced in the final third.
• Mwanga and Pajoy are very similar in terms of P per 90 both overall and in the final third, but Mwanga appears to be significantly more accurate.
• He only has 74 minutes but Freddy Adu’s numbers suggest the Union can use him on the field but quick.

Tomorrow we will return again to the Union’s passing numbers with some visualizations to better understand some of the data above.