The data from 2010 Brazil Serie A provides data on passing that has not been previously available. The first rule of good analysis is that when you get new data, the first thing to do is to look at the data. So I started by generating a historgram of distance of all 118,445 passes in the data set. The result (below) is much as expected with the mean distance of pass being 21.6 yards, with more passes below 20 yards than above 30 yards.

Now I started to slice up the data to get a good feel for what we can learn from it. From the data, we can see that 76% of passes are complete. (Note that we use a more restrictive definition of a complete pass: a completed pass may not include a deflection which changes the balls course).  So I split the data between complete and incomplete passes to see the difference in the distributions. This yields little in way of surprise as the incomplete passes tend to be longer (26.7 yards) than complete passes (20 yards), and incomplete passes have a wider distribution (stdev = 18.6 yards) than complete passes (stdev = 12.1 yards).

Now that we have established that the data confirms our general beliefs regarding what passing distributions should look like, we can start to see how some of the other variables in the data set effect pass completion, namely how does defense figure in to the passing game. Taking basic averages, we can see that passes that are Challenged (the highest level of pressure) are only complete 60% of the times, while 80% of passes that are not Challenged are completed.  Distance of the pass matters of course, so I ran a logit regression (an appropriate method for estimating probabilities) to estimate the probability that a pass would be completed based on the distance of the pass and whether the pass was challenged. The graph below shows the difference in completion percentages by distance for passes that were challenged and those that were not.

Once distance is controlled for the average difference in completion percentage for passes that are challenged grows to 27.6% vs those with no pressure. When the same analysis was run to determine the effect of Marked pressure (the second highest level of pressure) instead of Challenged pressure, the average difference in pass completion drops to only 6%.

This adds a significant piece to our knowledge of defense. Pressuring an opponent as strongly as possible has a 4.6 times larger effect on pass completion than just the next lowest pressure level. But does this really matter? Well, possessions that result in goals average nearly 3 passes while all other possessions average 2.2 passes. Challenging passes has a significant effect on reducing the number of passes on a possession, which will reduce the possibility of a goal being scored. 
This of course is just the beginning of the story. Over the coming months we will continue to push on this data so we can get an accurately gauge how skilled at passing a player is based on not only distance and pressure, but field location of the passer and target, type of pass etc. We will then combine these into a metric of passing ability, adjusted pass completion percentage, which will allow us to compare all players on an equal footing in terms of their passing ability.