Central Winger: Digging deeper on secondary pass – which players make life easier for teammates?

Back in June 2012, Central Winger introduced the concept of the "secondary pass," one of the many current methods of augmenting raw passing data. It's a part of the general ethos of this series to challenge raw numbers and rates and ask if we're accounting for an adequate amount of context to provide meaningful – if not yet actionable – results.

Our explorations into secondary passing is an exercise in just that.

As previously defined, a secondary pass is a pass that happened directly following a previous completed pass. Therefore, a player's "secondary pass completion rate" (or, 2RATE) is the percentage of passes that a player's teammate completes after he passed to them.

What 2RATE attempts to control for are incidents where a "completed" pass isn't actually beneficial to the continuing possession of a player's own team. For example, a "hospital ball" is a pass that immediately puts a player's own teammate under immense pressure – often leading to a crunching tackle and an immediate turnover. 2RATE takes aim at these types of incidents and crunches a less-biased metric than raw passing accuracy.

But, 2RATE is far from perfect, and also happens to succumb to similar biases that raw passing accuracy struggles with. Not all passes are equal; some are much harder to complete while others are much easier. Defenders and defensive midfielders are kings of raw primary and secondary passing rates because they tend to attempt more conservative passes in lower-pressure areas than their attacking counterparts. Therefore, we must attempt to adjust for pass difficulty and see which players are completing passes at rates different from what we would expect.

To do this, we use Opta's passing model that analyses each attempted pass and estimates just how likely that pass is to complete. To do this, the model weighs straightforward components such as pass distance and location as well as more specific qualifiers such as the current sequence of play. 

For example, here is pass difficulty broken down by area in MLS 2013. On a scale from blue to red, we get zones with low completion rates to high completion rates. The size of each bin is relative to the amount of passes attempted from those zones.

After adjusting for pass difficulty, these are the 10 best performers in adjusted 2RATE in MLS last season:

Player Ajusted 2RATE (percentage)
Brad Evans +2.56
Shane O'Neill +2.38
Darlington Nagbe +2.24
Nat Borchers +1.80
Robbie Keane +1.74
Davy Arnaud +1.60
Brian Carroll +1.49
Benny Feilhaber +1.34
Keon Daniel +1.22
Kosuke Kimura +1.20

By this measure, teammates who receive passes from Brad Evans are immediately 2.5 percent more likely to complete their next pass, the best in MLS last season. Going forward in this series, we hope to link up 2RATE with other metrics to evaluate the efficiency of not just individual players, but entire systems of play.