Contact Success and Failure Updates

With some suggestions from Jacob Liberman I’ve updated how I track contact success and failure. You may have noticed this in the stats I recently posted for Vancouver. At some point I will likely update the stats from previous tournaments as well.

The old method considered all made tackles as “contact successes” while lumping missed tackles and offloads together as “contact failures”. First, I never liked that offloads were coded as failures. I don’t believe offloads are inherently successful. Success is determined by the result of the offload; Did it create a break? Did it result in a turnover? This outcome-based classification should also apply to made tackles. Missing a tackle is bad, creating a turnover is good, and making a tackle is just something you do to prevent a missed tackle and create the opportunity for a turnover. As these shortcomings suggest, the old method was unsatisfactory in identifying the success of teams in contact situations.

Now made tackles that don’t result in a turnover and completed offloads that don’t result in a turnover or a break are considered “neutral events”, neither success nor failure. Given that turnovers and breaks are relatively uncommon, most teams will have a large portion of their contact situations classified as neutral. Missed tackles, half-breaks, breaks, and especially turnovers are the significant moments in a possession. The new method classifies these events accordingly.

To help illustrate the change, I included a chart that plots contact success rate and contact failure rate by team through the Vancouver tournament. It’s a first stab at determining whether the new method is useful for identifying teams strong or weak in contact. The rates seem to not only identify differences in teams but do so in a way that correlates with other important defensive metrics like opposition try rate, opposition breaks, and opposition meters. Considering how a large portion of defense is based on contact, this is encouraging.

There is certainly more to be learned from contact success and failure. But tracking those situations according to their outcomes – and, in doing so, describing more precisely which teams are “good at contact” – should provide a better foundation for answering future questions.

January 2017 Update

Over the last few weeks I’ve been posting tables of various stats from the first two rounds of the Sevens World Series. There are currently tables for:
Adjusted Possession and Scoring Rates
Raw Possession and Scoring
Passing and Passing Errors
Breaks and Half-Breaks
Penalties and Cards

I will post others for tackles, rucks, scrums, lineouts, and kickoffs. These tables will include each category’s corresponding errors. But I’m currently working through how best to address those errors. For example, when considering missed tackles, do you lump broken tackles and stepped defenders together or list them separately? Both show the ineffectiveness of a defense but they obviously aren’t identical. What about offloads? Some coaches consider these as missed tackles. Do I lump those in or break them out?

At some point there are too many numbers to make sense of on screen. Consolidation is necessary not just for applicability but readability. To that end I plan to update the tables to allow sorting as well as provide a better mobile experience. Since I don’t have fancy acronyms for the column headers, the long words are squeezed and often break across multiple lines when viewed on small screens. I think the current tables get the point across but need eventual improvement.

These are just a few of the things I’m spending my time on behind the scenes here. As always, if you have questions or suggestions, let me know.