You are here
Now that the Olympics are over—well, the rugby part anyway—it is time to reflect on the first season of PRO Rugby. Does this first year gie us a sense of the kind of on-field identity the league will have? Does looking at the data from PRO add to or challenge the conventional wisdom?
The game is wonderfully complicated, but when the whistle is blown for a penalty there is a moment in which the choices are finite and the outcomes traceable.
One of the things I’ve learned over the last year of watching rugby and tracking some data is how different the game is from league to league and level to level. If the goal is to understand which decisions are good decisions, one must look at expected outcomes. However, those expected outcomes seem to vary considerably from competition to competition, and then from team to team. The variance is such that the entire enterprise of trying to quantify decisions and their outcomes can seem futile.
I say that things seem to vary because for most leagues, the data available to the public is not that good.
But with PRO Rugby there was a chance to look at something fresh and new and build an understanding of a particular competition from its very beginning.
With the season over, there is a set of 30 matches to look at. Everything here is from my own data collection.
In PRO’s first year, the standard for goal kicking was not high, relative to other professional competitions. That increased the incentive to opt for lineouts, scrums, or quick taps.
The fact that PRO featured less predictable decision-making is good, in my book. However, the league-wide performance at lineouts was also relatively poor. This was another incentive to take penalties quickly or choose to scrum.
Here are the final lineout totals. For comparison, the Hurricanes were the worst performing lineout of Super Rugby with 83.6%.
Here is how valuable different penalty decisions were in different areas of the field. Inside 22 meters, the quick tap was most likely to lead directly to a try. Between 22-40 meters, no quick taps led to a try from a penalty, and no teams chose to scrum that far out.
These numbers use only points scored directly (from the ensuing possession). I do have information on decisions that led to indirect points (like a kick to touch that leads to another penalty but not to a try, and that second penalty results in points), but haven’t yet worked out an easy way to make sure points aren’t being double counted.
The fact that kicks to touch taken between 23-30 meters were so valuable is an oddity for which I have no explanation, beyond noting it is a relatively small sample size.
As mentioned, the standard for goal kicking was relatively low. Still, taking the kick remains the wise option in almost all cases.
Set Piece Value
A scrum inside 10 meters is valuable. Outside of that range, the value of the scrum dips quickly.
Looking at how valuable scrums are in other competitions, I have dragged my feet on conclusions because my assumption was that teams with strong scrums opt to scrum way more frequently than teams with weak scrums. Therefore, the opportunity to scrum near the try line is not inherently valuable; having a strong scrum is valuable. That was my thinking anyway.
In this first PRO season, Sacramento—who quietly ended up have the league’s best attacking scrum—was the team who opted to scrum most often, but San Francisco—who somewhat loudly had the league’s worst scrum—chose to scrum inside the 22 more often than they kicked to touch. Everybody wanted to scrum! The Rush scored three tries from the seven decisions to scrum inside the 22. That is just over two points per scrum. That is not bad.
If even the weakest scrum in the competition does that well, does that mean teams should scrummage more often than they kick to touch?
Looking at every scrum that took place within 10 meters of the attacking side’s try line, tries were scored directly from those scrums 37.3% of the time. Once outside of 10 meters, the likelihood of scoring a try dips considerably.
Looking at every lineout that occurred within 10 meters of the attacking side’s try line, tries were scored directly from those lineouts 34.9% of the time. A lineout inside 10 meters is twice as valuable as a lineout between 11-22 meters.
For this season of PRO, a five meter scrum was more valuable than a five meter lineout. Further, set pieces inside the 22 are not all equal.
So in the PRO matches, the scrums were more likely to lead directly to tries than lineouts. And PRO is not a crazy outlier; in the 2015 World Cup, a five meter scrum was more valuable than a five meter lineout.
Anecdotally, it seems that the decision most of the time for captains is usually whether to kick for the corner or take the shot at goal. If outside of 10 meters, yes, that decision makes sense. If inside 10 meters, the kick to touch is not the smartest play. With more data, the quick tap might turn out to be the smartest play inside 10 meters, but it sure feels like the riskiest most of the time.
As a bit of a digression, those players that can kick to touch from penalties and put their team inside 10 meters for the lineout are valuable. I took a look at all of the penalties between 22-50 meters that were kicked to touch. Individual players did not have enough opportunities to tell who stood out. However, the league-wide information can be used to get a sense of how good a good kick is. Nine percent of the kicks taken between 36-50 meters resulted in a lineout inside 10 meters. So, taking a rough look, if a player kicks to touch 40 meters out and the mark for the lineout is inside 10 meters, that kick is better than 90% of comparable kicks.
The conclusion that territory is valuable is pretty darn obvious, but it is at least quantified for the first PRO season. The conclusion that scrums are more valuable than lineouts—for all five teams in the competition—does not seem obvious to me.
Tries are cool, PRO is a league of tries, so let’s take a look at those.
The reason I say PRO is a league of tries is because the percentage of points scored from tries is relatively high. In the 2015 6 Nations, for instance, 48% of points came from tries; the 2015 World Cup saw 56% of points from tries.
The largest source of tries was lineouts. However, as a percentage, lineouts led to fewer tries in PRO’s first season than in most other professional competitions.
*Quick taps include penalties and free kicks taken quickly.
Unsurprisingly, most of the tries were scored within three phases.
PRO Rugby, which still has lines from the creases when it was in the box, has the chance to be a league that is compelling in part because patterns of play are different from other leagues. What works in PRO might be different from what works in other leagues, and that can be a good thing, so long as the skill levels get closer to those other leagues.