A recent article by the BBC discussed the recent changes in rugby’s laws which more harshly penalize high tackles. It condemned rugby for poor enforcements of its own rules to reduce high tackles and head injuries.
For instance, it said that:
“For a few weeks after the new sanctions were introduced in January referees flashed cards of both colours, then the Six Nations arrived and everything went quiet. An international rugby coach, who does not wish to be named, has said that rugby’s “Wild West has returned. What he meant was that the zero tolerance approach has been watered-down and that high tackles are going unpunished, or not sufficiently punished, in the way that World Rugby’s sanctions demand” (emphasis mine)
“For the first few weekends after the new sanctions came in the cards were out of the pockets. A point was being made and it looked like we were going to get somewhere. You could see the tackles going lower, but we seem to have drifted back. Cards were being dished out for high tackles, but where is that now? It doesn’t exist. You can say that players have adapted. No, they haven’t. Referees have adapted because they’ve stopped enforcing it with the same vigour.”
“It’s back to the Wild West, that’s correct. My sense is that there was a brief flurry of activity to demonstrate that things were happening and then the foot was taken off the gas and we’ve gone back to where we were before. So, effectively, nothing has happened. There’s been no meaningful change.”
Now, let me be very clear upfront – if high tackles are a primary cause of head injury in the sport (which data suggests they are, hence the law change), then a zero tolerance approach to them is required, and they do need to be eliminated from the sport. If they are being missed, or ignored, then it matters. I’ve no doubt that high tackles are being missed, and working towards closing the gap between actual and desired behaviour must be a priority for everyone in the game.
However, as is the case in every discussion, polarization of an issue is rarely helpful, and phrases like “it doesn’t exist”, “watered down”, “stopped enforcing”, “nothing has changed”, and “we’ve gone back to where we were before” are definitively polarized.
Why? Because take the question “Have the new high tackle laws been enforced properly?” In response to that, you could adopt a position that lies on a spectrum of opinions. At one end would be the position that “yes, high tackles are being officiated perfectly, 100% of the time”, and at the other would be the position that “No, nothing has changed, we are back where we were before”.
Adopting a position at either end of this spectrum is to polarize an issue where the answer may lie in the middle. It pulls the discussion apart, putting people who may well share ideals and purpose at opposite ends of a spectrum, and in my experience, that rarely makes for progressive action. It also makes for a terrible way to bring about change – on one hand, you’re shutting off any discussion and burying your head in the sand and claiming things are perfect. On the other, you’re dismissing any progress and coming at it from a position of “you’ve failed”, which is vastly inferior to constructive and informed discussion, especially when data suggest the opposite to that particular extreme, as we shall see.
So I wanted to write this post to refute some of that polarization. I write this NOT to defend referees for the high tackles that they’ve missed – like you, I watch the sport and wish that more would be punished, because yes, some are missed and there is always room to improve. I also work for World Rugby now, as a researcher whose job it is to help support, with evidence, decision-making around issues like player welfare and concussion. That gives me access to this data, but honestly, if you have the stats and want to do this analysis, go for it, and you’ll find exactly the same thing.
The BBC article carries numerous statements and opinions that are demonstrably, definitively incorrect, and I think presenting the data on them will help inform a better way to advance this welfare agenda. We can, and should, discuss this issue from a position of some success and progress, rather than from one of desperation and failure.
I hope that a considered, data-driven, informed position on this helps to frame the discussion in an informed manner, because what that BBC article reports and quotes is simply, objectively, incorrect.
First, a bit of background.
Background – changing the high tackle laws
The change to the high tackle laws was an attempt to reduce the risk of head injury, because research had shown that ‘higher tackles’ (shoulder and above) were four times more likely to cause a head injury than ‘lower tackles’ (below the shoulder), and that an upright body position for the tackler was 40% more likely to result in a head injury than a bent-at-the-waist position
(I actually led this research study, by the way. I have two papers currently in review with scientific journals – as soon as those are published, I will translate and explain the research here on this website too. Nothing must be hidden, full transparency is the way to do this properly).
Based on this research, a group of experts including coaches, players and officials met and decided that one intervention (there are a few more) should be to change the law to punish high tackles more strictly. The objective is to get tacklers bent at the waist, their heads in proximity to the trunk of the ball carrier, rather than head, shoulder and arm. Where before, no penalty was given, a high tackle would automatically result in a penalty, and specific situations would compel a yellow or red card. You can read about the law changes here, but this is a summary of what was changed:
The hypothesis – checking if it is working
If these law changes are being applied correctly, then there are some pretty obvious hypotheses, and one slightly trickier concept that has to be investigated.
First, the number of penalties for high tackles will increase. The number of yellow cards will go up, and so will red cards. That is, the direct application of those laws will see increases in all three categories of foul play sanction.
It makes checking their application over time easy – you just compare month by month figures for the law change to those from the same tournaments played the year before, under the old laws.
So let’s see what has happened to the penalty and card numbers since the law change was announced:
Penalties and cards
This figure shows the high tackle penalties (left panel) and dangerous charging penalties (right panel) in 2016, where old laws applied, and 2017, with new laws, for January, February and March.
I’m expressing the number as penalties per 10 matches for ease of use (gets away from fractions less than 1), and note that “charging” is defined according to the Law book as when a player tackles without their arms, and for this analysis, also makes contact with an opponent’s head.
That figure shows the opposite of “watering down”, and you cannot say, as the BBC article does, that “nothing has changed”. In fact, the number of high tackle penalties has doubled, from 5.5 every ten matches to 11.1 every ten matches this year under the new laws.
Even month by month, you can see an increase in the frequency of high tackle penalties in particular (left panel) – March had 13.4 high tackles/10 matches, compared to 10.5/10 matches in February. This is not “a return to the Wild West” or a “watering down” by any definition.
In terms of dangerous charging tackles, an important point is that high tackle penalties are ten-times more frequent than charging penalties, and with rarer events, there is more chance of random variation. The February 2016 number is higher than February 2017, mainly because Feb 2016 was an outlier (along with Nov 2016, as you can see on the right). Overall, dangerous charging penalties have increased by 59%, from 0.8 every ten matches to 1.3 every ten matches.
The next figure shows yellow cards awarded for high tackles, presented in two ways. On the left, you see the yellow cards awarded per 100 matches (because yellow cards are relatively rare, I’ve used this higher denominator), and on the right, the number of matches that passes per yellow card.
As was the case for penalties, there is substantial increase in the number of yellow cards (which makes the frequency of yellow cards higher, as you see on the right).
Overall, for the first three months of the year, yellow cards have increased by 120%, from 6.5 every 100 matches in 2016 to 14.2 every 100 matches in 2017. Strikingly, the yellow card number in March 2017 was 22.4 every 100 matches, which is almost 3-fold higher than the entire 2016 year average (7.9 yellow cards/100 matches).
That suggests that a directive has been issued to INCREASE the application of the law and severity of sanction, rather than to stop enforcing the law changes, as the BBC article suggested.
What about red cards? I’ve not shown these graphically, because they’re so rare that it doesn’t really add much and because in 2016, there was not a single red card for a high tackle in the first three months of the year.
2017 is different – there have been NINE red cards for high tackles, which works out to 3.9 reds per 100 matches (1 every 25 matches). In the whole of 2016, there were two red cards in 833 matches. The frequency is thus 16.5 times higher.
Relationship between cards and penalties
An interesting ratio to look at is that between penalties and yellow cards. This ratio tells you what proportion of penalties are also sanctioned with a card. If more penalties are being given, but high tackles are being let off relatively lightly without cards, then this ratio would increase, and that would suggest that referees are compromising, to some extent, their on-field decision by stopping at a penalty.
This figure shows that relationship:
Overall, combining the three months:
Jan to March 2016 – One card every 8.6 penalties (overall for the entire year is 8.1 penalties/card)
Jan to March 2017 – One card every 6.5 penalties (Bear in mind this is a mix of yellow cards and those nine red cards)
That’s a decrease, overall, of 19%, which says that referees are in fact giving cards out at a greater rate PER PENALTY than they were before the law change. This should have been expected given the previous data I showed you, that the number of penalties has increased by 100% in 2017, while yellow cards have increased by 120%. In other words, cards are being given disproportionately more than penalties for high tackles.
This is, of course, what you’d hope to see happen, given that the sanction for high tackles is supposed to have increased. There is an argument here, by the way, that the ratio should be even lower than it is, especially in Jan and Feb, where it was very slightly higher, pulled back in March. That is a viable argument, which would suggest that even more high tackles should be punished with cards. One in 4, one in 3, perhaps? I’d have no issue with that.
What would be interesting is an audit of whether those penalties that are not followed by a card are for a reckless tackle (when sanction is meant to be a card) or an accidental high tackle (sanction is only meant to be a penalty, so the decision is actually correct).
Conclusion on the data
A couple of conclusions. First, a caution – it’s early. We are only three months into the year, and so we have had 227 matches compared to 833 matches in 2016. Things may yet change, and that is why this data will be monitored continuously. I also have these data per competition, but it needs more matches for that to add value.
But the picture is very clear – the number of penalties given for high tackles has doubled. The number of cards has more than doubled. Nine red cards are being given where before, there were none. These changes indicate that the law IS being applied, and so referees deserve some commendation for acting on the directive.
Is it being applied perfectly? No, and like many, I watch matches and see incidents where a high tackle appears to have been missed. In an ideal world, they’d all be penalized and the game would look horrendous for a short time with a flurry of penalties and cards, before behaviour changed, and the problem would be largely solved (one hopes).
So this is not a rationalization or justification for those cases that are missed, and referee unions and managers around the world must make the continued reinforcement of these messages a top priority. They are auditing referee performance constantly, and identifying cases that should have been penalties and cards, and they are sending strong messages to referees to do more the next time. This is a treadmill, and we must keep running.
However, what the data shows is that the notion that World Rugby and various Union’s referees have ‘backed off’ or ‘watered down’ the application of the law is unequivocally false. The data all point in the same direction. Sweeping and inaccurate statements are not productive in this challenge. We’re all committed to reducing the risk, and I more than anyone accept that more must be done. But more needing to be done does not mean nothing is being done, or that nothing has changed, or that “we’ve gone back to where we were before”.
An alternative theory
That said, I can understand why this perception that little has changed is developing among observers of the sport, and I’d suggest the following is happening, and also that it is a good thing.
By making the law changes, and adopting a zero tolerance position to head injury, World Rugby is asking for the impossible in the short-term, because it’s not going to be possible to catch every single case.
However, the announcement and application has put into people’s minds the risk of high tackles, and drawn attention to the importance of eliminating them from the game. Then the referees have applied the law, and all of a sudden, twice as many high tackle penalties are being awarded. More than twice as many yellow cards are being given for them. Players are being sent off for high tackles, and that hardly ever happened! (this is shown by the data above)
The observer has thus been “sensitized”, first by communication and then by action. Having been alerted, they now watch matches conditioned to look for high tackles, and so begin noticing those that are NOT sanctioned by the referee. Awareness plus history has thus created a problem of consistency. For some, the (wrong) reaction is that nothing is being done, because they’re noticing the absence rather than the presence of a behaviour (punishment for high tackles).
I would suggest that the number of these “high tackle candidates” per match is the same, or hopefully even less, than it was in 2016, but they’re being noticed more because of the expectation created by the stated objective.
I have friends who tell me every Monday “Why was X penalized and not Y?” and they’re often right. But what they don’t realize is that a year ago, it’s likely that neither X nor Y was given. The observer’s response is often that X was unfair or wrong, rather than realizing that something is better than nothing. Or, as is the case in the BBC article, the reaction is that “nothing has changed” simply because it has not changed enough.
I fully accept that things can, and must, still improve. But I want to caution people against saying it’s all useless just because it’s not perfect. To borrow a metaphor I’ve used in doping conversations, if you were blind before, and I offer you vision out of one eye, don’t reject me and choose to stay blind just because I’m not giving you 20/20 vision in both eyes right away. We will work towards 20/20 vision, but let’s not shout falsehoods with our good eye closed. This language matters, by the way, because what we should be looking for is improvement, not replacement.
The same is true with the HIA process and number of concussions, by the way – the increased awareness of head injury and the introduction of temporary substitutions meant that so many more cases were being noticed, whereas before, a huge number was not even identified. But the result was that those that were missed were in a spotlight that did not exist before. And because the sensitivity of the HIA assessment is around 80% according to data, that was going to happen. But this doesn’t mean it should be demolished. It must be improved. That’s a constructive approach we should be seeking.
Finally, why is this sensitization of observers a good thing? It means there is awareness and recognition of what needs to change. People are recognizing and demanding, through their displeasure and criticism, more change. The authorities and referees now have a strong mandate to change things, and so the pressure on them must be maintained, as I’ve said above. But it’s pressure that should be informed and constructive. There is progress, and we’re pointing in the right direction, so we just need to keep going.
Where to next?
The big unknown in all this, by the way, is whether the actual number of high tackle candidates is changing. Or, put differently, we need a denominator. We need to know how many high tackles actually occur in a match, and how many have been identified correctly and sanctioned appropriately, compared to how many have been missed? That question needs to be asked of 2016 and 2017, so that we can establish whether player behaviour is changing, and whether referee performance is desirable given some “perfect score” for their decisions.
That is an audit I am planning to do. It will be onerous – someone, preferably a referee, will have to watch up to 4,000 tackles and ‘score’ the referee’s decision-making, and also count how many high tackles there are now compared to ‘then’? This historical comparison is really important, but neglected – you can’t frame this as “the sport is getting worse” when you don’t invest in understanding what it used to look like. Unfortunately, surveillance of these outcomes is either in its infancy, or not happening at all, but it’s really important to not look at a stat from 2017, and say “there’s your evidence”. It’s nonsense science in the absence of context and historical comparison.
My hypothesis, conceptually, is that for every 100 ‘high tackle candidates’, about 40 were picked up in the past under the old laws, and that number has now increased to 80, a doubling. It leaves 20 unaccounted for, and those are the ones that sensitized viewers, aware more than ever about high tackles, are noticing when they conclude that nothing is changing.
But that is a theory, a hypothesis, and it needs testing.
Concussions – the outcome of interest
Then finally, the key outcome of interest in all this is concussions. That must go down – if it doesn’t in time, then other interventions will have to be sought (as I mentioned, this high tackle law change is not the only one being considered anyway).
The reason I’m not yet showing the numbers for head injuries is because like charging, they’re rare, and so vulnerable to aberrations, but more than this, their entry into a central database is delayed. So if I presented data from 2017, there’s a chance it would be incomplete, and so I’d be presenting a distorted (and favourable, because the numbers would be too low) picture of what is happening, which would be insincere on my part. Penalties and cards are not the same – month by month, that data is finalized, as soon as the final whistle blows on each match, in fact. Concussions are not. So that must wait, unfortunately, for the lag time to unfold, and then it will be possible to assess whether change has occurred.
In conclusion, let’s keep the pressure up, let’s continue to reinforce zero tolerance, and give the referees a mandate to enforce the laws. But let’s also be real in our assessments, evidence driven, and avoid polarization of the issue.