Head injuries in Rugby Part 3: The key evidence and law change advice

03 Aug 2017 Posted by

This is part 3 of the 3-part series on head injuries in Rugby!  If you missed the first two installments, you can read them here:

Part 1: Overall summary of the research and the concepts of risk reduction

Part 2: What makes a tackle dangerous? Risk factors for head injury

You’ll recall that in part 2, I explained what our research had shown were the risk factors in tackles, and confirmed what you’d have expected regarding the energy transfer model for injury.  Namely, the higher the energy transfer, the greater the risk.  That’s why active shoulder tackles, front on tackles, high speed tackles and accelerating tacklers were more likely to cause a head injury.

However, the key pieces of the puzzle are yet to come, and that’s the focus of this, Part 3.  As I said, if this were a court case, you’ve already seen a number of exhibits and items of evidence.  Below are the two key ones, and in my opinion, were the most interesting findings of the study.

When the expert working group of coaches, players, officials and medics met in Dublin to hear about this research, these were the items that led to the recommendations made earlier this year, and these are the items that will again be discussed later this year as we seek to further reduce the risk in rugby.

For that, read on.



Head contact type

The nature of head contact is important because it tells you quite a lot about technique and the height of the tackle.  When the law change was announced for Jan 1 this year, many coaches and players were quite vocally critical, saying that trying to get the height of the tackle down would actually increase the risk because there would be more impacts between a player’s head and the knee or hip of the ball carrier.

We knew, however, what the relative risks of these situations would be, however, and so were quite confident that lower tackles would in fact reduce the risk, not increase it.  The figure below summarizes why we felt that way.

So head to head contact has a risk that is three times greater than head to knee, and 6.5 times greater than head to hip contacts.  That’s a ‘swap’ in behaviour we were happy to go with, because it would ultimately reduce the number of HIAs.

It stands to reason that head-to-head, head-to-elbow and head-to-knee contact would have higher risks, because bone-to-bone would create more force than say, head to upper body.

One important thing here – note that the risk I’m showing in this figure is actually just for the tackler, and if you factor that in, then head-to-head would have an even greater risk, because there’d be two “vulnerable” heads rather than just one.

So what I did next was to divide these up into contacts that are “high” and those that are “low”.  “High” means the tackler’s head made contact with a point above that blue line, sternum height, so head and shoulder.  “Low” is every contact below the blue line.

In total 130 tackler HIAs happened with “high contact”, and the risk was 3.75 HIAs per 1000 events.  Compare this to 165 tackler HIAs with low contact, at a risk of 0.88 HIAs per 1000 events.

In other words, a “high contact”, involving head-to-head or head-to-shoulder, is 4.25 times more likely to cause an HIA than a low contact.  If this were a court case, then this was to be a very important piece in the process later.


Body position

Final finding for this paper, body position.  The previous finding about head-to-head and head-to-shoulder contact (“high tackles”) points in a direction that body position tried to confirm.  We measured this because we are very aware that technique is a controllable factor, and that a big aspect of tackle technique is the position of the players in the tackle.  We identified three body positions – upright, bent-at-the-waist, and falling or diving (tacklers tend to dive, ball carriers can be either).

We also knew that head-to-head and head-to-shoulder contact (the “high contacts” I mentioned previously) would happen when two players got their heads into one another’s airspace.  We knew from the outset that this was a situation we need to avoid, and so looking at body position was an attempt to understand this a little better.

So, we asked what body position the player was in during the tackle?  Was he upright against an upright ball carrier, in which case the likelihood of that head-to-head contact would be greater?  Or was it a bent at the waist tackler and upright ball carrier, which cannot cause a head-to-head contact, but might cause head-to-hip or knee contacts instead?

I think this is one of the most interesting findings of the whole study:

What you’re looking at above are three panels:

A. Shows the HIA risk when the tackler is upright, and for three different ball carrier positions

B. The tackler is bent at the waist and there are three possible ball carrier position

C. The tackler is diving and there are three possible ball carrier positions

I’ve shown the total number of HIAs from each situation in italics on the figures, and identified the common pattern on each graph with the dotted line – see how the risk of a head injury is lowest for a bent at the waist ball carrier, and highest for a falling or diving ball carrier?  That’s true no matter what the tackler does.

In other words, it’s tackling game theory – it doesn’t matter what the tackler is doing, the best scenario is for the ball carrier to be bent at the waist, and the worst is to be falling or diving (the latter is because a diving ball carrier is unbalanced and unsighted and thus can’t prepare for the tackle – we have some data that shows that this increases the risk for the ball carrier, but I’ll leave that for another time).

One important point here – look at the number of HIA cases.  Only 5 + 17 + 10 = 32 HIAs happen with a diving ball carrier.  In contrast, 131 + 133 + 40 = 304 HIAs happen with an upright ball carrier.

So, realistically, that diving ball carrier is high risk, but not very common, and so between the very common situations, what we would want is a bent-at-the-waist ball carrier, and what we don’t want is an upright ball carrier.  The message from these graphs is simple: Ball carriers should be bent at the waist.

Now let’s do the same exercise, but from the perspective of the tackler:

So, what we see here is something quite similar.  When a ball carrier is upright (left panel, D), the HIA risk is highest when the tackler is upright and lowest when the tackler is diving.  Look at the middle panel – same pattern – highest for an upright tackler, lowest for a diving tackler.  Then look at the right panel, you see the same pattern.

So again, it’s tackling game theory.  The risk of an HIA is highest when the tackler is upright, and that’s true regardless of the ball carrier’s body position.  The risk of an HIA is lowest when the tackler is diving, also regardless of ball carrier position.

So what we would ideally take from this set of graphs is that tacklers should NOT be upright, but should be diving.  Problem is, diving tackles are a very specific type – they tend to happen from the side or back (much like that anomalous findings regarding speed that we looked at earlier), and they’re also less common, only accounting for 67 HIAs, and 23% of all tackles in rugby.  Bent at the waist tacklers account for 52% of tackles, and 233 HIAs, and so practically, they warrant attention.

What it means, then, is we want the tackler, in order of preference, to be: diving, bent at the waist, then least desirable of all, upright.

When we combine all these different scenarios, the result is that the HIA risk is 44% higher when a tackler is upright compared to bent at the waist, and 113% higher when the ball carrier is upright compared to bent at the waist.

The situation is even more striking when you factor in the type of tackle and the direction of the tackle.  I won’t overload you with graphs and figures, but this is the kind of combination analysis I alluded to earlier.

If we combine the risky tackle type (active shoulder) with risky body position (both upright), then the propensity for a head injury is 17.4 HIAs per 1000 such tackles.  That’s almost 10-times higher than our baseline risk, and it’s five times higher than the risk when the tackler is bent at the waist.  To reduce the risk of head injury in the sport, then, we need to discourage upright tacklers from executing forceful, front on active shoulder tackles.


Applying the findings

So where does that leave us?  Here’s a summary of the primary variables that I looked at in Part 2, and in this article, and remember the objective to reduce risk must be either be to eliminate the riskier behaviours completely, or otherwise shift behaviour from high risk on the right, to low risk on the left.

I’ve simplified the study findings into the ‘extremes’ of low risk on one end, high risk on the other.  The questions you have to ask now are:

  • “Which of those variables can be shifted from right to left?”
  • “Which offer the likeliest immediate solution, so that interventions might be prioritized?”

Those are the questions that the Expert group consisting of coaches, players and referees discussed in Dublin last year.  Remember, this expert group is the one that made the decisions, based on the data that we presented to them.  Now that you’ve seen the data (OK, some of it!), you can put yourself in their position, and think about solutions to the issue.

I’ve indicated on the figure above where they felt the likeliest and most immediate changes might be made.

Changing the type and direction of tackles were not considered viable solutions – coaches want physical dominance in the tackle, and that’s achieved through front-on, active shoulder tackles.  Taking this away, or asking any coach to buy into an alternative would be asking them to cede a performance advantage.  It might be theoretically doable, but we live and function in the real world, so it’s simply not viable.  A red cross for those two options.

There remain question marks about the speed and acceleration of the players.  If the game could be slowed down, the data suggest that the risk would be reduced.  Now, there are a couple of ways of achieving this.  One example would be to move the backlines closer together by changing the offside line at certain phases of the match.  This would mean less space to run into, and in theory, lower speeds.  This option has in fact been discussed in various places.  However, our evidence does not support that this would work, and may in fact backfire.

For one thing, we found that the head injury risk was similar after rucks as it was after scrums and lineouts, when the back lines are further apart (data not shown).

The other reason, which I think is very interesting, is that while an increase in tackler speed does increase the risk, so does an increase in the acceleration of the tackler, as you learned in Part 2.  However, remember that important teachable moment I brought up earlier?  The tackler is accelerating into contact only 11% of the time.

Now imagine the backlines were brought closer together.  Do you think this would change the behaviour of the tackler?  I think it would, and I think the change would be an increase in the number of times the tackler accelerates into the tackle, because they are still seeking physical dominance, and if they can’t achieve it through speed because there’s less space, I think they’d accelerate more often.

That 11% would increase, and so would the risk of head injury, based on this data.

Therefore, there are questions around whether it is practical and viable to change the speed of players, and there are questions around whether it would even have the desired effect.  I think that this is an avenue worth exploring, and it will be looked at, but in terms of the technique and what is coached of players, rather than rule change, as has been discussed and suggested.


The immediate viable solution – height and tackler behaviour

So that leaves height of the tackle and position of the players, and it’s here that the Expert Group felt that law change was the most effective immediate strategy to shift behaviour from right to left.

Remember the ‘clues’ – high tackles (as currently ruled by the referee) are 80-times more likely to cause head injury, “higher contacts” defined as head impact above the line of the sternum were 4.25 times more likely to cause head injury (to the tackler, note), and upright tacklers were 40% more likely to result in head injuries than bent at the waist tacklers.

All this points to the importance of lowering the height of the tackler.  Get players into a bent at the waist position, where the head of the tackler (the one most at risk, after all) is out of the greatest danger, and the overall risk of head injury will fall.

The premise behind law change is therefore to protect BOTH players, and not just the ball carrier.  This is a mind shift, a required change in paradigm for people, and when the change was announced, it was lost a little.  The point I am trying to emphasize is that if we can get the tackler lower, getting the height of contact below the line of the sternum, then it is the tackler who will benefit most from a reduction in risk.  Of course, the ball carrier also benefits if we can remove his tackler’s head from the scenario – a lower tackle means the only risk to the ball carrier is a head-to-ground or a whiplash type injury.  But a lower tackle also puts the tackler’s head in a much safer position than it is for higher tackles.

Consider the following:  for every situation with the exception of illegal tackles, the tackler is more likely to experience the HIA than the ball carrier.  Active shoulder tackle?  Tackler, 77% of the time.  Front-on tackle?  Tackler, 67% of the time.  High speed tackle?  Tackler, 78% of the time.  Upright ball tackler?  Tackler, 61% of the time.  Head-to-head impacts?  Tackler, 78% of the time.

At the risk of labouring the point, pretty much every legal situation exposes the tackler to more risk, and the solution to that, based on the data, is to get the tackler into a lower position.  That’s what the intention of the law change was.

Yes, this will protect the ball carrier too, which is a good outcome.  But just because the law is written from the perspective of the ball carrier does not mean that the change in behaviour it produces will not protect the tackler also.


Next steps

So that’s where we leave it for now, but not for good.  The data I’ve shared in these three posts is really only a fraction of what has been collected, and we are continuing to explore that.  For instance, I’m working on a couple of papers right now looking at a foul play analysis and whether the referee made the correct call on the field with respect to dangerous tackles, and another that is comparing HIA events to head injuries that are actually diagnosed as concussions (in other words, we can get an idea of “severity”, and what type of tackle is more likely to cause a severe head injury).

Then, in the not-too-distant future, we are going to explore technique in much more detail.  I’m sure many coaches reading this are wondering about the issue of technique, players getting their heads on the wrong side or the bodies into really poor positions.  That’s definitely a factor, and we have some evidence, courtesy a professional coach, that this is indeed happening.  So the plan is to work with expert coaches, using our data (some of which you see in this post), to identify ways to teach tackling more effectively.

Anyway, let’s leave it there for now, thanks for reading and I hope the findings are clear and give you some good ideas about how to play and coach rugby more safely!



Do NOT follow this link or you will be banned from the site!
%d bloggers like this: