When the Aviva Premiership rolls into action in September, the scrummage will be an area of keen focus after work from the IRB and Bath University to increase safety in this violent encounter is brought in.
There can be no starker illustration of the dangers of a rugby scrum than the death of Chris Tickle. The 23-year-old died from neck injuries sustained during a scrum while playing for Eccles RFC in 2009.
In rugby union, the scrum is a fundamental phase of the game that aims to restart play quickly, safely and fairly after a minor infringement or stoppage. To scrummage effectively requires a pack of eight forwards to coordinate a violent collision and sustained push to gain control of the ball. Given the nature of this encounter it is no surprise that forwards suffer from chronic degenerative injuries and on rare occasions, catastrophic spinal injuries.
Around 8 per cent of rugby injuries are sustained in scrums – a figure arguably skewed by the minor injuries prevalent in the sport. Of the spinal injuries, which often have life-changing consequences, 40 per cent are attributed to the scrum.
Despite this there has been very little research into the forces involved. The International Rugby Board (IRB) decided it was time that it investigated the biomechanical demands that the scrum places on rugby players. The IRB selected the Rugby Science Group at Bath University to investigate scrum mechanics and determine the causes of the injuries suffered throughout the sport. The team of researchers at Bath University follow two main disciplines – biomechanics and injury epidemiology – under Dr Grant Trewartha and Dr Keith Stokes.
In last year's Six Nations tournament 41 per cent of scrums collapsed. The area of the scrum that causes the most concern is when the scrum-half puts the ball into the scrum, so it was this – the initial impact and ball put-in – that received the most attention.
The IRB agreed to fund the research, which was conducted through two phases. Phase one focused on gathering data from real rugby teams scrumming against scrum machines during live training sessions, while phase two involved data collection during team-versus-team scrummaging.
One of the leaders of the project was Dr Ezio Preatoni, lecturer in biomechanics and motor control at the University of Bath. Although not a rugby player himself, Preatoni professes to be an avid fan having watched Six Nations Rugby on TV in Italy before coming to the UK in 2010, armed with a PhD as a biomechanical engineer from Politecnico di Milano, to join Bath University.
The study focused on injuries and the safety of players rather than increasing performance. The ultimate aim was to identify and compare the different control techniques and discover which would most likely reduce the risk of injury for players – not only catastrophic injuries, but also 'overuse' injuries. "Having repetitive loads on shoulders and necks may not only cause an acute injury, but over time will induce degeneration in the anatomical structure of the spine that will cause pain in later life," Preatoni explains. "We tried to understand the forces and movement in different techniques across all playing levels from international to recreational and youth teams."
Monitoring the scrum
The two phases allowed the researchers to establish baseline readings from a very controllable situation. Even though a scrummaging machine was used they tried to make the situation as realistic as possible. Previous research had been carried out in the laboratory under controlled conditions, but this time around the experiments were conducted outside on real training grounds.
"We put our instrumentation on the machine and then took this out onto the pitch," Preatoni says. "In phase one the main system comprised a scrum machine fitted with an NI cRIO-9024 controller. The CompactRIO modules provided the instrumentation for making measurements; the cRIO-9024 acted as an intelligent data logger.
"We fitted the scrum machine with strain gauges and accelerometers to measure forces at play. These measurements are synchronised with video captured from multiple cameras to analyse the player's technique side-by-side with the force data. We also mounted a loudspeaker on board the scrum machine so the CompactRIO system could mimic the referee by playing sound files containing shouts of crouch, touch, pause, and engage.
"In phase one we wanted to measure and compare six different engaging techniques with a controllable situation on one side of the scrum from a measurement point of view."
In phase two the front row players wore an F-Scan sensor from Tekscan that measured pressure exerted between front-row players and inertial measurement units, MTw from Xsens Technology, which measured the accelerations experienced on the trunk and forehead anatomical segments. Four video cameras (two side cameras and one top camera at 50Hz, and one top camera at 200Hz) were installed on the scrum machine's scaffolding to capture movement.
A real-time cRIO-9024 controller synchronised the measuring devices, and was tightly integrated with specially designed NI LabVIEW software. "We used a pre-recorded audio sequence simulating two referees' calls as a timeline for the synchronisation," Preatoni adds. "The recording played on a loudspeaker to ensure consistency for all teams during experimental trials. Lastly, the CompactRIO triggered LED arrays visible in each camera view at the instant of the engage or set referee commands for subsequent time synchronisation of video data and force data to within 1ms."
The step up in phase two was to analyse a live scrum with two teams involved. Whilst it was easy to install strain gauges and sensors on a scrum machine, it was trickier to find wearable sensors to put on the front row of the packs.
"The inertial measurements units were wireless so we had no concerns about cabling, but the pads were more problematic as there was no wireless solution available," Dr Preatoni says. "We had to route the cable through the horizontal structure that we used for video recording; this was eight metres above the scrum. These were the only cables present in the set-up and it was not a hindrance to players."
In phase one six different engaging techniques were used. The baseline for both phases was the crouch, touch, pause, engage sequence.
"In phase one we compared those results with the three-stage call that was implemented last season – crouch, touch, set – and then we introduced a fold-in procedure, very similar to this year's call," explains Preatoni. "It was crouch, touch, pause, engage, but at the engage instead of hitting the scrum machine they had to fold in their shoulders and only apply pressure on a further command. The other two sequences that we used were sequential engagements where we asked the front two rows to engage and then added either the number eight or back row to engage later.
"In the second phase we only had two different call sequences. We had two alternatives which were 'crouch, touch, set', where the only difference was there was no pause; and the other was 'crouch, touch, set' but this time with the bind maintained. With so many teams at different levels and a variety of set-ups it became a very complicated analysis."
The end result is that in this season's professional rugby fixtures the 'crouch, bind, set' sequence will be trialled and if a success will be written into the rulebook for 2015. The main difference is that at touch, the front rows will no longer touch and then move away but need to maintain the binding with the arm. To avoid any confusion the IRB decided to change the second call from touch to bind. Finally, the players don't push until the referee is happy that the platform is stable, he will then call 'yes nine' allowing the scrum-half to feed the ball into the scrum.
The results showed that this new engagement sequence reduces the mechanical load on the players by between 18 and 25'per cent depending on the parameters. "We focused on peak'forces or peak acceleration and peak velocity at engagement and all these were reduced under'the new arrangements," Preatoni says. "This is a result of the fact that the two front rows set up closer to each other because they have to maintain the bind, which reduces the velocity. This, of course, reduces the inertia and therefore the force at impact."
The scrum itself comprises two parts, the initial contact and then a sustained push. If the ball is played into the scrum when it is stable it is the sustained push that is the indication of performance; the force measured during this push is how good the team is at scrummaging. So while this new sequence reduces all the forces at impact, it doesn't influence the force at the sustained push stage of the scrum and thus doesn't decrease the capacity of the team to scrummage.
"The feedback that we are getting from coaches and players around the world is that this is a positive change that enhances player welfare and scrum stability," Brett Gosper, IRB chief executive explains. With the English Aviva Premiership season kicking off in early September, they will not have to wait long for the system to be trialled in earnest.