March 26th, London’s Emirates Stadium and 200 sporting minds in attendance at the Innovation Enterprise Sports Analytics Innovation Summit. Whereas the MIT Sloan Sports Analytics Conference in Boston is (naturally) a very US-sports-centric event, the London summit, now in its second year, had promised to be based somewhat more around European sports, notably soccer, rugby and Olympic events.

Who was in attendance? Quite a varied crew. Soccer, rugby (union and league), australian rules football, bobsleigh, swimming, sports law… and more. Even the odd journalist. Hat tip to Brent Rogalski of the AFL’s West Coast Eagles who flew to London from Perth after the Eagles’ last game and will be back in Melbourne for their next one next Sunday.

Performance Determinants, Swimming and Injury Prevention

The early morning sessions on Wednesday were a mixture of Olympic sports and soccer. Steve Ingham of the English Institute of Sport led out the morning with a talk on “Determinants of Performance: the framework for training intervention”. Ingham’s talk centred around the notion of understanding what marks were really important in a sport in terms of impacting the end result (i.e. success) and then linking training to those marks. In multidisciplinary sports such as heptathlon this gets a little more difficult, and decisions have to be taken on what areas should be concentrated on for a particular athlete to gain most improvement in their overall points score.

Swimming followed. Karl Cooke of British Swimming broke down a swimming race into its component technical disciplines, including the bit that the typical viewer would usually call “swimming” but also adding the start, the turn and time spent swimming underwater. Being able to break down a race in this manner allows swimming analysts to plan for a “perfect” race, then benchmark an individual swimmer’s marks against each of those technical disciplines. When expanded beyond a single swimmer to entire groups of swimmers (e.g. a potential Olympic training squad) it’s then possible to analyse which areas the group as a whole need to work on and which are already being executed as a high level. This is of course important in terms of use of both funds and training time, both of which are usually limited in some manner.

Sam Erith took the stage before the morning caffeine recharge to chat about the various things being done in football to manage injury prevention. Erith is currently employed by Manchester City and formerly of Tottenham Hotspur; he said he was speaking of his experiences as a whole rather than MCFC specifically.

Erith noted the battery of pre-training tests that are done before a player steps on the field in the morning as well as using data to quantify training and rehabilitation. They can now tell how easy or hard a training session is, or is planned to be, and pull players in or out accordingly. One interesting point he made was in the use of data to adjust players’ recovery schedules in busy times of the season; if a player exceeded certain thresholds based on his individual maximums over the previous six weeks or so, with a game just three days later they would need more recovery than usual.

The coffee break saw a dash for various brown, caffeine-filled liquids along with the expected assortment of biscuits and mini-muffins. These breaks are for many the most valuable times at conferences as it allows them to chat to people in the same or different sport in an informal, collegiate setting.

Rugby: GPS and Technical Direction

Although there was an off-field analytics stream in the late morning I stuck to the on-field side of things and took in rugby talks from Dr. Kevin McDaid of Dundalk-based company STATSports and Steve Aboud, the IRFU’s head of Technical Direction. STATSports are in the GPS game, offering data on distance, acceleration and collisions. McDaid mentioned the potential use of collision data by broadcasters during game action, for example a graphic showing that a player had been involved in a collision of a large number of G forces, but also noted that this would need careful handling by the rugby powers that be. After all, he said, a mother of a nine year old might not be hugely impressed by the thought of her offspring being involved in such collisions. However, the potential was there to get across the collision intensity of a top level rugby match to TV viewers.

Steve Aboud, fresh off Ireland’s Six Nations victory, talked about developing effectiveness as a team. He presented this in terms of the IRFU’s Technical Model that was developed in order to shape the direction of the nation’s rugby efforts and ensure it was ultimately related to impacting the top level of the game via appropriate training/interventions at the various stages of a player’s rugby life. A key point in Aboud’s presention was the importance of developing an understanding of the realities of the game/sport in which one is involved, rather than what one’s perception of it might be.

Rankings and a Holistic Approach to Training

After lunch Peter Vint of the US Olympic Committee presented on Application Methodologies to Evaluate Competitiveness. It was interesting stuff, applying relatively straightforward models to sports to estimate which team would be more likely to beat another on any given day. In the NFL such models have an accuracy rating of 65-66%* which, Vint said, compares favourably with the average Vegas house winning rate of 68%. In his world this provides insight in a number of areas, including measuring the competitiveness and standard of particular competitions which then provides the USOC with context for the results of their athletes and even whether to send athletes to particular competitions in the first place.

Rene Wormhoudt gave a quite wonderful talk on the Athletic Skills model of developing talent. He said that early specialisation was not good for development and it is better for kids to play and train in multiple sports and disciplines rather than simply concentrating on one. Why? For one thing that person will become a better allround mover, with better balance and coordination. They will likely be less susceptible to injury (Wormhoudt said that kids who played just one sport were four times more likely to be injured than kids who played multiple sports) and can be more creative in coming up with mid-game solutions. All in all it was a fascinating talk and, for somebody in attendance from Ireland, it suggested that kids being exposed to multiple sports such as rugby, soccer, GAA and hurling at a young age, to say nothing of individual sports, is potentially a very good thing in the Irish sporting landscape.

More Rugby – the Saracens Analytics Experience – and a Panel Discussion…

With the Heineken Cup quarter finals on the horizon it was apt that there were three members of the Saracens back room team at the conference, including Bill Gerrard who talked about Creating a High Performance Analytics Culture: The Saracens Experience. Gerrard brought the audience through the stages of the analytics journey at Saracens and provided a brief insight into what he thinks might come in the future. The cultural part of the equation was mentioned more than once, with Gerrard reckoning that although many other teams have access to far more data than them, they make very good use of it indeed.

The day’s sessions were concluded with a panel – Challenges, Gaps & Future of Analytics in Sports – hosted by Blake Wooster, the day’s M.C., with Simon Wilson of Manchester City, Adam Beard of the Welsh Rugby Union and Chris Anderson, co-author of The Numbers Game. Beard mentioned a time when he worked in Aussie Rules back in 2000 when they put heartrate monitors on players. There was a 10,000 dollar fine for this but it was decided to do it anyway in order to get a competitive advantage.

Wilson said that the starting point for people starting out as analysts now was massively different, even in terms of the availability of data in our society being so much greater. He also said that they used to spend the majority of the week on opposition analysis rather than analysing their own players. His point was that in retrospect this was odd, given that ones own players are more controllable than the opposition.

Chris Anderson made the point that good decisions are based on logic and evidence; they don’t necessarily have to be quantitative. He also said that there’s an enormous amount to be learned from looking outside ones own usual world both in terms of hiring and insight.

Beard made an interesting comment about wanting to understand the contact load on players but there potentially being an inherent dilemma in reporting those loads given the recent experiences of the NFL. He also said that some senior players on the 2013 Lions tour from other countries (i.e. not Welsh) had not wanted to wear GPS units. Beard said that Warren Gatland told those players that he’d respect their decision but they’d be playing on Wednesdays. When asked if the large Welsh representation on the Lions tour perhaps impacted the team’s subsequent form in terms of fatigue Beard immediately said no. England and Ireland both had large representation, he said, and England almost won a Grand Slam while Ireland won the Six Nations championship (and Ireland did a very good job on Wales tactically). A very generous attitude.

A few thoughts…

And so, the day drew to a close at the Emirates. A day spent talking to interesting people about interesting sporting things. Over a few beers various wonderings were put forward, such as the decision to go down the individual speaker route over panel discussions. Panels can potentially provide more entertainment, certainly, but (networking aside) whether standing alone at a lecturn or seated in comfortable armchairs the audience are mostly there to pick up little nuggets, glimpses of insight that they can perhaps use in their own sports or at their own clubs. 30 minute talks, including Q&A and changeovers, perhaps are a little short to develop issues or insights in more depth.

The absence of Premier League soccer clubs was notable at the Sloan conference in Boston and their attendance at this summit was not overwhelming, especially when considering its location in London. In all, 16 soccer teams were represented (including lower league and non-English sides) while there were 12 from rugby (including international sides). Considering the small size of the rugby world compared to soccer, that was surprising. I asked two delegates if most soccer clubs are so far ahead of the curve they didn’t see value in such conferences? Both laughed.

“Sports Analytics” is alive and well, but that term shouldn’t always be associated with log tables and pocket protectors. Behind all the big words, graphs and methodologies lies the one desire shared by every athlete and coach there has ever been – the desire to gain a competitive edge.

*Updated 27/03/14 15.01: originally written in error as 65-68%

Related pieces that you might find of interest:

Sloan Sports Analytics Conference 2014: Day 1 Review

Sloan Sports Analytics Conference 2014: Day 2 Review

Podcast: Second Captains chat about Stats in Sport