Blog
New York, NY, USA
Tuesday 8th November 2016

The final post reporting on the inaugural Orreco Science Summit will concentrate on the area of data science.  The previous posts focused on the presentations delivered at Glenlo Abbey Hotel, Galway, covering sports science and medicine as well as the applied sports science aspects of current and future performance science knowledge.  However, the final session covered a growing field of knowledge that may not have been given airtime 5 years ago, that of data science.


As "big data" becomes a more critical influence in industries from medicine to retail, the sporting playing field is developing an interest as well.  But as more data is collected, less is understood by those filling traditional roles in sports teams and so many gigabytes of unreliable, invalid and noisy data are being misinterpreted and poorly analysed, risking poor decision-making by those in the coaching and management offices. 


Whilst individual teams don't employ the expertise of data scientists fluent in statistical modelling, artificial intelligence, cognitive computing, neural networks, evolutionary game theory and machine learning, these are the specialists who are trained to interrogate the data intelligently, clean the data collected and accurately answer the questions asked by coaches, management, performance science and medical services staff.  As a result, Orreco has naturally evolved by employing a small army of analysts with PhDs in these areas, to ensure they stay at the frontline of the sporting evolution. 


The smart teams recognise this fact and are working with external consultants to build the banks of information that become more and more insightful with every data point collected.  The more traditional teams, or those that are failing to embrace the changes, are continuing with age-old habits and asking sports science staff, or in some cases even strength coaches, to conduct essential tasks in which they have little or no training.  The analogy is that they are asking a cardiologist to conduct brain surgery or a knee replacement because they think a doctor is a doctor...a dangerous and potentially costly mistake to make.  Ironically, in some cases in North American team sports, this is because orthopaedic surgeons or athletic trainers are being tasked with making performance-related decisions, for which they have little or no appropriate training to make.



SESSION 3 - DATA SCIENCE


Stats to the Future.  Professor John Newell.


Professor John Newell is Head of the Biostatistics Unit in the HRB-Clinical Research Facility, NUI Galway, and a Science Foundation of Ireland Funded Principal Investigator in Insight, Europe's largest data analytics research organisation.  Leading the Orreco-Insight partnership, aimed at developing computationally intensive statistical methods to model the likelihood of injury occurrence and identify optimal training load and recovery strategies, Prof Newell is building a team of extremely talented data analysts that are tasked with answering some of the questions posed by research and applied sports science and medical professionals in elite sport, using the masses of data that is collated on a daily basis.


John opened with an example of team data interrogation, to answer a question regarding the relationship between exposure to load and risk of injury.  The example was used to illustrate the fact that, too often practitioners collect large amounts of data and then generate questions based on the data observed.  John underlined the importance of starting with a hypothesis to test from the outset, considering the statistical design needed, collecting data in an intelligent manner, identifying the key set of variables that carry the signal of interest and using an appropriate statistical model, often incorporating prior data, and domain specific expertise.  Once the data has been analysed, the results need to be displayed in a meaningful and insightful manner, to ensure that feedback can be relayed to those who can act upon it, in a format they can digest.


John acknowledged that we are at an early stage of utilising "big data" and stressed that to ensure effective management of the athletes of tomorrow, we need to move to a focus on high quality data, not just high volumes of data.


The Rise of Computational Intelligence in Sport. Colm O'Riordan.


Dr. Colm O'Riordan is an authority on information retrieval, evolutionary computation, artificial life and multi-agent systems and evolutionary game theory, lecturing in the Department of Information Technology, NUI Galway and leading Orreco's cognitive computing and machine learning R&D team.


Colm thankfully introduced the topic with some explanations of terms commonly understood by those working in his field, such as "machine learning", "cognitive computing" and "game theory", as well as those terms used to describe his expertise at NUI.  After setting the scene, Colm continued to discuss some of the challenges for data scientists when it comes to the fusion of both structured (e.g. GPS data, sleep study results) and unstructured information (e.g. coaching notes, social media feeds) that have been retrieved for analysis.  Once collected, Colm underlined the importance of justifying and contextualising results in order to create valid reports, appropriate visualisations to support knowledge feedback and personalise information for different individuals, which can explain situations and inform decisions or predictions.


Colm hypothesised that computational intelligence will continue to evolve, exploring the realms of game theory (the study of strategic decisions and outcomes) to inform reason about conflicting strategies, for example, to understand how the athletes of tomorrow are going to behave or act.




Following on from the data analytics presentations and the subsequent questions fielded by Keith Wood, the day's proceedings were brought to a close by a commentary and gaze into the future by John Nosta.



The Future Begins Now!  John Nosta.


John Nosta is the founder and president of NOSTALAB, a think tank dedicated to the advancement of digital health, which consults to large biotech and pharmaceutical companies, advising them on the future of digital healthcare.  John is also a member of the Google Health Advisory Board and author of the Forbes technology blog, 'Health Critical'.


In a thoroughly engaging summary footnote, John described 'big data' as the "third window into humanity", following the advances enabled by the inventions of the telescope and microscope.  John described a future, where the worlds of diagnostic detection and treatment shift back to back, welcoming in the advent of "stage zero" medicine, made possible by the introduction of nano particles that constantly navigate the body, monitoring the overall homeostasis.


John highlighted the explosive impact that new technology is playing, relentlessly driven forward by actors such as the "power of cool", the "empowerment of the quantified self", big data and, of course, money, on creating an environment of blistering change.  Echoing a discussion I had a little over a month ago, John's statement that "the progress made by mankind in the next 100 years will exceed the cumulative progress achieved over the last 20,000 years", reinforced the magnitude of recent advances in bioscience...and as he poignantly concluded, there's no point in fighting it, because "the future begins NOW!".



Sitting amongst a table full of passionate enthusiasts at dinner, barely an hour later, the conversations ramped up another level and the cross-table collaborations that ensued were incredibly exciting.  If the proverb suggests that, by being the smartest guy in the room, you're in the wrong room, I was extremely comfortably that I was, most certainly in the right room, sitting alongside some new co-conspirators in a freshly formed "collaboratory"...one in which I have every faith will positively and significantly contribute to the changing landscape of biomedical science.



To conclude my three-part review, I would like to thank Brian (Moore), Orreco, the panel of speakers, Keith Wood and my fellow attendees, for an extremely positive and energising learning experience during my three days in Galway.


See you all soon!!!!




References


Cohen, S., Doyle, W.J., Alper, C.M., Janicki-Deverts, D. And Turner, R.B. (2009).  Sleep Habits and Susceptibility to the Common Cold.  Archives in Internal Medicine, 169 (1); pp. 62-67.


Kruger, M.J. And Smith, C. (2012).  Post-contusion polyphenol treatment alters inflammation and muscle regeneration.  Medicine & Science in Sports & Exercise, 44 (5); pp. 872-880.


Milewski, M.D., Skaggs, D.L., Bishop, G.A., Pace, J.L., Ibrahim, D.A., Wren, T.A. and Barzdukas, A. (2014).  Chronic lack of sleep is associated with increased sports injuries in adolescent athletes.  Journal of Paediatric Orthopaedics, 34 (2), pp. 129-133.


Ramakrishnan, S., Wesensten, N.J., Balkin, T.J. and Reifman, J.A. (2016).  A unified model of sleep performance: validation of its predictions across different sleep/wake schedules.  Sleep, 39 (1); pp. 249-262.


Sharp, C. (1996).  The sub-24-minute 10,000 metres, 2040 AD.  British Journal of Sports Medicine, 30; pp. 181-182.


Wright, K.P., McHill, A.W., Birks, B.R., Griffin, B.R., Rusterholz, T., and Chinoy, E.D. (2013).  Entrainment of the Human Circadian Clock to the Natural Light- Dark Cycle.  Current Biology, 23 (16), pp. 1554-1558.



blog comments powered by Disqus