Shaping Analytics In Rugby

How an economics expert transformed Saracens’ use of data


Compared to the obvious formational diversity of sports like football and soccer, the nuances of rugby’s tactics are less clear to the uneducated eye. The nature of the game dictates that teams line up with eight forwards and seven backs. These units line up in a one-to-one reflection of their opponents, meaning the real tactical battle lies in playing style rather than formational change.

In this sense, rugby is more a sport of marginal gains than football or soccer - both sports in which formational overhaul can drastically affect the outcome of a competition. It’s not so much a sport where a single moment of brilliance can determine a result; it’s one where a relatively clear winner often emerges gradually over the course of the 80 minutes. Because anomalous results are less likely and success is more directly tied to performance, improvements made through analytics-driven decisions are more easy to quantify.

For example, a longer kicking game snatches territory, territory yields points. Therefore, a team that commits to kicking long for territory should, in theory, see more points. This may be a simplistic view, but it’s an example of one of a plethora of decisions on style a coach has to make to affect a sport without formational fluidity. Rather than lining up in a particular fashion, rugby teams are forced to alter the specifics of their playing styles to counter threats or exploit weakness.

Bill Gerrard, who came into the sport by his own admission ‘as a complete outsider with absolutely no background whatsoever in rugby union,’ became an important figure in the Saracens RFC analytics team. With an initial background in economics, Gerrard came to sport with the skills to pull apart data and work it into actionable insights, for which he has been lauded in the sporting world. Referring often to the ‘evidence based approach’ made possible through analytics, Gerrard is an expert at identifying important metrics and turning these into KPIs for teams to work towards improving.

For Gerrard, the work he does ’combines the coaches’ experience and intuitions with video analysis and data analysis as part of that - helping us to inform decision on training priorities, on team selection, on tactics, and longer term on player recruitment and player retention issues.’

In an industry struggling to make sense of the multitude of data it now collects, Gerrard’s approach is refreshing. Rather than scanning reams of passing data to identify patterns and hope to uncover insight, for example, Gerrard first establishes what it is the team are specifically trying to achieve. Once he has this, he can identify the metrics that influence these goals and, from these, form KPIs. ‘If it isn’t going to influence decisions, don’t waste your time and resources getting the data and analyzing it. You’re wasting your time.’ The notion that analytics must fundamentally be focused on impact, on changing behavior, underpins Gerrard’s approach to the practise.

‘We’ve talked so much about big data in analytics conferences, but you almost get the sense we’re getting back to basics, that small is beautiful,’ he said. ‘It’s that expert, detailed knowledge and data on our individual players, within the context of what it is that we’re trying to do within our organization.’ Essentially, context is key. The more you have, the less variety of data you need.

Big data is effective in providing a background and identifying generic trends and patterns, but it has to be integrated with small data to provide the why. ‘You need causality. More and more, big data is about correlation, not causation. It’s interesting and useful to know what the correlations are, and indeed you can make a business plan if you’re Amazon or Facebook on that. But coaches, sport scientists, teams, and athletes can’t. They need the small data.’

So rugby is a sport of fine margins. Small tweaks to playing style can have a major impact on results. It’s fitting, then, that analytics teams should focus on small data to inform decision making, as well as the vast seas of big data they otherwise identify trends within. For Gerrard, the sport will reach a point where the coach’s bench responds to analytics in real-time. It’s up to people like Gerrard and his team, however, to ensure that the insights available are both valuable and actionable for the decision makers. 

Vision small

Read next:

Big Data Forecasting In Pharma