Stats are most effective during games when they're tailored to players' needs

Stats are most effective during games when they're tailored to players' needs
Kerry's Darragh Ó Sé tries to halt Tyrone's Sean Cavanagh, in an era when the Kingdom didn't place any emphasis on the statistical side of the game. Picture: INPHO/Tom Honan

AFTER Tyrone defeated Kerry in the 2005 All-Ireland final, Kerry manager Jack O’Connor really embraced statistical analysis.

O’Connor had never given the practise much thought but the way in which Tyrone clinically picked Kerry off convinced O’Connor that there was merit in studying data and trends.

Around that time, only a handful of inter-county football teams were heavily focusing on in-depth analysis. Armagh, Tyrone and Dublin certainly were but Kerry embraced that culture as the decade progressed.

When O’Connor returned for a second term in 2009, he brought Eamonn Fitzmaurice on board as a selector but one of the Finuge man’s main briefs was video analysis.

O’Connor departed after the 2012 season but he took over the minors in 2014, leading them to that year’s All-Ireland. Video analysis framed a large part of his preparation, but O’Connor had a certain cut-off point with the practise.

He was dealing with younger players. They didn’t have the same attention span, but O’Connor also believed that there was a middle ground to be found with statistical analysis.

Turnovers have always been a key pillar of stats analysis, but O’Connor got rid of that category with his Kerry minors in 2014-15. He understood the importance of those stats, but O’Connor believed that the focus on turnovers was becoming a hindrance – he felt that it was making some of the players afraid to kick the ball, for fear of turning it over.

Of course, there are different definitions of turnovers. There’s a marked comparison between cheaply coughing up possession from a lazy hand-pass, and another player’s attempted pin-point pass, which is just cut out at the last second by a sweeper or an extra covering defender.

The key is how that data is coded and interpreted because, as O’Connor noted that time with young players, the last thing players want with data and analysis is how it can promote a culture of fear.

It may have been in a different sport but a recent study by researchers from the University of Bath’s Department for Health revealed how a ‘Big Brother’ data culture in rugby, driven by performance management, threatens to create heightened distrust, anxiety and insecurity among players.

The qualitative research, based on interviews with ten players, coaches and analysts at an English Premiership club, suggests that data culture in the professional game can have unintended negative consequences on team morale.

It was in a professional environment, but interviewees told of how metrics surrounded their every movement of their lives, on and off the pitch.

As a result, the researchers concluded, players increasingly focused on their own statistics rather than the performance of the team as a whole.

Players were feeling increasingly judged on numbers rather than their true performance. Yet the researchers stressed in the study that it is not necessarily the amount of data collected in training and in matches that is a concern, but rather how data is interpreted and communicated by clubs.

Too many numbers can shackle players’ natural affinity and decision making, because they become almost subordinate to the metrics.

“It is a reflection of where life is,” said Dr Shaun Williams, who conducted the study along with Andrew Manley. “We live in the big data society in which everything is broken down into numbers. The club we researched was an extreme example of that and one conclusion was that a lot of the data lacked relevance.”

Williams told a story of an interview he had read with George Ford (the England and Leicester fly-half), who said that when he goes home from training, he watches reality television to try and switch off from the pressures of rugby.

Different:

George Ford celebrates scoring the opening try for England against Ireland. The out-half doesn't like an overload of stats. Picture: INPHO/Dan Sheridan
George Ford celebrates scoring the opening try for England against Ireland. The out-half doesn't like an overload of stats. Picture: INPHO/Dan Sheridan

Rugby is a vastly different game to Gaelic football or hurling. Andy Farrell, the Ireland manager, watches games in the coaches’ box beside a handful of statisticians with laptops. The game is 80 minutes long while there are a lot more breaks in play to watch footage of what is happening.

Gaelic football and hurling are much faster games. There isn’t the same scope for in-depth analysis during games as there is in rugby, but players and coaches still want to be informed.

The trick is how you relay that information accurately and concisely and within the proper context. Analysts can’t really tell the manager what to do but if the information is strong enough, it will inform their decision in what they may need to do.

With every team forensically analysing each other, how do you manage that information without overloading players with more data and statistics than some can, or want, to take in?

If players' attention span isn’t as wide as it once was, and analysts and coaches keep feeding them small information, how are they going to improve their attention span? That is the balance. If analysts and coaches can provide information that is engaging, then there is more critical thinking involved.

Relevance is everything, especially when the group dynamic is so different. Some players eat information up. More have little or no interest in forensically studying the opposition, but most have a threshold as to how much data they can process.

NUI Galway's Conor Whelan with Huw Lawlor of UCD. Hurling requires analysis even with its off-the-cuff nature. Picture: INPHO/Oisin Keniry
NUI Galway's Conor Whelan with Huw Lawlor of UCD. Hurling requires analysis even with its off-the-cuff nature. Picture: INPHO/Oisin Keniry

There is often a coldness to the clinical nature of statistics and analysis, but player empathy has never been more important.

Analysis, and applying it to the training ground, plays a key part. But the best managers, coaches and analysts still always care about the individual much more than the data.

More in this section

Sponsored Content