Data-led team building: increasing the odds of success

“You don’t put a team together with a computer” said Grady Fuson to Billy Beane in the 2011 movie Moneyball. “Baseball isn’t just numbers. It’s not science. If it was, anybody could do what we do, but they can’t because they don’t know what we know. They don’t have our experience and they don’t have our intuition.”

It may be a quote straight out of the Hollywood playbook, but Moneyball is based on author Michael Lewis’ non-fiction work Moneyball: The Art of Winning an Unfair Game, which tracks the Oakland Athletics baseball team’s ascent under general manager Billy Beane and his use of an analytical, evidence-based, sabermetric approach to assembling a competitive team.

The probability of success

Once a novel approach, data-led team selection is now a commonly deployed tactic in many sports. Similarly, technology and data analysis are being adopted by businesses to recruit personnel, select and form teams, and inform strategic thinking. So what are the advantages to using a data-centric approach in both disciplines?

Jonathan Leeder, physiologist at the English Institute of Sport explains that increasing the probability of success by finding areas for marginal gains is the whole point of data analysis. “You can’t guarantee success, but you can increase the odds of it,” he says.

In team sports in particular, technology and data is used to aid team selection, in addition to shaping training and tactics. Bill Gerrard, professor of business and sports analytics at Leeds University, is perhaps better known for his collaboration with Moneyball’s Beane around football, or what Beane might refer to as soccer. He explains how using technology and data helps to “establish a culture of evidence-based decisions which forces coaches to consider all of the factors involved in a decision on team selection, to clarify what the evidence is on each factor, and to make explicit the relative importance of different factors.”

So using data to inform team selection provides a more comprehensive view of likely outcomes. The more you can understand what a player’s peak performance levels are as a benchmark, the more you can assess their condition and readiness for a match. It stands to reason that if you aren’t using the data capabilities open to you, your competitors will be, so the likelihood is you will be behind before you’ve started.

Indeed, data-led team selection has become so integral to strategy and tactics in rugby that the RFU declined to comment for this article. Both the England and Ireland camps cited the “sensitive nature of the metrics” as the reason.

Data analysis, says Gerrard, has changed the tactics deployed in the Six Nations. “It has helped teams identify their strengths and weaknesses and clarify the critical success factors, some of which can be surprising.”

In the business world, data analysis and metrics-led recruitment is also gaining momentum. James Webb is the managing director at Propel London, a recruitment consultancy specialising in technology and digital. Performance data, he says, “plays a key role in a sales person, for example, getting through the door to an interview.”

Scott Ross, chief technology officer at global marketing and technology agency, DigitasLBi, explains that the recruitment market moves so fast that “the more we can pre-qualify our candidates, the more likely we are to find talent within their ever-shortening window of availability.”

People, not robots

Data analysis in sport has its limitations, as it does in recruitment. “Some key aspects of player performance are not particularly amenable to analysis,” says Gerrard. “How to blend together a group of individual players into a team so that the whole is more than the sum of the parts is more a matter of judgement than analysis,” he continues.

Angus McNab, head of rugby at OptaPro, also cites player influence on the group as a whole and leadership qualities as being very difficult to measure.

Leeder agrees, adding that you can’t quantify culture or psychology with data. “As technology gets better and better, it’s easy to become too distracted by it, by too much data. You need to measure things that are contributing to winning the medal – anything that isn’t is arguably a waste of time,” he says.

In recruitment, data-driven selection has similar limitations. Relying purely on data-led recruitment removes the human aspect, which is “the most important part of any team”, says Webb. Teams “are about development of relationships, interactions, challenges, and shared experiences. Purely using data to inform team selection removes any consideration of these factors.”

In fact, it is the recruiters personal relationship with a candidate and understanding of their preferences, ambitions and career objectives that enable recruitment agencies to make “informed recommendations of [a candidate’s] relevancy for a specific position or company.”

Ross breaks down what he considers to be the right balance: “Data is an incredibly important tool in the recruitment process to lower the signal-to-noise ratio, but the face-to-face conversation still remains a mandatory step, and will continue to be the primary decision mechanism for us.”

Top10, a hotel recommendation app based in London, launched last year and was named one of the App Store’s ‘Best of 2014’. The company now needs a strong technical team to work on its app and Harry Jones, co-founder and CTO, explains he is very analytical in how he hires, particularly for the technical team. However, despite using technical proficiency tests and data analysis to inform his decision making Jones still only weights the analytics as 50% of his final decision. “We still want to work with people we get along with, not robots who just complete tasks. The data can’t tell you that.”

A game of two halves

In sport, as in the recruitment of high-performance business teams, all experts interviewed for this feature agree that when it comes to successful team selection, the key is in balancing data-led analysis with gut instinct and other, usually cultural, factors. The skill of a coach is to take all data sources - some quantitative and some qualitative - into account and find the right blend.

“Data gets the person in the door,” says Webb, but gut instinct “is very important and often outweighs data.”

OptaPro’s McNab believes the days of solely relying on gut feeling have gone, but is also a believer in making bold decisions. “This is done with a game-plan and strategy in mind though, so it is a combination of the data and also knowledge of the players and their intangible qualities that will help.”

‘If you challenge conventional wisdom, you will find ways to do things much better than they are currently done,’ wrote Michael Lewis. As the sporting world and the recruitment world continue to invest in data-led team selection, the probability of success will continually rise. A reliance on statistics alone however, will leave a team, an athlete, or a recruitment team, coming up short.

Eleanor Dallaway is editor of Infosecurity magazine

The Creative Data hub is funded by Accenture. All content is editorially independent except for pieces labelled ‘Brought to you by Accenture’. Find out more here.