The framework of winning: 4 principles for data-driven success

While moments of profound, once-in-a-lifetime success may feel unpredictable, finding a formula for winning consistently over time doesn’t have to be a mystery.

But for many leaders moving into a new opportunity with an organization, it can be difficult to know exactly what allowed them to be successful in the past—and how to repeat that success going forward.

Paul DePodesta—chief strategy officer for the Cleveland Browns and longtime sports executive whose unique analytics-driven approach was featured in Michael Lewis' 2003 novel Moneyball—recently talked with Pluralsight CFO James Budge about the parallels between sports strategy and technology strategy when it comes to big data and leadership, unearthing some key principles of success along the way:

1. Challenge your assumptions frequently

When asked what it means to have a truly data-driven mindset, DePodesta was quick to call out a single trait at the center of good data professionals.

“It really starts with humility,” DePodesta said. “A lot of experts in the field—when they’ve been around awhile—can start to think they know success when they see it, or automatically assume they know what’s happening. It’s hard to admit to ourselves as decision-makers that maybe we don’t actually know.”

DePodesta recounted his experiences around Billy Beane, the former general manager and current EVP of the Oakland Athletics he worked under in the early 2000s, as a good example of this kind of humility. 

“Billy used to tell our scouts: ‘Look, I was a first-round draft pick. I was a minor league player, then a major league player. I’ve been an advanced scout and assistant GM. I’ve been in the game for over 20 years, so I’ve seen it from every angle possible. And yet I still can’t walk into a high school game, point at some player and say that he’s going to be a star. So if I can’t do that given my set of experiences, I don’t think that anyone can. At least not well.’” 

With humility as a foundation, data professionals (in sports or technology) are thus forced to find better and more consistent decisions from insights that are objective and actionable, not simply based on the “eye test” or gut feeling.

“We can’t predict the future,” DePodesta said. “So instead, we have to constantly focus on tools and ways of looking at data that can narrow down choices and give us better odds at success.”

2. Investigate your biases

Once you’re in a position where you understand your own decision-making fallibility as a human being, you should begin to also examine what other biases impact your organization.

“You really have to understand the forces acting against your best interest as a decision-maker, so you can arrest them before they lead you down a wrong road,” DePodesta said.

Two of the most common biases DePodesta sees taking hold for sports teams is recency bias and confirmation bias. 

“If a player has a great game one week, then oh, suddenly they’re a star, but if they have a bad game the next week, suddenly it’s time to trade them,” said DePodesta. “Or you have a player that you really want to be good—maybe you’ve invested a lot of time or money into them—so you sort of ignore the things they don’t do well.” 

So how do you level the playing field and eliminate biases with data? DePodesta said that engaging in internal unconscious bias training can help “allow yourself to see the whole picture” and mitigate the impact that emotions have on decision-making.

3. Get as many people involved as possible

While a bias toward a single individual’s experience can hinder data-driven decisions, when you pull back to a macro view, it can actually become a strength to have a large group of people involved in decision-making—because of their unique experience, not despite it. 

And it’s not just engaging a large quantity of people that makes a difference, but ensuring that the people you draw insights from are as diverse as possible—in terms of demographic factors, professional background, personality and skill sets.

“People's individual expertise drove a lot of insights we’ve gotten from data over my career, because the viewpoints or angles from which they were looking at things allowed us to tease out findings or conclusions we never would or even could have thought of on our own,” DePodesta said of involving people outside his leadership team. 

Not every high-level decision made by a sports team or technology organization can be democratized; accountability is still important, and the buck does have to stop somewhere. But many companies could benefit from engaging their employees in thoughtful collaboration more often, which has the added benefit of getting stronger organizational buy-in from your employees sooner.

DePodesta referenced a recent personnel decision the Cleveland Browns made where the team involved some 60 people, from equipment managers and conditioning coaches to scouts and public relations professionals, in a decision typically reserved for the team owner alone. This resulted in a faster hiring decision that people were excited about.

“It goes back to the concept of knowing that you don’t have all the answers,” DePodesta said. “There are probably parts of the organization that have some of the answers, and another part that has some of the answers, and they can help you piece the puzzle together over time. You can get high-quality information from unexpected places.

“They don’t need to sit in on every interview and meeting in order to be involved. They just need to feel like they’re part of the solution.”

4. Create a shared vision from concrete goals

It’s hard to turn success into a repeatable process if you don’t know exactly what it is that allowed you to accomplish past successes. This is where the concept of a shared vision comes in.

“As an organization, you have to agree on some fundamental principles that everybody buys into,” DePodesta said. “This then helps you make your decisions, because it’s a shared vision for success that everyone, no matter their role, knows that if they execute on it, the team is ultimately going to be successful.” 

In the case of sports teams, DePodesta described the process of drilling down on your goals until you are at the most discrete, tangible pieces of that goal. From there, you’re able to take action.  

“If you ask any NFL team what their ultimate goal is, they’re going to say ‘We want to win the Super Bowl,’” DePodesta said. “So you have to ask yourself, ‘OK, how are we going to win enough games to make it there?’ And that means you have to ask yourself, ‘How are we going to put the right talent on the field to win those games?’ And that requires you to ask, ‘What skills and metrics are we prioritizing as we evaluate players?’ You just keep backing it up until you get to your core principles for winning.”

A final thought on trusting your data

Even with all these steps in play, it can be nerve-wracking to jump into a data-driven strategy with both feet, as contradictory as it sounds.

“We like to say that we are some of the most uncertain people in the building,” said DePodesta. “But at the end of the day, you have to do what you believe is right, and what’s going to give you the best chance of being successful.

In the face of that uncertainty, DePodesta stressed the need to trust what might look like a risky decision at face value when you have enough quality data to back up your decision. 

DePodesta gave the example of fourth-down strategies in football, one of the most scrutinized, high-pressure decisions that team and coach will make over the course of the game. Even with a strategy that produces a hypothetical 70-percent success rate on fourth downs—which by all considerations should be considered an enormous accomplishment—you have to acknowledge and be comfortable with the 30 percent of failure that’s part of the process.

“A lot of teams are now realizing that it pays to go with the math, to understand the risk but to also understand when the upside makes it a worthwhile bet,” said DePodesta. “But that doesn’t mean you’re not nervous as it’s playing out in front of your eyes every single time.”

Data professionals are often considering an incredibly wide range of possibilities and directions to go when working with raw data, but once you feel good about the insights you’ve gathered, you can’t be constantly second-guessing your decision-making—even if you do have the occasional “Are we really doing this?” moment.

“Even if it has a good payoff, you’re still going to miss, and in each of those situations the critics and journalists are going to point to the misses as an example of why your strategy doesn’t work,” said DePodesta. “It’s tough, but you have to get used to facing that criticism and remind yourself that the bet is still paying off.”