The Best Candidate is a Lie

I make video games with people who don't like video games. I made AAA games for 10 years. The turning point that sent me here was years ago when I was crunching on a particularly difficult game, contemplating where it had gone wrong, and noticed how one colleague of mine had tried to consolidate power by pushing away people who were different (in gender, sexuality, and nationality) and pulling close people who was similar. They called themselves the Wolf Pack.

What do you look for when you are building a team? Do you choose people who demonstrates they fit the job description the best? For example, the best programmers from the best schools with the best track records you can find?

Do you choose people who have the same successful traits you value in yourself? Maybe you value people who have a vision of how things can be done in the future. Or people who really get things done. Or people who examine problems from every angle and think of what could go wrong. Or people who have a client-focused approach like you do. Or maybe people who have experience of what has worked and hasn't worked on previous projects and can bring that wisdom.

You've recruited a team of superstars and you have a solid process that has already proven itself. And then you fail. There are many, many factors that can contribute to this.

After that crunch I read a lot, and I learned that one factor can be the failure to take in account that group intelligence is more than the sum of individual intelligence.

The Whole is Not the Sum of its Parts

We didn't used to know this. Around the 1970s some psychologists were involved in what is known as the Person-Situation Debate: Is behaviour largely due to the person, or to their context? The consensus now is that it is due to both, variably, depending on the strength of the personality traits involved and the strength of the situation, and-this is where it gets really interesting-personality traits and situations influence each other as well. Group behaviour is a complex system with feedback loops.

A strong personality trait will tend to be expressed the same across situations. For example, a person who is very quiet will be quiet in both a movie theatre and a house party, whereas a person who is less quiet will be quiet in a movie theatre but not in a house party.

"A strong situation will tend to predict behaviour no matter the people involved. Most people are quiet in movie theatres"

A strong situation will tend to predict behaviour no matter the people involved. Most people are quiet in movie theatres.

But also, people will choose new situations based on their personality traits. A quiet person may tend to stay home more often and go to less house parties.

And people will change situations depending on their personality traits. A house party with only quiet people may become a very quiet house party. And then, if someone arrives who is very charismatic, she may be able to draw out the quiet people and make them less quiet.

And personality traits can change slowly over time. A person who tends to be quiet across situations may slowly over time be quiet less and less until eventually they are no longer a quiet person.

Contexts that can influence behaviour on your team are such things as: Goals, processes. Outcomes. Corporate culture. And the other people on the team. Each individual in the team brings unique traits and perspectives to the equation.

Building a team by considering individuals and not the group you are creating may lead to imbalances and to unpredictable outcomes. If, for example, you build a programming team by finding the best programmer you can find, and then finding the next best programmer you can find, etc., you may not be building the best programming team. If you recruit programmers into your company based on how well they match to a job description, you are doing the same thing.

And if you pull people close to you who are like you, and push people away who aren't like you, you risk losing perspective, losing group intelligence, and losing team performance. While you're busy bonding over things that aren't relevant to work like the local sports team or national pride, you are missing out on leveraging valuable skills and experience and novel approaches to problem solving. In retrospect, I see that it wasn't just that one woeful Wolf Pack. All the more difficult projects I've worked on had this problem. They were all teams that were homogeneous and hostile to outsiders.

You Are Not Your Mind

One reason that we tend to end up with imbalanced and less intelligent teams is because we all have cognitive biases. We don't think correctly.

You might think this doesn't apply to you. You're smart. You're rational. You think that because your mind's own trickery is hidden from you. In fact, the more objective you think you are, the more biased you tend to be. And, insidiously, when one of your biases is brought to your attention, you will believe you made your decision for different reasons than you actually did.

This study demonstrates this: Two applicant profiles were created for a police chief position. One profile was streetwise and one was formally educated. When the streetwise candidate was named Michael and the formally educated candidate was named Michelle, participants tended to choose Michael-and when questioned why, would believe that streetwise characteristics were more important for the job of police chief. But when the streetwise candidate was named Michelle and the formally educated candidate was named Michael, participants would still tend to choose Michael, and would now believe that formal education was more important. This is how we are as humans. Many people who discriminate "feel especially convinced that their selected candidate is the obvious and objective choice."

"The knowledge and intelligence that are required to be good at a task are often the same qualities needed to recognize that one is not good at that task."

The Dunning-Kruger effect

What biases are most interesting when it comes to hiring? I picked three of the many we all share.

The Fundamental Attribution Error is our tendency to attribute behaviour to the person that is actually due to their context. It's why we don't consider the context in the first place and why we have such terms as "the best candidate" when we are talking about hiring.

Then there is the Dunning-Kruger Effect: "The knowledge and intelligence that are required to be good at a task are often the same qualities needed to recognize that one is not good at that task." This is why I think I can do my boss' job better, and then see how wrong I was and how many more subtleties are involved once I get their job. Or why you perhaps not only not value, but also not even see, contributions to a team's success outside of the kind that you yourself provide. It's how you might value hiring people who remind you of yourself and not value hiring people whose skills, approach, methods, or style don't resonate with you.

Third, the Implicit Association Bias is our tendency to make unconscious decisions about people based on messages we've received throughout our lives about people of a certain race, age, gender, appearance, or other grouping. It's why Michael was a better fit for police chief than Michelle. And it's separate from your conscious opinions. It's why when I was once interviewing a young woman for a programming position on my team, my colleague, who was committed to hiring more women, interrupted me in the middle of a question and asked her "Do you play games? How is your math?"

The Best Team is a Balanced Team

You are never going to be able to understand all your own cognitive biases, discover all your blind spots, clear away all your own false perceptions. Then what? Balance your team. It's hedging your bets. It's the right thing to do.

After I read about group intelligence and bias, and wanted to do better going forward, my colleague Ashraf Ismail gave me his framework for hiring teams. I've used it since then.

The first time I used it was for the Child of Light programming team. There was a different atmosphere on this team. It was professional, happy, efficient, and creative. We developed interesting, unexpected innovative features like a mixed 2D/3D pipeline that gave the main character a liquid, surreal, almost underwater movement especially with her hair.

"If you are part of a dominant group, learn more about other people. Study your biases"

Ash's framework is easy to use and effective. You make a list of all the characteristics you want represented on the team: ownership areas, hard skills, soft skills, roles, personality traits, background experience, interests, whatever. And then as you hire people you check off the boxes that they cover. Each person will cover multiple boxes. You try to keep it balanced and you try to check off everything.

This framework will help reframe your thinking so you consider the interactions between your team members and consider your own biases. Go outside your usual networks to find people. Hire more people of colour and women. Don't stereotype them negatively or positively-leverage different talents and perspectives without expecting certain talents and perspectives.

If you are part of a dominant group, learn more about other people. Study your biases. If you are white, watch movies directed by people of colour. If you are a man, play games made by women. Etc.

No Wolves Allowed

One last note about good teams. Diversity is not an excuse to keep dysfunctional individuals on your team. The benefit of diversity is in the increase in perspectives and communication and the decrease in bullying. If your team is balanced and someone is still derailing progress, that person should be removed.

The best teams I've worked with are united around their purpose, and not around extraneous things like outside hobbies. The best teams are headed toward the same goal, from different perspectives and different backgrounds. These teams are more efficient and more creative. I love video games, but I want to make video games for people who don't, so I work with those people to do so.