West Point has a dropout problem it cannot solve with SAT scores. Every summer, roughly 1,200 new cadets show up and immediately enter Beast Barracks, six weeks of physical and psychological punishment designed to break them down and rebuild them. About one in 20 will quit before the summer ends. The academy spent decades trying to predict which ones. They built a composite metric called the Whole Candidate Score that combined SAT results, high school class rank, leadership ratings, and physical fitness assessments. It was the product of institutional obsession with getting the selection right.
It did not get the selection right. Not for Beast Barracks, anyway.
A 12-question survey that outperformed decades of institutional data
Angela Duckworth, a psychologist at the University of Pennsylvania, showed up at West Point with a questionnaire. Twelve items. Statements like “I finish whatever I begin” and “Setbacks don’t discourage me,” rated on a five-point scale. She gave it to 2,441 cadets across two entering classes before Beast Barracks started, then waited to see who was still there when it ended.
The results, published in 2007, were embarrassing for anyone who believed that talent metrics predict toughness. Cadets who scored one standard deviation higher on Duckworth’s Grit Scale were 60% more likely to survive Beast Barracks. SAT scores? Not a significant predictor. Physical fitness? Also not significant on its own. The Whole Candidate Score, the academy’s crown jewel of selection criteria, told you almost nothing about who would endure the hardest six weeks of their lives.
A 12-question survey about whether you tend to finish what you start outperformed everything West Point had built over decades.
The thing the Whole Candidate Score was actually measuring
This is where I think most people misread the study. The Whole Candidate Score was not useless. Over the full four years at West Point, cognitive ability was the strongest predictor of academic grades. Physical fitness predicted physical performance ratings. The composite score did what it was designed to do: identify people who would perform well under normal operating conditions.
But Beast Barracks is not normal operating conditions. It is sustained misery with no clear end point, where the rational decision on any given day is to quit. The cadets who left were not the weakest or the dumbest. They were the ones who, when everything was stripped away, did not have a reason to keep going that was stronger than the pain of staying.
Duckworth’s Grit Scale measured two things: consistency of interests (do you stick with goals over time?) and perseverance of effort (do you keep working when things get hard?). The combination predicted who would survive a crucible that talent alone could not explain. Later, when Duckworth expanded the research to 11,258 cadets across nine entering classes over a full decade, the pattern held every single time.
Your company’s version of Beast Barracks
Every organization has one. It might be the first 90 days at a startup when nothing works and the product keeps breaking. It might be a reorg where everyone’s role changes and the roadmap evaporates. It might be a quarter where the biggest client leaves and the team has to rebuild pipeline from scratch with no new headcount.
These are the moments when your best-credentialed people are often the first to check out. They update their LinkedIn. They start taking recruiter calls. They do the math on their equity and decide the pain is not worth it. This is not a character flaw. It is a predictable outcome of hiring for talent and never screening for grit hiring perseverance in any systematic way.
I have watched this happen at three different companies during rough patches, and the pattern is always the same. The people who stay and fight through it are rarely the ones with the most impressive resumes. They are the ones who have a history of sticking with hard things, whether that is a sport they played for a decade, a side project they refused to abandon, or a previous job where they were the last one standing after layoffs.
Why you cannot just give candidates a grit test
The obvious response is to add Duckworth’s Grit Scale to your hiring process. This is a bad idea, and Duckworth herself would tell you so. Self-report questionnaires work at West Point because cadets have no incentive to game them. In a job interview, everyone claims they finish what they start.
But the signal is still available if you know where to look. Instead of asking people to rate their own perseverance, look for evidence of it. A candidate who spent four years building a community theater program and is still involved tells you something different than a candidate with six 18-month job stints at increasingly impressive companies. Neither is automatically better, but they predict different things. The first one is more likely to survive your Beast Barracks. The second one is more likely to impress in the first quarter and leave in the third.
The interview question that gets closest to the grit signal is not “tell me about an accomplishment.” That selects for polish. Try instead: “Tell me about something difficult you stuck with for more than a year, and what made you want to quit along the way.” The answer, or the inability to produce one, tells you more than their resume does about what will happen when your team hits a wall.
The critique Duckworth’s own colleagues raised
Marcus Crede at Iowa State ran a meta-analysis arguing that grit’s predictive power comes almost entirely from the “perseverance of effort” subscale and that the “consistency of interests” component adds little. Others pointed out that grit overlaps heavily with conscientiousness, a personality trait psychologists have studied for decades. If grit is just conscientiousness with better marketing, the argument goes, then Duckworth has not discovered anything new.
I think this critique is technically valid and practically irrelevant. Whether you call it grit, conscientiousness, or “the tendency to finish what you start,” the workplace lesson is the same: people who persist through difficulty outperform people who are merely talented when conditions get hard. And most hiring processes measure talent extensively and persistence almost not at all.
The more useful takeaway from the academic debate is that perseverance of effort matters more than consistency of interests. You do not need someone who has been passionate about supply chain logistics since childhood. You need someone who, when the supply chain breaks at 2 a.m., will still be working the problem at 6 a.m. Those are different traits, and Duckworth’s data says the second one is what predicts survival.
If your team is going through something hard right now, and you are wondering why some of your best people are the first to disengage, the West Point data offers a clean explanation. You selected for SAT scores. You never measured who finishes what they start.