Take every cadet who entered West Point between 1949 and 2022, pull the ones with usable data on cognitive scores, personality, and performance, and pool them into one dataset. That’s 29,102 people, drawn from 15 sources across 21 independent samples, spanning seventy-three graduating classes. Then ask the obvious question the institution has been trying to answer since before most of these cadets’ grandparents were born: what actually predicts who becomes a good leader? That’s the design behind Brain, Personality, and Leadership: A Meta-Analysis of Performance Predictors Among West Point Cadets, published Oct 3, 2026 in Military Psychology by Benjamin J. Elliott and Nathan R. Kuncel of the University of Minnesota, covered the same day by PsyPost.

The finding, in short: cognitive measures — standardized test scores, high school class rank — predicted academic success well and leadership performance weakly, and were almost entirely unrelated to physical performance. Conscientiousness did the opposite. It was the strongest non-cognitive predictor of leadership, the thing cognitive ability was supposed to be measuring and mostly wasn’t. Heraclitus never saw a regression table. He didn’t need one. Sometime around 500 BCE he wrote seven words — ēthos anthrōpōi daimōn — that a University of Minnesota meta-analysis just spent 73 years of institutional data confirming: character is fate for a person, not intelligence.

The Quick Verdict

Question	What we found
What was studied?	Which traits — cognitive and non-cognitive — predict academic, leadership, and physical performance among West Point cadets
Who ran it?	Benjamin J. Elliott and Nathan R. Kuncel, University of Minnesota
Where and when?	Military Psychology, published Oct 3, 2026; covered the same day by PsyPost
Sample	29,102 cadets, West Point graduating classes 1949–2022, pooled across 15 sources and 21 independent samples
Cognitive measures	Test scores and class rank strongly predicted academic success, weakly predicted leadership, barely touched physical performance
Core finding	Conscientiousness was the strongest non-cognitive predictor of leadership performance
Grit and hardiness	Grit showed only a modest leadership link and smaller academic/physical ties; hardiness showed a very weak leadership link
Other Big Five traits	Extraversion, openness, agreeableness, and neuroticism showed little to no relationship with leadership ratings
What Heraclitus already argued	Fragment 119: ēthos anthrōpōi daimōn — character is fate for a person, roughly 2,500 years before “personality trait” was a phrase anyone used

The Quick Version

A meta-analysis of 29,102 West Point cadets found that test scores and class rank — the things institutions lean on hardest to predict future performance — strongly predict grades and barely predict leadership. Conscientiousness, a trait nobody puts on a transcript, predicted leadership better than anything else measured. Heraclitus said something close to this before anyone had the word “personality”: character is fate for a person. The study didn’t discover a new idea. It put a number on an old one.

What the Study Actually Found

The useful thing about this meta-analysis isn’t the headline — “personality matters too” is not a controversial claim in 2026. It’s the scope. Elliott and Kuncel pulled 29,102 cadets across seventy-three graduating classes, used artifact-distribution methods to correct for measurement unreliability and range restriction (the statistical headache that shows up whenever you study a population, like West Point admits, that’s already been filtered hard before the study even starts), and sorted outcomes into three buckets that don’t usually get analyzed side by side: academic performance, military leadership performance, and physical performance.

Cognitive measures did exactly what you’d expect in one of those buckets and almost nothing in the other two. Standardized test scores and high school class rank were strongly associated with academic success — no surprise, that’s what they were built to predict. But their relationship with leadership performance was much weaker, and with physical performance it was close to zero. A cadet who tested brilliantly in high school was likely to do well in the classroom at West Point. Whether that same cadet could lead a platoon under pressure or hold up physically was, statistically, almost a separate question.

Conscientiousness crossed all three buckets. It was the single strongest non-cognitive predictor of leadership performance specifically, and it also showed real (if more modest) ties to academic and physical outcomes — the trait that generalizes where the test scores didn’t. The rest of the Big Five didn’t come close. Extraversion, openness, agreeableness, and neuroticism showed little to no relationship with leadership ratings at all, which should sting a little if you’ve ever assumed the loudest, most socially fluent person in the room is the natural leader.

Grit, the trait that spent the last decade as a TED-talk fixture, came out weaker than its reputation. It showed only a modest relationship with leadership performance, and an even smaller one with academics and physical outcomes. Hardiness fared worse still — a very weak relationship with leadership. Both traits are, roughly, about persistence under difficulty. Neither came close to conscientiousness, which is a narrower, less dramatic thing: the tendency to be organized, dependable, and disciplined about follow-through. Grit is the trait people want to have a TED talk about. Conscientiousness is the trait that actually shows up to the TED talk on time, finishes its slides early, and double-checks the slide clicker beforehand.

Does IQ Predict Leadership?

Not well, according to this data. Specifically:

1. Cognitive scores predicted academic success strongly — test scores and class rank did what they were designed to do in the classroom.
2. The same scores predicted leadership performance only weakly — the correlation dropped sharply once the outcome shifted from grades to how cadets actually led.
3. Cognitive scores were almost entirely unrelated to physical performance — a near-zero relationship, the weakest of the three outcome domains.
4. Conscientiousness predicted leadership better than cognitive ability did — the trait that doesn’t show up on a transcript outperformed the one that does, on the exact outcome IQ was assumed to forecast.
5. Most other personality traits didn’t predict leadership either — extraversion, openness, agreeableness, and neuroticism showed little to no relationship with leadership ratings, so this isn’t simply “personality beats cognition.” It’s conscientiousness, specifically, beating cognition.

What Did Heraclitus Mean by “Character Is Fate”?

Heraclitus of Ephesus, writing around 500 BCE, left behind a fragment — ēthos anthrōpōi daimōn — usually rendered “character is fate for a person” or “a person’s character is his destiny.” Ēthos means habitual disposition, the settled pattern of how someone acts, not a passing mood. Daimōn means something closer to fate, fortune, or guiding spirit than “demon” in the modern sense. The claim: who you are, repeatedly and reliably, is what determines what happens to you — not a separate force called luck.

Scholars have argued for a while about which direction the fragment points. One reading treats daimōn as something handed down externally — the gods apportion both your character and your fate, the way Heraclitus talks elsewhere about divine forces assigning human roles. The competing reading treats it as a claim about agency: your habitual disposition isn’t assigned to you from outside, it’s something you build through repeated action, and that built pattern is what generates your outcomes. The second reading fits better with the exhortations elsewhere in Heraclitus’s fragments — you don’t lecture people about their choices if their character was handed to them at birth like a coin flip.

Either way, the fragment is doing something specific: refusing to separate “who someone is” from “what happens to them.” Most of us keep those as different categories. Talent, intelligence, circumstance — the stuff that determines outcomes — on one side. Character — the stuff we praise or blame people for — on the other, treated as a nice-to-have, a moral decoration sitting on top of the “real” predictors. Heraclitus is saying that’s backward. The habitual disposition isn’t decoration. It’s the mechanism.

Why the Study and the Philosophy Rhyme

The connection here isn’t “an ancient Greek predicted a 2026 personnel study.” It’s narrower, and it holds up better for being narrower: Heraclitus’s specific claim was that ēthos — settled, habitual character — is the thing generating a person’s outcomes, not an external stroke of fortune or talent. This study measured a modern proxy for ēthos (conscientiousness is, almost definitionally, habitual reliability) against a modern proxy for daimōn-as-talent (cognitive test scores) and found the fragment’s bet pays off on the one outcome — leadership — that actually requires sustained, repeated conduct under pressure rather than a single test performance.

It connects to the practical wisdom Aristotle called phronesis — the capacity for good judgment that, unlike raw intelligence, has to be built through repeated practice rather than measured on a single exam. Phronesis isn’t something you have. It’s something you’ve done enough times that it’s become who you are. That’s a conscientiousness-shaped virtue dressed in Aristotelian language three centuries after Heraclitus.

It sits close, too, to Aristotle’s and Cicero’s account of courage — the idea that courage isn’t the absence of fear but the habitual, repeated choice to act rightly despite it. A cadet leading under pressure isn’t drawing on a single burst of bravery any more than a conscientious person is drawing on a single burst of discipline. Both are patterns, not events. That’s the whole difference between a trait a meta-analysis can detect across 29,102 people and a test score that only measures what you knew on one Saturday morning.

There’s a sharper fit still in Mencius’s account of the moral sprouts — his argument that virtue starts as something small and native in a person but requires cultivation, repeated tending, to become reliable character rather than an occasional impulse. Conscientiousness, in the personality literature, is also understood as something with a baseline but real room to develop through practice and habit. Mencius and the Big Five researchers are closer than either camp would probably admit.

And it rhymes, from a different angle, with Heraclitus’s own argument that you can’t step in the same river twice — the fragment about flux, logos, and the rational structure running through change. Ēthos anthrōpōi daimōn isn’t a contradiction of that doctrine. It’s the stable thing logos was pointing at all along: not fixed circumstances, not a fixed self, but a reliable pattern of disposition that holds its shape while everything around it moves. Character is the logos of a person — the structure inside the flux of a life.

The Gap Heraclitus Didn’t See Coming

Here’s where the comparison gets more interesting than a clean match, and it’s worth sitting with rather than smoothing over.

Heraclitus’s fragment is genuinely ambiguous about whether character is given or built — the determinist and free-will readings of daimōn have coexisted in the scholarship for a reason, because the Greek doesn’t settle it. The modern personality literature isn’t ambiguous on the equivalent question: conscientiousness has a real heritable component, but it’s also known to shift with age, role, and deliberate practice — it’s not fixed at birth the way a cognitive test score at seventeen tends to be fairly predictive of a cognitive test score at forty. If anything, the more “developable” half of the ancient ambiguity has the better modern evidence behind it.

That matters practically. If the study had found conscientiousness just as fixed as cognitive ability, the finding would be fatalistic in the way the determinist reading of Heraclitus is fatalistic — interesting, but not actionable. Instead it points toward something closer to the free-will reading: character is fate, but it’s a fate assembled out of habits, which means it’s a fate with some say in it. Heraclitus left that door ambiguously open twenty-five centuries ago. The personality researchers, without citing him, walked through it.

What This Study Doesn’t Show

It’s worth being precise about the size of this, because the framing invites more confidence than the data earns.

A Single, Highly Selected Institution

Every cadet in this dataset already cleared West Point’s admissions bar before the study ever touched their data — which means the cognitive range in this sample is narrower than the general population’s, even after statistical correction for range restriction. Whether the same pattern holds among populations with a wider cognitive spread is a separate, untested question.

Military Leadership, Specifically

“Leadership performance” here means what West Point’s evaluators rated as leadership performance, inside a military command structure with a particular culture, hierarchy, and definition of what good leading looks like. Whether conscientiousness predicts leadership the same way in a startup, a hospital ward, or a classroom is plausible but not what this study measured.

Seventy-Three Years of Changing Instruments

Pooling data from the class of 1949 through the class of 2022 means pooling across decades of different test instruments, different personality measures, and a military and cultural context that changed enormously over that span. The meta-analytic correction methods are built to handle some of this. They can’t make a 1950s-era assessment and a 2020s-era assessment into the same instrument.

Correlational Within an Already-Selected Group

This is pooled observational data, not an experiment. Nobody randomly assigned conscientiousness to cadets. The study tells you which traits moved together with which outcomes inside a population that had already been selected on some of those same traits — it doesn’t tell you what would happen if a military academy tried to select differently.

What Actually Follows From This

1. Stop treating cognitive screening as a leadership proxy. If your selection process leans on test scores and class rank to predict who will lead well later, this meta-analysis is a direct challenge to that assumption — those measures are doing real work on academic outcomes and comparatively little on leadership.
2. Take grit’s reputation down a notch, and conscientiousness’s up one. The culturally louder trait (grit) showed up weaker here than the less glamorous one (conscientiousness). Dependability and follow-through, not passion and perseverance through hardship, carried the leadership signal.
3. Treat character as mechanism, not garnish. Heraclitus’s fragment and this dataset agree on the same reframe: habitual disposition isn’t a nice-to-have sitting on top of “real” ability. In the domain that actually requires sustained conduct under pressure, it’s closer to the whole story.

Heraclitus never ran a meta-analysis and never met a cadet. He had Ephesus, a handful of cryptic fragments, and a conviction that who you reliably are matters more than what fortune or native talent hands you. Two and a half thousand years later, a University of Minnesota researcher pulled 29,102 records spanning seventy-three graduating classes and found the same thing, with a correction for measurement unreliability attached. The test scores still mattered — in the classroom, where they were built to matter. Everywhere character actually got tested, character won.

---

This post discusses a 2026 meta-analysis published in Military Psychology and Heraclitus’s Fragment 119. The study’s findings come from West Point cadets, an already-selected military population evaluated on that institution’s definitions of academic, leadership, and physical performance across 1949–2022 — they describe correlational relationships within that population, not a causal account of leadership in general, and not a claim about civilian workplaces, other leadership contexts, or whether conscientiousness itself can be deliberately trained.