BRAINBITES
WEEK OF JUN 19–25, 2026 · ISSUE 17
BRAINBITES — Your weekly health, fitness, and longevity digest. Five minutes that cover the highest-signal podcasts and science before they hit the mainstream.
This week: relationship researcher Paul Eastwick walked Huberman through 15 years of data on what predicts romantic attraction — spoiler: stated preferences predict almost nothing. And nutrition scientist Layne Norton returned in Essentials with the most evidence-based breakdown of dieting misconceptions on the platform.
From: Huberman Lab · Huberman Lab Essentials
🔢 The correlation between stated mate preferences and actual partner choice
🎙️ Paul Eastwick on attraction, stated vs. revealed preferences, and what speed-dating data shows
💡 Deep Dive: the mismatch between what you say you want and what you actually choose
📰 News Bits: the ideal-partner paradox · metabolic adaptation and dieting · the protein leverage hypothesis
GM.
The most surprising data set this week came from Paul Eastwick on Huberman. Eastwick has spent 15 years studying mate selection with the tools of experimental psychology, and his core finding is uncomfortable: the traits people list as important in a partner (height, income, sense of humour) have near-zero correlation with who they're actually attracted to when they meet someone in person.
⏱ ~5 min read
🧠 THE MATCHING ENGINE
What the science of mate selection actually says — and what dating apps get wrong.
Two bodies of research collided this week. One: your stated preferences in a romantic partner — the checklist in your head — predict your actual attraction to real people with a correlation near zero. Two: your stated dietary goals — the plan you set on Monday — predict your actual eating behaviour about as well. Both domains run on revealed preferences that bypass conscious intention.
The thread this week: you think you know what you want. The data says you don't — in love, in food, and probably in everything else.
🔢 THE VITAL SIGN · PSYCHOLOGY / RELATIONSHIPS
r ≈ 0
The correlation between stated ideal-partner traits and actual attraction in speed-dating and longitudinal studies.
Eastwick's speed-dating studies are his most striking. Before the event, participants rate how important various traits are to them (physical attractiveness, earning potential, sense of humour, kindness, ambition). After the event, researchers measure who participants were actually attracted to. The correlation between stated preferences and actual attraction is near zero — statistically, knowing someone's stated preferences gives you almost no ability to predict who they'll like.
The finding holds across genders, age groups, and cultures. Men who say they prioritise physical attractiveness don't consistently choose the most attractive partners. Women who say they prioritise earning potential don't consistently choose the highest earners. The stated preferences are real beliefs — people aren't lying. They're just not predictive. The mechanism that generates attraction in a live interaction operates below the level that stated preferences capture.
Eastwick's explanation: stated preferences reflect abstract ideals generated in the absence of a specific person. Actual attraction is an emergent, holistic response to a specific person — driven by interaction dynamics (conversational flow, humour synchrony, nonverbal cues) that can't be decomposed into trait lists. You can't know what will attract you until you're in the interaction.
The checklist in your head is a fiction. Attraction is an emergent computation that runs on data your conscious mind doesn't have access to.
Dr. Paul Eastwick on Huberman Lab · June 2026
🎙️ THE EPISODE · PSYCHOLOGY / RELATIONSHIPS
Your dating checklist predicts nothing — the speed-dating data is clear
Paul Eastwick on 15 years of mate-selection research and why stated preferences fail to predict actual attraction.
Dr. Paul Eastwick · Huberman Lab · June 2026
Eastwick is a relationship researcher at UC Davis who publishes in the top journals in social and personality psychology. The conversation with Huberman is evidence-rich, counterintuitive, and directly applicable to anyone who's ever wondered why they keep falling for people who don't match their 'type.'
Dating apps optimise for stated preferences — which don't predict attraction.
Eastwick's most practical implication: dating apps are built on filter logic — swipe based on photos (physical attractiveness), age, height, occupation. These are stated-preference variables. The data says stated preferences don't predict attraction. Apps that prioritise interaction quality over filter matching would, in theory, produce better outcomes. But interaction quality is harder to monetise than profile scrolling.
🎙️ Skip to ~28:00 for the dating-app critique and what would work better.
Interaction dynamics drive attraction — not trait checklists.
Eastwick's research identifies the variables that do predict attraction in speed-dating: conversational responsiveness, perceived partner interest, humour co-creation, and the subjective sense of 'clicking.' None of these are things you can list in a profile or filter for in an app. They emerge in real-time interaction. His recommendation: increase the number of in-person interactions rather than optimising the filter criteria.
Attraction is not a trait match — it's an emergent computation.
Eastwick's theoretical model: attraction isn't the brain checking a list of traits against a stored template. It's an emergent property of the interaction — a holistic response to a specific person in a specific context. The same person can be attractive in one conversation and not in another. This explains why 'type' feels real in the abstract but fails predictively: your 'type' is a post-hoc narrative your brain constructs to explain a process it doesn't consciously control.
🎙️ Listen → Huberman Lab, Dr. Paul Eastwick, June 2026
💡 DEEP DIVE · STATED VS. REVEALED
Eastwick gives you the relationship data. Norton gives you the nutrition parallel. The pattern: what you say you want and what you actually do are different systems.
Layne Norton's Essentials episode, dropped the same week, lands on the same stated-vs-revealed gap from the nutrition angle. People state dietary preferences and goals (eat clean, track macros, avoid sugar). Their actual behaviour deviates within days. Norton's data on dieting shows that stated intentions predict dietary adherence for about two weeks before revealed preferences — the habitual, context-driven eating patterns that run below conscious planning — reassert themselves.
Norton's meta-point mirrors Eastwick's: the conscious, rational plan (the checklist, the diet) is overridden by a deeper system that optimises for different variables. In nutrition, the deeper system optimises for energy density and palatability (the protein leverage hypothesis suggests the body eats until it hits a protein target, regardless of total calories). In attraction, the deeper system optimises for interaction quality and compatibility signals that can't be pre-specified.
The practical implication is the same in both domains: stop optimising the checklist. Start optimising the environment. Norton's dietary recommendation is to design your food environment so the default option is aligned with your goals (don't buy the chips). Eastwick's relationship recommendation is to increase the number of real interactions rather than refining your filter criteria (go on the date instead of optimising the profile).
Both researchers are saying: your conscious preferences are a rough draft. Your revealed behaviour is the final copy. The intervention isn't to fight the gap with willpower. It's to design systems that work with the revealed preferences rather than against them.
The conscious mind thinks it's the author. The data says it's the editor — and the first draft is already written.
📰 NEWS BITS
The ideal-partner paradox: the more specific your checklist, the less predictive it is.
Eastwick's data shows an ironic pattern: people with the most specific stated preferences (must be tall, must have a graduate degree, must love hiking) show the weakest correlation between those preferences and actual attraction. The explanation: a long checklist creates a rigid filter that screens out the interaction-quality signals that actually drive attraction. The most 'open' daters — those with minimal stated preferences — reported the highest satisfaction with their eventual partners.
— Huberman Lab, Dr. Paul Eastwick, June 2026
Metabolic adaptation: your body fights weight loss by reducing energy expenditure.
Norton's Essentials episode covered metabolic adaptation — the well-documented phenomenon where sustained caloric restriction reduces resting metabolic rate beyond what weight loss alone would predict. The body 'defends' its starting weight by becoming more metabolically efficient: moving less unconsciously (NEAT reduction), reducing thyroid hormone output, and increasing mitochondrial efficiency. Norton's practical advice: diet breaks (1–2 weeks at maintenance calories every 8–12 weeks of dieting) partially reverse metabolic adaptation and improve long-term adherence.
— Huberman Lab Essentials, Layne Norton, June 2026
The protein leverage hypothesis: your body eats until it gets enough protein — regardless of calories.
Norton discussed the protein leverage hypothesis, which proposes that the human appetite system prioritises protein intake. In environments with low-protein, high-calorie foods (the modern ultra-processed diet), people over-consume total calories trying to hit a protein target their body is seeking. The intervention: increase the protein density of your diet early in the day and total caloric intake tends to self-regulate downward. Multiple observational and controlled studies support the mechanism.
— Huberman Lab Essentials, Layne Norton, June 2026
P.S. Think about your actual partner (or your last relationship). How many items on your 'checklist' did they actually match? Hit reply and tell us the number — we're betting it's lower than you'd guess.
What did you think of today's BrainBite?
🧠🧠🧠🧠🧠 The stated-vs-revealed framework just reframed how I think about decision-making in general
🧠🧠🧠 The protein leverage hypothesis is the dieting tip I didn't know I needed.
🧠 Less relationship psychology, more biochemistry.
We listen to every health podcast so you don't have to. Weekly. No noise.
The checklist is a fiction. The interaction is the data. Trust the data. See you next week.
