Galpin's 10-category evidence framework scores every claim 0–100 across meta-analyses, chronic studies, interventions, acute responses, mechanisms, teleology, natural reasoning, anecdote, expert consensus, and intuition — and the strength of your conviction must match that composite score, not just one flattering slice.
2
Acute and molecular data are the most manipulated tiers: a 50% increase in a biomarker (e.g., mTOR, CRP) almost never translates linearly to a 50% improvement in a clinically meaningful outcome — this single insight deflates most supplement and longevity-drug hype.
3
Identify where any speaker sits on the guru-to-scientist spectrum before weighing their claims: both extremes fail in predictable ways, and the most dangerous commentators are those who hide their position on that spectrum.
4
Galpin's strongest practical filter: if a claim fails teleological, anecdotal, natural, and expert tiers while scoring only on molecular mechanism, treat the conviction level as approximately 15–20 out of 100 — not action-worthy until the clinical tiers catch up.
Protocols
Concrete recipes — what, when, how much, and why
8 items
Run a 10-category evidence radar plot before forming a strong conviction on any health claim
WhatWhen evaluating any claim — supplement, training method, diet, medical intervention — mentally score it 1–10 across all ten evidence tiers: meta-analysis/systematic review, chronic/epidemiological data, intervention studies, acute studies, mechanistic/molecular data, teleological sense, natural sense, anecdote, expert consensus, and personal intuition. Sum the composite; calibrate conviction to that composite.
WhenWhenever a new claim comes up in a podcast, book, social media post, or conversation that you are considering acting on — especially before spending money or changing a long-standing practice.
Dose5–10 minutes of deliberate mental review. Galpin does this in real time without a spreadsheet; the habit takes months of practice to automate.
For whomAny educated consumer of health, fitness, or nutrition content who wants to stop being fooled by sophisticated-sounding arguments that rely heavily on one category.
WhySingle-tier evidence can be faked or overstated in every category. Multi-tier assessment forces breadth and makes cherry-picking immediately visible.
CaveatsNot all ten categories are warranted or possible for every question. For a lifestyle question with no intervention studies, low meta-analysis scores don't condemn the claim — absence of data is not data against. Weight categories relative to the type of question.
Galpin's worked examples show how the system produces actionable conclusions: a high mechanism + high acute score with low chronic and intervention scores = 'probably over-interpreted, don't act yet.' A high anecdote + high expert + high teleology but no science = 'might be the parachute case — you don't need science.' A strong score across 7–8 of the 10 = 'I'm probably gonna try it.' The system explicitly accommodates non-scientific evidence without treating it as inferior by default.
Mechanism
Forces the evaluator to locate missing evidence consciously rather than confusing 'no study on X' with 'X does not work' or 'X is proven.'
Every single thing will have necessarily score 100 — 0 to 100 — that means higher scores better — so there are plenty of examples that I will show you where 20 out of 100 is all I need to be 100% confident in it.
Locate your own position on the guru-scientist spectrum before consuming or producing health content
WhatIdentify honestly whether you personally lean toward early adoption (willing to act on weak evidence, high error tolerance) or toward scientific conservatism (requires strong evidence, slow to update). Locate the speaker or author you are evaluating on the same spectrum. Use that positioning to calibrate how to read their output — not to dismiss them.
WhenBefore following any new influencer, reading any new book, or forming a practice from a single podcast episode.
For whomEveryone consuming health content online; especially useful for people who have been repeatedly burned by trend-chasing or, conversely, people who have been too slow to adopt well-evidenced practices.
WhyA guru who says 'I don't care about being wrong' is not lying — that is their operating model. If you don't know this, you mistake their confidence for evidence strength. A scientist who says 'there's no good evidence' may be protecting prestige, not describing reality.
Galpin's key diagnostic behaviors: guru-end communicators don't return and retract, they just move on to the next thing. Scientist-end communicators hold their position long after contradicting evidence accumulates. Both are predictable failure modes. The practical fix is not to avoid either end of the spectrum — 'without the red end we would be in big big trouble, they find things scientists would never find' — but to adjust your prior when you know where on the spectrum the speaker sits.
Once you understand where they lie on this continuum then you're a lot less likely to be tripped — they could be genuinely trying to trick you or they could not — you just don't understand where they are in terms of how they value evidence.
Apply the falsifiability test to any claim that seems to have an answer for every challenge
WhatWhen a speaker answers every counterpoint with a new caveat that conveniently protects their central claim from testing, apply the 'dragon in the garage' test: if no conceivable piece of evidence could change their position, the claim is non-falsifiable and should be discounted regardless of how sophisticated it sounds.
WhenWhen evaluating any long-form argument where each challenge is met with 'yes but in this case it's different because...' — especially in supplement, longevity-drug, and dietary ideology spaces.
For whomAnyone navigating high-stakes health decisions where the evidence base is thin and advocates are motivated by financial or identity interests.
WhyNon-falsifiable positions cannot be true or false by definition. When practitioners defend such positions as science, they are exploiting the authority of science while violating its core requirement.
CaveatsSome claims are genuinely hard to falsify for practical reasons (e.g., very long study timelines), not because they are unfalsifiable in principle. Distinguish between 'not yet falsified' and 'structurally unfalsifiable.'
Galpin borrows the dragon-in-the-garage thought experiment from Carl Sagan's Demon Haunted World: a claimed dragon that is invisible, floats (so leaves no footprints), emits no heat (blocks infrared), and sheds paint is indistinguishable from no dragon at all. The health equivalent: 'This supplement works in all populations, but in your case the dose was wrong, and in the study the population was wrong, and in the meta-analysis the baseline was wrong...' At some point the claim has acquired so many protective qualifications that the evidence base itself becomes unfalsifiable.
What's the difference between that and actually there being no dragon at all — pretty tough to say — and so your evidence has to be falsifiable at some point.
Apply the Dave Beck 8-point quality screen to any piece of new health evidence
WhatRun every claim through eight rapid questions: (1) Is it from multiple independent sources or one lab/author? (2) Is the speaker an expert (deep domain experience) or merely an authority (credential + popularity)? (3) Can the claim be measured/tested? (4) Is the simplest explanation being favored (Occam's razor)? (5) Are there other plausible explanations that haven't been ruled out? (6) Is the entire argument true, or only mostly true? (7) Is the claim falsifiable in principle? (8) Is the speaker protecting an ego-driven position?
WhenWhen evaluating any new-to-you health claim, especially from books, documentaries, or social-media influencers with high follower counts.
Dose2–3 minutes per claim. Can be done while listening.
For whomGalpin adapted this from an eighth-grade science curriculum — it is designed to be accessible to non-scientists. Directly applicable to any adult consuming health content.
WhyEach of the eight points targets a distinct manipulation technique: authority fallacy, cherry-picked single source, unfalsifiability, ego protection, etc. Together they cover the most common ways honest-seeming people mislead audiences.
Galpin describes his father-in-law Dave Beck, a middle-school science teacher who built his entire curriculum around teaching students to evaluate evidence quality. The eight points are the curriculum's output. Galpin calls them 'the BS detector' for practical day-to-day use, as opposed to the full 10-category spider plot which requires more domain knowledge. Together the two systems form a two-tier toolkit: the 8-point screen for quick filtering, the spider plot for deeper analysis.
He teaches eighth graders — you're adults — but nonetheless he teaches them eight points and so when you're trying to identify the quality of evidence you have to look: number one is it coming from multiple sources.
Also said
“It can't be mostly true — it has to be totally true or it's not true. We've already talked about is it generally falsifiable and then ego — are people protecting and sticking to their explanations because of the bears — are they willing to change when the evidence changes.”— The final two criteria that are most commonly omitted in informal analysis: structural falsifiability and ego protection.
Avoid documentaries and books by people with a vested commercial or identity interest in the conclusion
WhatTreat documentaries about health/nutrition/fitness topics as the lowest-quality information source by default. For books, cross-reference at Red Pen Reviews before investing time. For any author, identify whether their livelihood or social identity is tied to the conclusion they are advocating.
WhenBefore starting any new health book or documentary; when evaluating any new influencer whose content you plan to follow regularly.
DoseOne 5-minute check at Red Pen Reviews per book. For influencers, check whether their social media bio/handle encodes an ideological commitment (e.g., 'keto_dave' or 'conjugate_strength').
For whomAnyone who has been misled by a health documentary or a book that turned out to be ideologically motivated.
WhyDocumentaries are editorially crafted for persuasion, not accuracy. Authors who sell supplements, coaching, or branded diets have an undisclosed conflict of interest as real as a pharma-funded trial. Those whose identities are encoded in their handles have made updating their position personally costly.
CaveatsConflict of interest doesn't mean wrong — it means the reader bears extra burden to verify. Galpin explicitly applies this equally to scientists who protect research agendas and to commercial influencers.
Galpin's hierarchy of worst information sources: (1) documentaries — 'I haven't watched one probably since the mid-90s, they're terrible, they're the worst place ever to get information from'; (2) books by people with companies or strong identity investment in the conclusion; (3) news of any kind — 'just gone, leave it.' He is specifically not saying avoid anecdote-driven or practitioner-derived content — those have legitimate evidential weight — but sources whose business model requires a particular conclusion to be true.
I don't watch documentaries about anything sports science related and I haven't for decades — they're terrible they're awful they're the worst place ever to get information from.
Differentiate expert lane before weighting any expert opinion
WhatBefore applying an expert's opinion as evidence, verify that their expertise domain matches the specific claim. An MD cardiologist recommending a training protocol is out of lane. A legendary speed coach recommending a nutrition protocol is out of lane. Domain-specific expertise does not transfer.
WhenAny time a highly credentialed or experienced person makes a recommendation outside their documented domain.
For whomAnyone who relies on podcasts or social media where hosts routinely invite experts to comment beyond their domains.
WhyThe appeal to authority fallacy is most dangerous when the authority figure is genuinely expert in something adjacent to the claim — enough expertise to sound credible, not enough to be reliable.
Galpin uses Jimmy Radcliffe (legendary sprint/power coach at University of Oregon, ~25 years) as his positive example: 'I will trust his depth of knowledge and experience in those realms' — but explicitly says he doesn't care what Radcliffe says about nutrition. He uses 'TV doctors' as the negative example: a cardiovascular surgeon giving nutrition advice is 'completely out of your lane — your MD has no relevance at all to what you're trying to sell me on.' His specific extension: people who list their identity as a credential on social media often use it to talk beyond that credential.
I'd say would be another one so there was Hindus and shamans and things like that that were doing this breathwork for a long time — I mean it's been years meditating for example — but I will trust his depth of knowledge and experience in those realms. Downside: how the hell do you know who's qualifying to be an expert?
Also said
“TV doctors are the worst about this like your cardiovascular surgeon why are you telling me anything about nutrition — you're completely out of your lane.”— The practical negative example most listeners will recognize immediately.
Treat aging / longevity molecular research with a specific skepticism discount
WhatSystematically apply a higher skepticism discount to all longevity claims derived from molecular, cellular, or rodent-model research. Ask explicitly: was this effect replicated in humans? At what sample size? Over what time horizon? Did the effect scale linearly, or did it attenuate dramatically in larger species?
WhenAny time a longevity claim cites earthworm, fruit fly, cell-culture, or rodent data as its primary evidence.
For whomEducated longevity enthusiasts who follow the research and are at risk of acting on mechanism-tier data before the clinical tiers catch up.
WhyMetabolism does not scale linearly across species. Effects seen in fasting, caloric restriction, and senolytics in small animal models have systematically failed to reproduce at the same magnitude in human clinical trials. The mechanism tier is almost always the only strong tier for longevity claims.
CaveatsThis is not a blanket dismissal of animal model research — it is a specific warning about extrapolating longevity magnitude claims from small organisms to humans. The direction of effect (positive/negative) may be correct even when the magnitude is completely different.
Galpin's explicit statement: 'Basically all of aging research for the most part is based on this molecular stuff and a huge percentage of these are either extremely exaggerated or not clinically relevant so it wouldn't actually make a functional difference in humans. We have species compatibility issues — fasting is another classic example — when you looked at it in the cell culture stuff on earthworms it looked amazing and fine in fruit flies but when it was reproduced in humans it showed it didn't do anything.' He quantifies the gap: 'They lived 40% longer — okay sure — what happened when you did it with humans? Yeah they lived eight days longer on average. Great.'
Mechanism
Metabolism scales roughly to body mass to the 3/4 power (Kleiber's law). Interventions affecting cellular energy status have proportionally smaller effects in larger animals with lower mass-specific metabolic rates.
Basically anything under the aging umbrella — they take studies from earthworms or at best rodent models — hey we fasted them or reduced their calories by 13% they lived 40% longer — okay sure — what happened when you did it with humans? Yeah they lived eight days longer on average.
Use the epidemiology canary-finder function correctly — then stop
WhatTreat epidemiological / chronic database findings as pattern detectors ('canaries in the coal mine') that generate hypotheses for deeper research — not as conclusions. When you see a correlation in epidemiological data, the correct next step is to design or seek an intervention study, not to prescribe behavior change.
WhenAny time a health recommendation is being derived primarily from observational/epidemiological data without a corresponding intervention study.
DoseApply as a single gating question: 'Is there also intervention data, or only observational correlation?'
For whomAnyone who relies on media coverage of population-level health studies to make personal health decisions.
WhyEpidemiology almost never establishes causation, but it is systematically misused to establish causation in health journalism and popular books. The HPV case history (epidemiology → intervention → causation confirmed) is the correct use pattern; 'meat causes cancer' (epidemiology → causation assumed) is the incorrect use pattern.
CaveatsSome questions can never be addressed with intervention studies (e.g., 50-year lifestyle effects). In those cases, epidemiology plus strong teleological and mechanistic support may be the best available evidence — but the confidence level should reflect that limitation explicitly.
Galpin's full articulation of the canary function: 'How epidemiology specifically is supposed to work is they're supposed to find Canaries — so good examples of this are things like HPV — we figured out the cure for HPV because of Epidemiology — then we dive deeper — that is how epidemiology is supposed to work. It is not the answer itself.' He pairs this with the VO2max/mortality example: strong epidemiological signal → justified investment in intervention research → now a cornerstone of longevity practice.
Epidemiology is supposed to find Canaries and so good examples of this are things like HPV — we found figured out the cure for HPV because of Epidemiology — then we dive deeper — that is how epidemiology is supposed to work. It is not the answer itself.
What's new
Personal practice updates, fresh positions, predictions
8 items
The 10-category evidence spider-plot — a practical scoring system for everyday claim evaluation
~12 min
Galpin presents a structured 10-axis radar plot where each evidence category (meta-analysis, chronic, intervention, acute, mechanism, teleology, natural, anecdote, expert, intuition) receives a 1–10 score. The overall 0–100 composite determines confidence level — not just whether one tier looks strong.
Why this matters: Most people unconsciously weight one tier heavily (usually the one they heard on a podcast). The radar plot forces breadth of evidence assessment, making cherry-picking visible at a glance.
Background
Galpin built this system after observing that every category of evidence can be faked or misrepresented in isolation. Multi-category assessment makes red herrings stand out.
Galpin explains the system operates as a mental process, not a literal spreadsheet: 'I don't literally do this — I don't have a spreadsheet and I don't have scores — this is well mental.' The value is training the habit of scanning across all ten categories before arriving at a confidence level. Claims scoring 90+ stop warranting new counter-evidence review. Claims scoring below 10 don't merit energy. The vast majority of contested fitness/nutrition questions live in the 30–70 band, and this is exactly where the radar plot earns its keep.
I tend to view all the evidence and I'm using on quotations and I put into ten categories each category gets a score of one to ten based on the strength of the evidence so 10 times 10 gives you 100 so if something scores a 95 out of 100 it's very close to green.
Also said
“All I really ask is not that you don't do things that are can out of a hundred — all I ask is that the strength of your conviction equals the strength of the evidence.”— States the core ethical principle of the whole framework: calibrated confidence, not certainty.
Acute-response evidence is the most easily weaponized tier in health and fitness communication
~28 min
Acute studies (zero to a few weeks, including animal and cell-culture work) can reliably show cause-and-effect in the moment, but the correlation between acute response and chronic adaptation 'can be really bad.' Galpin calls this class of evidence the easiest to manipulate against a lay audience.
Why this matters: Supplement companies and longevity-drug promoters routinely present acute molecular data as though it guarantees long-term outcomes — Galpin names this specifically as a category-level trap, distinct from fraud.
Background
The mTOR–hypertrophy example is canonical: greater mTOR activation in one training protocol does not produce proportionally greater muscle growth at 8 weeks — the relationship is real but non-linear.
Galpin walks through multiple real examples: a higher protein-synthesis rate post-workout does not guarantee more total muscle; a fasted versus fed acute performance difference does not mean chronic fasted training produces better endurance athletes (because adaptation to fasting occurs over weeks). The inflammation argument is even more direct: 'Just because some food causes more inflammation doesn't mean it's worse for you — it may cause a manic response and you may actually reduce your inflammation because of that.' He frames this tier as a positive lead generator — useful for identifying what to investigate — but a deeply unreliable answer generator on its own.
The acute response doesn't necessarily guarantee you the chronic outcome and this is really rampant in the protein synthesis research or the muscle signaling stuff.
Also said
“You took caffeine and ran faster right now — eight weeks later will you actually have better endurance when you go to race in your competition? We assume that a lot and a lot of the times it is true but it's not necessarily true.”— Concrete example of the acute-to-chronic gap that even credible-seeming acute data cannot close.
Teleological reasoning as a long-game error-correction tool
~38 min
Teleological evidence — does this make evolutionary/functional sense? — tends to 'win the long game': ideas that look scientifically supported but fail a basic teleological check have a history of being reversed by later science. Galpin uses fasting and thick-soled shoes as personal examples.
Why this matters: Gives non-scientists a heuristic to pre-filter before evidence matures, without needing to read papers.
Background
Teleological reasoning is formally unreliable (evolution is not goal-directed), but as a rough screen it's historically useful because it catches extreme overclaims early.
Galpin's examples are concrete: 'Carbs are bad for your brain — would it make any logical sense for a carbohydrate to be bad for humans to consume? We saw the same thing when we said humans didn't evolve to sit all day, therefore sitting is bad — okay, that makes sense — but where they went wrong was therefore you should stand all day.' The misfire is the leap from 'sitting is bad' (plausible) to 'standing is optimal' (not plausible). Similarly, fasting scored high on molecular earthworm data but was teleologically questionable at extreme protocols, and the human clinical data later confirmed moderate usefulness with no spectacular longevity effect. The caveat is that teleology can also justify anything via post-hoc evolutionary storytelling, which Galpin acknowledges explicitly.
Ideas that pop up — you're like okay it looks like some science is strong here but this doesn't make any sense teleologically — give it a few years and the science will probably correct and be like actually this doesn't work at all.
Metformin as a case study in dangerous molecular-tier overconfidence
~50 min
Galpin describes the mid-2010s wave of preventive metformin advocacy in the longevity space as a textbook example of mechanism-tier evidence being extrapolated into clinical practice without chronic, intervention, or expert-consensus support. Clinical trials later showed metformin was impairing muscle mass.
Why this matters: High-profile and directly relevant to the longevity audience: if you cannot name specific failures of molecular overconfidence in the health space, you cannot teach the lesson. Metformin is the most cited concrete failure.
Background
The mechanistic case for metformin rested on AMPK activation and caloric-restriction mimicry in rodents. The leap to preventive use in healthy adults ignored the absence of chronic human data.
Galpin's framing: 'There was a lot of people who were like wow let's start taking metformin in our 40s to protect us against aging because of overexaggerated or just misinterpreted sorry molecular stuff — despite no evidence anywhere else in the platform — and in spite of the fact that the logic behind it was really poor people pushed it.' His point is not that metformin is useless, but that the advocacy preceded the evidence by years, and when clinical trial data came back showing muscle mass impairment, those advocates should have publicly retracted. The failure to do so is the second-order problem: protecting prestige over updating belief.
Despite no evidence anywhere else in the platform and in spite the fact that the logic behind it was really poor people pushed it — until the clinical trial started coming out showing this is crippling people's muscle mass.
Also said
“Those people should have been like look I'm doing this there's not good evidence for it across the board but this is what I'm doing — and then when that stuff came out they should have been retracting and apologizing.”— Adds the accountability norm: the failure is both epistemic and ethical.
The guru-to-scientist spectrum as a prerequisite for evaluating sources
~5 min
Before applying any evidence-quality framework, Galpin argues you must locate both yourself and your source on a spectrum from 'guru' (early adopter, innovative, high error rate, rarely updates) to 'scientist' (conservative, slower, more often right, protective of prestige). Understanding a speaker's position on this spectrum explains most apparent contradictions without assuming bad faith.
Why this matters: Reframes 'is this person wrong?' as 'where are they on the spectrum and am I calibrating to that?' — a much more useful and less adversarial starting position.
Background
Galpin places himself near the scientist end but not fully blue: 'I personally believe most of us should be more in this purple end — have some blue folks have some red folks.' He uses the example that national health guidelines should be conservative by design.
Key diagnostic for the guru end: they don't come back and update after being wrong — 'they just do next next next.' Key diagnostic for the scientist end: 'they tend to be very stubborn and they tend to be very protective of their prestige.' Both can mislead, but in predictable and different ways. Galpin applies this to himself explicitly: his prior negative view on steady-state cardio was a spidey-sense failure that he has since updated. The framework does not require the source to be honest — it only requires the listener to correctly classify them.
Once you understand where they lie on this continuum then you're a lot less likely to be tripped right — because they could be genuinely trying to trick you or they could not — you just don't understand where they are in terms of how they value evidence.
N=1 self-experimentation as a legitimate evidence tier — with specific safeguards
~44 min
Galpin defends personal self-experimentation (anecdote category) as 'very clinically relevant, individualized, and scalable' while identifying its two primary failure modes: placebo and the 'true for me equals true for all' fallacy. He frames his book Unplugged as a protocol for running N=1 experiments properly.
Why this matters: Balances the scientist's aversion to anecdote with a practical recognition that meta-analyses cannot personalize prescriptions — both extremes are wrong.
Galpin names the classic self-experiment failure mode as spontaneous remission: 'A classic example of instead of taking my cancer drug treatment I ate only juiced celery and then my cancer ameliorated — well yeah maybe the juice celery, or we also have these things in biology called spontaneous remission where things go away and you don't know why.' He also criticizes scientist-side anecdote-aversion: 'We should promote people doing their own self experiments — start with science, but you've got to self-experiment from there.' The practical advice: use science to find the starting point, then verify on yourself, staying aware of the placebo problem.
The benefits of anecdote right — it's very clinically relevant but by definition almost — and it's also very scalable and individualized.
Risk versus hazard conflation as a specific manipulation technique in chronic/epidemiological data
~22 min
Galpin flags the risk-versus-hazard distinction as a concrete literacy gap that content creators exploit when presenting epidemiological findings. A '50% reduction in mortality risk' from a baseline of 1% to 1.5% is not clinically significant — but the headline never says that.
Why this matters: Gives the audience a single, specific question to ask of any epidemiological claim — what was the absolute baseline? — that defuses most headline-level health journalism.
Galpin's example: 'If your risk was 1% and it took you from 1% to 1.5%, that's a 50% improvement but I could say this improved lifespan by — reduced risk by 50%. Not really — it took me from one to one and a half percent, that's not clinically significant.' He pairs this with the single-factor causality problem in epidemiology: teasing out one factor from a lifetime of co-variables almost always overstates its independent effect, which is why 'meat is bad for your health' fails so comprehensively on the full spider plot.
Does it reduce your risk of mortality by thirty-five percent or your instances of mortality? That's very very different.
Keto as a lived example of updating beliefs when the evidence base evolves
~55 min
Galpin documents his own position change on ketogenic diets: initially skeptical, now seeing it as reasonable for sedentary non-athletes as the meta-analyses, chronic, and intervention tiers filled in. He uses this as the main example of intellectual honesty operating correctly.
Why this matters: Most fitness communicators either defend their original keto position or attack it — Galpin models the nuanced outcome: 'I was too negative early, the evidence moved, my position moved, and specific populations still should not use it.'
He contrasts this with the continuing failure on longevity claims for keto (still no strong evidence) versus body composition claims (now reasonable). The broader lesson: 'The keto folks got a lot of things wrong but they did have some things right that we missed on.' Being dismissive toward the guru end of the spectrum caused legitimate insights to be overlooked for too long.
I shouldn't have been so negative to keto when it first came out but I have changed my tune because more evidence has come along.
Recommendations
Products, supplements, and tools mentioned in the episode
4 items
Red Pen Reviews (redpenreviews.org) — accuracy audit site for health/science books
Tool
Galpin recommends this website as a pre-filter for health books before investing reading time: look up the book, see how its factual claims rate, then decide whether to read it.
Mentioned as a counterweight to the recommendation to be wary of books from influencers and MD-authors out of their lane. Galpin frames it as a quick shortcut for non-scientists who cannot independently evaluate the quality of citations: 'You can start there if you're interested in a book — check it out at Red Pen Reviews — and then you can come back and decide to read the book or not.' Particularly valuable for popular longevity and nutrition books where claims are often overstated.
One thing I won't recommend is a website called red pen reviews — it's really cool to go to to figure out kind of the quality of accuracy of books so you can start there if you're interested in a book.
Source of the 'dragon in the garage' thought experiment that Galpin uses to teach the falsifiability criterion — the core epistemological concept underlying the entire BS detector framework.
Galpin credits Sagan directly: 'I borrowed this from Carl Sagan in his famous book Demon Haunted World.' The dragon-in-the-garage story is Sagan's illustration of why unfalsifiable claims are epistemologically equivalent to false claims. Galpin uses it as the capstone of the lecture because it closes the loop on what makes an evidence framework useful at all: without falsifiability, no score on any tier can ever move downward, and the spider plot becomes meaningless.
I borrowed this from Carl Sagan in his famous book Demon Haunted World — and Carl's dragon in the garage goes like this — he says hey look I want you to come to my garage I want you to see my dragon.
PubMed / Google Scholar / ResearchGate — primary literature access
Tool
Galpin recommends these as the correct sources for meta-analyses, systematic reviews, and intervention studies — with the specific note that ResearchGate often provides free access to paywalled articles.
Practical tip embedded in the meta-analysis section: 'If you're ever on Google Scholar or PubMed and you find an article behind a paywall you can also try ResearchGate — a lot of times they're up there free.' He also recommends subscription-based evidence-synthesis services (two to five PhDs reviewing the literature for $5–15/month) as an intermediate-accessibility option. For nutrition specifically he names examine.com as a well-curated free resource.
Good sources to find them — PubMed is fantastic — Google Scholar can get them — if you ever find a paywall you can also try the third option I have there which is ResearchGate — a lot of times they're up there free.
ISSN and NSCA conferences — primary practitioner-grade evidence translation
Service
Galpin specifically recommends ISSN (International Society of Sports Nutrition) and NSCA (National Strength and Conditioning Association) as the best conference and position-stand sources for sports science and nutrition intervention evidence.
His framing: 'I would strongly encourage if you're in this area of sports science to head to ISSN or NFSA — maybe ACSM but not really — or a similar event and go to these national conferences or local clinics or regional ones and you can learn a lot of information about this level of science.' He contrasts ISSN position stands (high quality, reviewing full literature) against generic medical conferences where sports nutrition is treated as peripheral. He also recommends examine.com for the same tier of synthesized nutrition evidence without a subscription cost.
I would strongly encourage if you're in this area of sports science to head to ISSN or NFSA maybe ACSM but not really or a similar event and go to these national conferences.
Galpin's own book on how to balance technology, science, and personal self-experimentation to figure out what actually works for your individual biology — mentioned as the practical implementation of the anecdote/self-experiment tier.
DisclosureGalpin is a co-author and references the book multiple times as the practical guide to N=1 self-experimentation.
Referenced twice in the episode: once in the natural reasoning section ('I actually put this in my book unplugged — hand sanitizer') and once in the anecdote section ('this is basically what my entire book unplugged is all about — is how to do your own self experiment and how to balance the science and the technology with what your body is actually telling you'). The book's premise is the operationalization of the anecdote tier: start with science as a foundation, but verify on yourself.
This is basically what my entire book unplugged is all about — is how to do your own self experiment and how to balance the science and the technology with what your body is actually telling you to figure out what's right for you.
Lines worth pulling out — contrarian, specific, or perfectly phrased
6 items
The strength of your conviction must equal the strength of the evidence — so if you're like hey man I totally believe the earth is flat — oh really where's your evidence — damn it there is no evidence but I just really feel it — I don't care.
The one-sentence rule that is the entire framework's moral foundation. Calibrated confidence, not certainty, is the goal.
Basically anything under the aging umbrella — they take studies from earthworms or at best rodent models — they lived 40% longer — okay sure — what happened when you did it with humans? Yeah they lived eight days longer on average.
The most targeted single statement in the episode: directly indicts the majority of longevity supplement and drug advocacy that Galpin's audience is exposed to.
Just because one thing has a higher rate of protein synthesis doesn't mean eight weeks later you have the exact same amount of total muscle growth — they're not unrelated but they are not — it's not linear.
Destroys one of the most prevalent supplement marketing arguments (the mTOR/protein synthesis cascade) in a single clean sentence.
The plural of anecdote does not equal data — you saw a website and they have 50 pre-post pictures — it's not as good as science — they probably had 5,000 pre-post pictures that sucked but they're not telling you.
Names the specific anecdote manipulation most commonly used in fitness marketing — and quantifies exactly why 50 success stories is not evidence.
I don't watch documentaries about anything sports science related and I haven't for decades — they're terrible they're awful they're the worst place ever to get information from.
Surprising and memorable coming from a science communicator; licenses audiences to stop treating health documentaries as credible.
It was metformin — there was a lot of people who were like wow let's start taking metformin in our 40s to protect us against aging — and in spite the fact that the logic behind it was really poor people pushed it — until the clinical trial started coming out showing this is crippling people's muscle mass.
The most concrete real-world harm example in the episode — turns an abstract epistemological lesson into a specific cautionary case study.
Sign in to share feedback
Tell us if this brief hit the mark or missed it — feedback feeds back into the next iteration of the prompt.
Reading is free for everyone. A free account adds the personal layer: save protocols, follow experts, and see how the other experts weigh in on this same topic.
Educational summary of the cited expert source — not medical advice. Open the source recording linked above and consult a qualified physician before acting on any protocol.
