Name: Hypertrophy: The Few Principles That Actually Drive Muscle Growth (And the Surprisingly Short List of Things That Don't)
Creator: Realised
Published: 2026-06-01

Summary

Muscle growth runs on a short list of real drivers — mechanical tension as the primary stimulus, enough hard weekly volume (to a point, with diminishing returns), training effort near failure, and progressive overload — and an almost-as-short list of discredited myths (soreness equals growth, muscle confusion, the narrow post-workout anabolic window, light-versus-heavy mattering once effort is matched); the honest move is to tell users the real levers plainly while refusing to launder the influencer scene's over-precise set-numbers and reps-in-reserve prescriptions into a certainty that small,

Why Strong

Strong Evidence because the entry's load-bearing claim is the core model — mechanical tension as primary driver, weekly volume dose-response to a point, load-equivalence-for-size when effort is matched, the anabolic-window myth, and progressive overload — and those directional claims replicate across multiple independent meta-analyses spanning years. The myth-debunks (soreness, muscle confusion, window) are independently well-supported.

NOT Foundational because the topic carries genuine clinical and commercial judgement and a two-sided controversy (oversold precision on one side, residual bro-science on the other), not a single undisputed axiom.

NOT Moderate for the headline, because the spine is not "a few suggestive studies" — it is multiple independent meta-analyses agreeing on direction. The lower-certainty pieces are specific downstream parameters, which the entry marks rather than inheriting.

The per-driver split (read this, not just the headline):
• Mechanical tension as primary: Strong; the relative weight of metabolic stress and damage is Moderate-to-Emerging.
• Volume dose-response: Strong for direction; the exact set-number is a soft heuristic, method-dependent — explicitly NOT Strong.
• Load equivalence for whole-muscle size: Strong; fibre-type-specific differences Emerging (biopsy data too sparse).
• Proximity to failure: Moderate — directionally supported but small, borderline effect sizes; precision overstated by influencers.
• Anabolic-window myth: Strong debunk.
• Soreness ≠ growth: Strong debunk.
• Muscle confusion: Moderate debunk; targeted regional variation an Emerging nuance.
• Measurement caveat: Strong as a methodological fact — and the reason the precise numbers stay below Strong.

Practical takeaway

The framing to hold: the things that grow muscle are few, boring, and free of novelty. You do not need to chase soreness, switch exercises constantly, or time shakes. Consistent hard-enough work with gradual loading is the whole game.

Do the few things that work.
• Train with enough hard volume. Aim for roughly ten-to-twenty hard sets per muscle per week as a range, not a target — more is generally better up to a point, then returns diminish. Start at the lower end and add volume only if recovery allows.
• Take sets close to failure. Stop a few reps short to near failure on most sets. You do not need to grind every set to absolute failure; the growth difference is small and the recovery cost of constant failure is real.
• Progressively overload. Add reps, then load, over time. The progressive-overload mechanic itself is owned by physical_progressive_overload — route there for the how.
• Pick a load you can take near failure. Heavier builds strength better; for size, anything from moderately heavy to fairly light works as long as the set ends near failure. Use what your joints and equipment allow.

Ignore the things that don't.
• Don't chase soreness. It is not a growth gauge. A productive session can leave you barely sore; a useless novelty session can leave you wrecked.
• Don't "confuse" the muscle with random exercise churn. Pick a small set of movements and get progressively stronger at them. Variation for variety's sake buys motivation, not growth.
• Don't worry about the post-workout window. Hit your total daily protein and the timing is a rounding error. (Dose: diet_protein_intake.)

Hold the precision loosely.
• Treat any exact prescription — "twelve sets, one-to-two reps in reserve" — as a reasonable starting point, not a law. The evidence supports the direction; the decimals come from small, short, surrogate-outcome studies. Adjust to your own recovery and response.
• If you are older, detrained, or a true beginner, do NOT assume the young-trained-male set-numbers transfer directly; start conservative and defer to sarcopenia_prevention and resistance_training_and_body_composition.

Evidence detail

Why This Entry Exists

Hypertrophy is one of the most over-explained topics in fitness, and the explaining happens in two failure modes at once. The old bro-science layer sells novelty and sensation as the engine of growth: chase the pump, chase soreness, keep the muscle "guessing," and slam a shake inside the anabolic window or the workout was wasted. None of that survives contact with the evidence. But the "science-based" influencer scene that debunks all of it has quietly built its own problem: it sells precise prescriptions ("do exactly twelve to fourteen hard sets, stop one-to-two reps from failure") with a confidence the underlying data do not earn. Those data are mostly small samples of young men trained for eight-to-twelve weeks, measured with ultrasound muscle-thickness surrogates that disagree with biopsy and MRI, and produced by a handful of commercially-active researchers.

So this entry is the one-stop for "what actually grows muscle," built to hold both truths at once. It states the real drivers confidently because they replicate across independent meta-analyses, and it states the myths as discredited because the evidence genuinely discredits them. Then it refuses to over-prescribe: ranges, not magic numbers. The load-bearing claim — tension plus enough hard volume plus effort near failure plus progressive overload, and nothing exotic — is Strong Evidence and survives regardless of how the precise set-counting debates eventually resolve.

What bad advice this protects against, in all directions:
• "No pain, no gain — if you're not sore the next day, it didn't work" → soreness tracks unaccustomed and eccentric stress, not stimulus quality; growth happens in protocols that produce little soreness, and damage is not a prerequisite.
• "Keep your muscles guessing — constantly switch exercises so they never adapt" → muscle confusion as a growth driver is unsupported; when volume and effort are equated, varied and fixed selection grow muscle similarly.
• "You must hammer a protein shake in the thirty-minute window or you lose the gains" → the narrow anabolic window is a myth once total daily protein is adequate; this entry defers protein dose itself to diet_protein_intake.
• "Lift light or you won't build size — only heavy weights count" / "only light, high-rep pump work builds size" → both wrong; load barely matters for whole-muscle growth once sets are taken near failure. Heavy still wins for strength.
• "Do exactly fourteen sets at one-to-two reps in reserve, that's the optimised protocol" → the meta-analyses support the direction, not the decimal; that precision is sold with more confidence than n=20, eight-week ultrasound studies can carry.

This entry owns the evidence-based model of muscle growth and the myth-debunking. It does not re-argue general resistance-training or body-composition (resistance_training_and_body_composition), protein intake and dose (diet_protein_intake), DOMS physiology and recovery (muscle_soreness_recovery_doms), the progressive-overload mechanic itself (physical_progressive_overload), or muscle preservation in ageing (sarcopenia_prevention). It states those scope boundaries and defers the specifics.

Evidence

Organised by driver, with the tier signal inline. The headline is Strong, but the precision is not — read the tiers, not just the thesis. Every number below is a direction with a confidence interval, never a prescription.

Mechanical tension — the primary driver (Strong for tension-as-primary; the other two mechanisms are weaker).

1. Mechanical tension is the primary stimulus for hypertrophy; metabolic stress ("the pump") and muscle damage have weak, indirect causal support. The classic framing listed three proposed mechanisms — tension, metabolic stress, and muscle damage. Subsequent mechanistic work has held tension (via mechanotransduction and downstream mTOR signalling) as the genuine primary driver, while downgrading metabolic stress and muscle damage to indirect or minimal contributors; damage is increasingly viewed as a byproduct rather than a necessary stimulus. (Schoenfeld BJ, "The mechanisms of muscle hypertrophy and their application to resistance training," J Strength Cond Res 2010;24(10):2857-2872 — originated the tension/metabolic-stress/damage triad; later mechanistic reviews downgrade the latter two. Strong Evidence for tension-as-primary; the relative contribution of the other two is Moderate-to-Emerging, mechanistic and contested. BOTH-WAYS: Schoenfeld is the field's most prolific hypertrophy researcher AND a commercial fitness author/educator — that publication-and-influence concentration in one figure is itself a bias vector to flag.)

Weekly volume — the dose-response, to a point (Strong for direction; the number is soft).

2. Weekly volume drives growth in a dose-response, with diminishing returns — not unlimited gains. More hard sets per muscle per week produce more growth; the original analysis found fewer than five sets suboptimal and ten-or-more best, and the largest and most recent meta-regression confirms growth keeps rising with volume but with diminishing returns. The working heuristic of roughly ten-to-twenty hard sets per muscle per week is a soft band, not a hard finding. That band is the growth-OPTIMISATION range with diminishing returns, not a floor: the minimum-effective floor — most of the benefit from far fewer sets — is owned by minimum_effective_dose and physical_progressive_overload and deferred there. (Schoenfeld BJ, Ogborn D, Krieger JW, "Dose-response relationship between weekly resistance training volume and increases in muscle mass," J Sports Sci 2017;35(11):1073-1082; updated by Pelland JD et al., "The Resistance Training Dose Response," Sports Med 2025, 67 studies and 2,058 participants — diminishing-returns best fit. Strong Evidence for direction — replicates across independent meta-analyses spanning eight years. NUMBER CAVEAT: Pelland 2025 also showed that how you count a "set" — the fractional-set method — changes the headline figure, so the exact set target is method-dependent. Do NOT present a precise weekly set number as Strong.)

Load — heavy versus light barely matters once effort is matched (Strong for whole-muscle size).

3. Load barely matters for whole-muscle hypertrophy once effort is equated; strength still favours heavy. Low-load (≤60% 1RM) training taken to or near failure produces hypertrophy similar to high-load (>60% 1RM) training. Size is roughly load-independent when effort matches; strength is not — it still favours heavier loads. (Schoenfeld BJ, Grgic J, Ogborn D, Krieger JW, "Strength and Hypertrophy Adaptations Between Low- vs. High-Load Resistance Training: A Systematic Review and Meta-Analysis," J Strength Cond Res 2017;31(12):3508-3523 — trivial hypertrophy difference, strength favoured high load. Strong Evidence for whole-muscle equivalence; fibre-type-specific differences remain Emerging, the biopsy data are too sparse to meta-analyse. CAVEAT the authors themselves flag: the equivalence rests on whole-muscle ultrasound/MRI surrogates that could not be corroborated by biopsy.)

Proximity to failure — matters less than prescribed (Moderate; the signal is genuinely borderline).

4. Training to momentary failure is not clearly superior to stopping short. The meta-analysis found only a trivial advantage for set-failure overall (effect size ≈ 0.19, p = 0.045) and no significant benefit of momentary failure specifically (effect size ≈ 0.12, p = 0.343); a follow-up RCT found similar quad growth training to failure versus stopping one-to-two reps in reserve. "Near failure is enough" is directionally supported, but the effect sizes are small and statistically borderline — the precision influencers attach to it is overstated. (Refalo MC, Helms ER, Trexler ET, Hamilton DL, Fyfe JJ, "Influence of Resistance Training Proximity-to-Failure on Skeletal Muscle Hypertrophy: A Systematic Review with Meta-analysis," Sports Med 2022;52(12):2693-2713, 15 studies; follow-up RCT Refalo et al., J Sports Sci 2024, ~26 trained adults. Moderate — directionally supported, effect sizes small and borderline. DISCLOSED COI, both-ways: co-authors Helms and Trexler disclose earning income as writers and practitioners within the fitness industry; the study received no funding and the disclosure is present — but the people producing the "you don't need to grind to failure" message are embedded in the coaching market that message serves.)

The anabolic window — a myth (Strong debunk).

5. The narrow post-workout "anabolic window" is a myth when total daily protein is adequate. It is specifically the NARROW (~30-minute) post-workout window that is a myth; total daily protein still matters across a wide (~24-hour) window, owned by diet_protein_intake. Timing protein immediately around training has no meaningful independent effect on hypertrophy or strength once total intake is controlled; in the meta-analysis the apparent timing effect disappeared after adjusting for total protein — the timed groups had simply eaten more. (Schoenfeld BJ, Aragon AA, Krieger JW, "The effect of protein timing on muscle strength and hypertrophy: a meta-analysis," J Int Soc Sports Nutr 2013;10:53; narrative companion Aragon AA, Schoenfeld BJ, "Nutrient timing revisited," J Int Soc Sports Nutr 2013;10:5. Strong Evidence for "no narrow window when total protein is adequate." Defers protein-dose specifics to diet_protein_intake. FUNDING NOTE, both-ways: published in the ISSN journal, which has historically had supplement-industry ties — but the finding itself is anti-supplement-marketing, debunking "must slam a shake now," which cuts AGAINST a pro-industry bias and strengthens it.)

Soreness — not a marker of growth (Strong debunk).

6. Muscle soreness (DOMS) is not a valid marker of muscle growth. Soreness correlates poorly with actual muscle-damage markers and with hypertrophy; growth occurs in muscles and protocols that produce little or no soreness, and damage is not a prerequisite for growth. Soreness reflects unaccustomed and eccentric stress, not stimulus quality. (Damas F, Libardi CA, Ugrinowitsch C, "The development of skeletal muscle hypertrophy through resistance training: the role of muscle damage and muscle protein synthesis," Eur J Appl Physiol 2018;118(3):485-500; plus the broader DOMS-versus-damage dissociation literature showing soreness tracks poorly with strength loss, range-of-motion, and creatine-kinase. Strong Evidence — "soreness ≠ growth" is well-supported. No industry incentive distorts this; if anything it removes a common upsell. Strong recovery-framing fit: chasing soreness is exactly the optimisation-trap Realised reframes away from. DOMS physiology and recovery owned by muscle_soreness_recovery_doms.)

Muscle confusion — not a driver (Moderate debunk; targeted variation a separate Emerging nuance).

7. "Muscle confusion" — constantly changing exercises to keep the muscle guessing — is not supported as a growth driver. When volume and intensity are equated, varied versus fixed exercise selection produces similar hypertrophy; the main measured benefit of variation was higher self-reported motivation, not more growth. Some systematic variation may aid regional growth, but random churn offers no advantage. (Baz-Valle E, Schoenfeld BJ, Torres-Unda J, Santos-Concejero J, et al., "The effects of exercise variation in muscle thickness, maximal strength and motivation in resistance trained men," PLoS One 2019;14(12):e0226989; review Kassiano W et al., "Does Varying Resistance Exercises Promote Superior Hypertrophy?," J Strength Cond Res 2022. Moderate — variation-for-confusion debunked; targeted variation for regional growth is an Emerging nuance, not a mandate. Cuts against the commercial program-selling incentive — novelty sells subscriptions, and the null finding is the unglamorous one.)

The field-level caveat — why precision is oversold (Strong as a methodological caveat).

8. Much hypertrophy evidence rests on ultrasound muscle-thickness surrogates measured over only eight-to-twelve weeks in small samples — and the surrogates disagree. Ultrasound thickness, biopsy fibre cross-sectional area, and DXA/MRI can yield contradictory results, so over-precise prescriptions outrun the measurement precision itself. This is the load-bearing reason NOT to hand out magic set-numbers. (Haun CT, Vann CG, Roberts BM, Vigotsky AD, Schoenfeld BJ, Roberts MD, "A Critical Evaluation of the Biological Construct Skeletal Muscle Hypertrophy: Size Matters but So Does the Measurement," Front Physiol 2019;10:247; plus Stokes T et al., ultrasound methodological validation, Physiol Rep 2021;9(1):e14683. Strong Evidence as a methodological caveat — the measurement disagreement is itself well-documented. NOTABLE: several of the SAME researchers who produce the optimistic dose-response findings, including Schoenfeld, co-author the measurement-skeptic papers — credit for intellectual honesty, but it underlines how concentrated the field is in a handful of commercially-active labs.)

Mechanism

Why mechanical tension is the engine. A muscle fibre senses load through mechanotransduction: high mechanical tension across the contractile apparatus activates signalling (centrally, the mTOR pathway) that raises muscle protein synthesis above breakdown, and when that net balance stays positive across repeated sessions, the fibre adds contractile protein and grows. Tension is the input the machinery reads. This is why the amount of hard work matters (volume) and why effort matters (sets taken near failure recruit and fatigue more high-threshold motor units, exposing more fibres to high tension) — both are routes to delivering tension to the fibre, not separate magic levers.

Why the pump and soreness are byproducts, not the cause. Metabolic stress ("the pump") and muscle damage co-occur with growth-producing training, which is why they got mistaken for the cause. But the pump is transient fluid shift and metabolite accumulation, and soreness reflects mechanical and inflammatory disturbance from unaccustomed or eccentric loading. Neither is necessary: growth happens in protocols that produce little pump or soreness once tension and volume are adequate. Correlation with hard training is not the same as causation of growth.

Why load is roughly interchangeable for size but not strength. Take a set near failure and the final reps recruit the high-threshold motor units regardless of whether the load is heavy-and-few or light-and-many — the fibres exposed to high tension end up similar, so whole-muscle hypertrophy ends up similar. Strength is different: it is partly a neural and load-specific skill, so it keeps favouring heavy loads that train the nervous system to express force under that load. Same fibres grown; different skill practised.

Why total protein beats timing. Muscle protein synthesis stays elevated for many hours after a training bout, so the relevant variable is whether enough protein arrives across the day, not whether it arrives in a thirty-minute window. When studies appeared to show a timing benefit, the timed arm had usually just eaten more total protein. Adequate daily intake makes the window a rounding error. The dose itself is owned by diet_protein_intake; this entry only debunks the window.

Risks And Contraindications

• Over-precision is the central hazard. The biggest trap is laundering soft heuristics ("ten-to-twenty sets," "one-to-three reps in reserve") into hard certainty. The meta-analyses support direction — more volume to a point, effort near failure helps a little — not the exact numbers, which are method-dependent and built on short surrogate-outcome trials. State ranges and uncertainty, never decimals dressed as law.
• Population scope. Most data are young (~25y) men trained for eight-to-twelve weeks. Do NOT silently extrapolate set-numbers to older adults, postmenopausal women, or detrained beginners; defer to sarcopenia_prevention and resistance_training_and_body_composition.
• Constant failure-training has a real cost. The marginal growth from grinding every set to absolute failure is small, but the fatigue, recovery debt, and injury exposure are not. "Near failure on most sets" is the safer and roughly equally effective default.
• Lane discipline. This entry must NOT re-argue protein dose (diet_protein_intake), DOMS physiology and recovery (muscle_soreness_recovery_doms), or general body-composition (resistance_training_and_body_composition). It cites the myth-debunks only as they bear on growth, then defers.
• The debunks are not a licence to slack. "Soreness doesn't equal growth" does not mean training can be easy; "the window is a myth" does not mean protein is optional. The myths are wrong about the mechanism, not about the need to train hard and eat enough.

Controversy

Nature: a well-established core model (tension, volume to a point, effort near failure, progressive overload) wrapped in two layers of error — an old bro-science layer that is simply wrong, and a newer "science-based" layer that is directionally right but over-confident. The interesting fight is not bro-science versus science; it is between honest direction and oversold precision, and the precision is sold by a small, commercially-active field.

Position A — "The drivers are real and well-established." The mainstream evidence-based take.
• Best evidence: strong and replicated. Mechanical tension is the primary stimulus; volume drives growth in a dose-response with diminishing returns; load barely matters for whole-muscle size once effort is matched; progressive overload is required. These replicate across multiple independent meta-analyses.
• Where it stays honest: it acknowledges that the exact volume number and the proximity-to-failure effect are softer than the headline, and that the measurement tools disagree.

Position B — "The bro-science is wrong, AND the science-based scene oversells its own precision." The harder, both-ways take.
• Best evidence: the myths are genuinely discredited (soreness, muscle confusion, anabolic window). But the trials underpinning the precise counter-prescriptions are small-n, weeks-long, and lean on ultrasound muscle-thickness surrogates that disagree with biopsy and MRI; many of the most-cited researchers earn income inside the fitness and supplement industry. "Do exactly X sets at Y reps-in-reserve" is sold with more confidence than the data support.
• Where it's wrong if pushed too far: this is not nihilism. The directional claims DO replicate; discounting the magic numbers is not the same as discounting the model.

The funding/bias dimension — cui bono, both ways. The bro-science myths pay the supplement market (anabolic-window shakes), the novelty-program and "muscle confusion" market (ever-changing routines), and "feel the burn" programming. The evidence here debunks those — pointing AWAY from those sales incentives, which strengthens the debunks. But the "science-based" counter-scene has its own cui bono: the field is concentrated in a small number of commercially-active researchers (one figure authors or co-authors a large share of the cited meta-analyses while also working as a paid fitness educator; another paper's co-authors explicitly disclose earning income as fitness-industry writers and practitioners; several papers sit in a journal with historical supplement ties). None of this invalidates the findings — disclosures are present and several findings cut against industry — but it means the AUTHORITY is narrow and the PRECISION is oversold.

Realised Position: Hold both. Tell users the short list of things that actually drive growth — tension, enough hard volume, effort near failure, progressive overload — and name the myths (soreness, muscle confusion, anabolic window) as discredited, plainly. Then refuse to over-prescribe: give ranges, not magic numbers, because the precision implied by influencer culture exceeds what small surrogate-outcome studies can support. The recovery-framing matters here: you don't need to chase soreness or novelty, and you don't need the perfect set-count. Consistent hard-enough work with gradual loading is the whole game, and that costs nothing and sells nothing — which is the tell that it is tracking truth rather than a SKU.

Cross-Pillar Connections

Hypertrophy sits in the physical pillar but its real drivers and myths reach across diet and recovery.
• Physical (resistance_training_and_body_composition): the general resistance-training and body-composition home; this entry is the growth-specific deep-dive that defers the broader training-and-composition story there.
• Physical (physical_progressive_overload): owns the progressive-overload mechanic itself — the required driver this entry names but does not re-explain.
• Diet (diet_protein_intake): owns protein dose and distribution; this entry debunks the timing window and hands the amount there.
• Physical (muscle_soreness_recovery_doms): owns DOMS physiology and recovery; this entry only uses "soreness ≠ growth" as it bears on the growth question.
• Physical (minimum_effective_dose): the recovery-framing sibling — enough hard work, not maximal novelty; the volume-to-a-point and diminishing-returns story is a minimum-effective-dose argument.
• Physical/Longevity (sarcopenia_prevention): owns muscle preservation in ageing, where the young-trained-male set-numbers do NOT transfer cleanly; this entry defers older and detrained populations there.

What would change our mind

Falsifiability: explicit upgrade/downgrade criteria from source

• We'd firm up the volume numbers and proximity-to-failure toward higher confidence if long-duration (six-to-twelve-month) RCTs with biopsy or MRI endpoints — not just ultrasound thickness — showed the dose-response, load-equivalence, and near-failure effects hold at the same magnitudes. If they shrink or reverse under better measurement, we'd downgrade.
• We'd blunt the cui-bono concern if a large independent meta-analysis from labs with NO fitness-industry conflict replicated the core findings — that would raise confidence in the directional claims and reduce the narrow-authority worry.
• We'd justify a more precise prescription if evidence emerged that a specific weekly-set target (rather than a range) reliably maximises growth across populations. That evidence is currently absent, which is exactly why we give ranges.
• We'd qualify "load doesn't matter for size" if robust biopsy data showed low-load and high-load training differ at the fibre level. The whole-muscle equivalence currently rests on surrogates that can't be corroborated by biopsy in the meta-analysis.
• We'd re-open the myth verdicts if any well-powered trial showed soreness or exercise-variation independently predicts growth when volume and effort are equated. None does now.
• What would NOT move us: the core model (tension primary, volume to a point, near-failure effort, progressive overload), because it replicates across independent meta-analyses; the myth-debunks, which are well-supported and cut against sellers. Across all of it, independent (non-seller) funding and harder endpoints are the decisive variables.

Industry bias note

Structural incentives the evidence base may reflect

This is a topic with commercial pressure at both ends, which is why the multi-meta-analysis replication and the disclosed conflicts are the anchors.
• The bro-science seller end: the supplement market profits from the anabolic-window myth (the "must slam a shake now" upsell); the novelty-program and "muscle confusion" market profits from ever-changing routines that sell subscriptions; "feel the burn / no pain no gain" programming profits from soreness as a proxy. The evidence here debunks all three — pointing AWAY from those sales incentives, which strengthens the debunks.
• The "science-based" counter-scene end: the field is concentrated in a small number of commercially-active researchers. One figure (Schoenfeld) authors or co-authors a large share of the cited meta-analyses AND works as a paid fitness educator; the proximity-to-failure paper's co-authors (Helms, Trexler) explicitly disclose earning income as fitness-industry writers and practitioners; several findings sit in the ISSN journal with historical supplement ties. None of this invalidates the findings — the disclosures are present and several findings cut against industry — but it means the AUTHORITY is narrow and the PRECISION is oversold. Confident influencer prescriptions built on n=20, eight-week ultrasound studies are marketing-grade, not settled science.
• The clean signal: trust the directional claims, because they replicate across labs and across years; discount the magic numbers, because they don't. The honest read — train with enough hard volume near failure, overload progressively, ignore soreness and novelty, hit total protein — costs nothing and sells nothing. That is the tell that it is tracking truth rather than a product.

Hypertrophy: The Few Principles That Actually Drive Muscle Growth (And the Surprisingly Short List of Things That Don't)