The Long Game is a living reference for nutrition and fitness. Its governing priority is lifelong health, physical capability, and quality of life. Hypertrophy—muscle development—is a major supporting goal, not the criterion that decides every tradeoff.
The founding edition introduced three reader-facing chapters, this specification, an evidence register, and a revision history. The guide remains a foundation as those chapters are revised. It is not a complete encyclopedia, a formally conducted systematic review, or an independently reviewed clinical guideline. The initial research and writing are AI-assisted. No independent clinical or specialist review has been completed.
The daily process is intended to improve the same guide. It is not a requirement to produce a new recommendation every day. A recorded search that does not justify a change is a legitimate result.
Scope and intended reader
The initial audience is adults seeking understandable general education. Readers need no specialist terminology. Every chapter should introduce its own necessary concepts, explain calculations, and distinguish a finding from advice constructed using that finding.
The core subjects are nutrition, resistance training, aerobic activity, everyday movement, body-composition decisions, and recovery across adulthood. Relevant medical context belongs at the boundaries, but this guide does not replace individualized diagnosis, treatment, rehabilitation, or exercise clearance.
Children, pregnancy, breastfeeding, eating disorders, disease-specific nutrition, medication changes, and rehabilitation require separately developed expertise and evidence. They must not be handled by casually extending an example for generally healthy adults. Disability and older age must not be treated as afterthoughts, either: applicability and access should be stated rather than implying that one gym routine represents everyone.
We do not rank people by body size, prescribe an appearance as a health requirement, or moralize missed sessions and food choices. The purpose is to improve decisions, not to recruit readers into an identity.
What a source must support
A substantive factual claim needs traceable support for its actual wording. A citation about a nearby subject is insufficient. The population, intervention or exposure, comparison, outcome, and timeframe matter. Evidence about a short-term response cannot silently become evidence about decades of survival. C09 C10
The evidence register gives consequential claims permanent identifiers. Each record contains a bounded statement, statement type, descriptive confidence label, applicable population, outcome, limitations, sources, and dependencies on other claims. Source records show what was accessible and inspected. An official summary and an abstract-only record are not presented as full-text appraisals.
Our labels—such as guideline-based, qualified, reasoned, and unconfirmed—are project descriptions. They are not formal GRADE ratings. We borrow useful appraisal questions about bias, consistency, applicability, precision, and missing evidence without claiming to have completed the formal process. C10
Repeated publications, overlapping reviews, and reports from the same underlying study must not be counted as independent confirmations. The registry includes an evidence-group field to support that distinction; the presence of the field does not mean every overlap in the literature has already been investigated.
Three different contributions
A research finding states what an inspected source supports, with its limitations. A practical synthesis adds explicit assumptions and choices to turn evidence into an example or decision rule. An original hypothesis proposes something that has not been established.
The worked week is a synthesis. Its component guidance does not prove that the whole schedule is best. A hypothetical risk calculation is a teaching device, not an observed study result. An original hypothesis remains labeled wherever it appears; repeated use inside this guide cannot supply independent evidence for it.
An open hypothesis
H01 — unconfirmed: a preplanned, reduced-demand fallback might improve long-term participation compared with an equally detailed plan that offers only its full schedule. Guidance supporting feasible training and some activity rather than none motivates the question, but does not establish the comparative advantage. H01
A useful test could randomly assign adults to plans with or without an explicit fallback, prespecify how participation is measured, follow them long enough to capture interruptions, and examine function and unintended effects as well as adherence. A competing explanation is that extra planning adds complexity without improving participation. Another is that a fallback reduces activity when the fuller plan was feasible.
The guide may offer a smaller example because readers sometimes have less time available. It must not advertise the untested hypothesis as proof that the smaller-plan option improves health or adherence. Original reasoning is welcome; invented experimental validation is not.
A complexity budget for the reader
Every new obligation must explain what decision it serves, its likely importance, its burden, and what it might displace. More reference detail does not automatically mean more tasks for the reader.
A proposed measurement must answer “what would I do differently with this information?” before it becomes a default tracking requirement. A recommendation with small or uncertain benefits should not acquire the presentation of an indispensable foundation merely because it is interesting to research.
Examples should include ordinary budgets, limited time, different food preferences, and different access needs. Alternative examples must be honest about evidence gaps; demographic variety is not permission to invent tailored prescriptions.
This section is editorial policy. It is not a validated score for measuring the value of an intervention.
What happens during a daily review
The reviewer starts from the current repository, not conversational memory. They read the guide’s structure, all central recommendations, the claim register, unresolved questions, and the previous review. While the guide remains small, the full reader-facing text is read. As it grows, the record must distinguish a whole-guide structural review from a fresh appraisal of every source; summaries are never represented as full-text verification.
One focused question is then selected for its importance, uncertainty, or a known explanatory gap. Searches should include conflicting evidence and foundational work, not only recent papers supporting the current position. Routine work has a bounded scope: usually one question, up to ten candidate sources, and up to four detailed source inspections. These are planning limits, not quotas or permission to call an incomplete search exhaustive.
The reviewer decides whether the evidence changes an action, applicability, confidence, explanation, or nothing important. All affected chapters, examples, summaries, and future calculators are considered together. The data model tracks dependencies so that changing a protein premise can flag both its explanatory chapter and meal examples using it.
A weekly structural audit supplements daily review; a monthly domain audit checks neglected subjects. Neither audit authorizes needless rewriting. Delete weak or distracting material when that improves the guide. Preserve stable links or add redirects when restructuring.
Publication gates and correction policy
Routine explanatory and educational changes may be published after source, consistency, calculation, and build checks. Consequential new clinical advice, experimental treatment claims, or higher-risk recommendations require appropriate qualified review and must remain held when that review is unavailable. A model-generated second pass does not count as independent expert review.
The repository validator checks declared source links, claim dependencies, page coverage, metadata, unsupported default hypotheses, and several release conditions. Rehearsal tests introduce withdrawn sources, conflicting premises, missing coverage, cycles, and unjustified confidence upgrades. Passing software checks does not prove that a scientific claim is true or that every unmarked sentence has been correctly sourced. Semantic source verification and prose review remain separate editorial responsibilities.
Evidence corrections must be conspicuous. A materially changed recommendation receives an explanation of what changed and which readers are affected. Old Git revisions remain available. A formatting edit must not advance an evidence-review date. A source being inaccessible must be recorded, never concealed by a claim of verification.
The revision history distinguishes publication, evidence review, and practical recommendation changes. The current dates describe work actually performed, not a promise that tomorrow’s review has already happened.
Access, independence, and what comes next
This section does not collect personal health measurements or offer personalized clinical assessment. It requires no paid supplement, wearable, or gym product. No affiliate links or sponsorship claims are introduced in the founding edition; future relevant interests must be disclosed rather than assumed absent.
No paid source access, subscription, or additional recurring service is authorized by the editorial process alone. An inaccessible source may limit what can be concluded. The reviewer should document that limit or pursue an appropriate alternative, not invent access.
A targeted full-text review of the 2026 resistance-training position stand is now recorded in S14. This does not complete appraisal of its constituent reviews. The food-foundation chapter now gives dietary patterns their own place in the reading path, using institutional guidance and labeled meal-building examples. Detailed nutrient adequacy and the underlying dietary-pattern trials still need deeper appraisal. Further priorities are the constituent resistance-training evidence; progression and hypertrophy dose across experience levels; and adapting activity through older age and different physical capabilities. Body-composition change, recovery, supplements, and monitoring also need dedicated treatment. These are acknowledged gaps, not empty chapters presented as finished coverage.
Success is not a growing word count. It is a reader finding a useful answer, understanding why it is offered, seeing where it stops applying, and being able to inspect the support. When a revision fails those tests, the right outcome is to hold it and record the reason.