HowToLiveBetter: 550 Evidence-Based Life Decisions in 2026
Hello HaWkers, a Chinese repository called HowToLiveBetter has surpassed 6,000 stars on GitHub by organizing 550 recommendations about health, money, work, emergencies, family, and safety. Rather than selling a magic formula, the project lays out the cost, benefit, evidence level, and primary source for each suggestion.
What is worth borrowing from this project, and what requires caution outside China? In this article, we will turn the guide into a practical method for choosing one or two useful actions, checking the evidence, and preventing a massive list from becoming guilt, misinformation, or one more abandoned task.
What HowToLiveBetter Is and Why It Went Viral
HowToLiveBetter is an open source book and search engine that brings together recommendations for everyday problems. The version reviewed on September 20, 2026, had 550 items spread across 32 thematic files. Its content ranges from longevity, sleep, and first aid to finance, layoffs, renting, parenting, travel, account security, and legal risks for technology professionals.
The format explains much of the interest. Each item shows what a person needs to spend, the expected outcome, a plain-language translation of the data, the evidence level, and a link to the source. The author says the project uses scientific papers and official documents rather than secondary articles. When there is controversy, a limitation, or a figure that has not yet been verified, the note appears in the text itself.
The project does not tell readers to complete 550 tasks. Its documentation describes it as a list of options ranked by cost-benefit and recommends starting with one or two. That distinction may look small, but it protects against the checklist effect: accumulating advice without assessing whether it fits your reality.
Four resources are at stake: longevity, time and energy, money, and personal freedom. The guide does not try to convert everything into a single unit. Avoiding debt is not mathematically compared with gaining hours of sleep; each benefit remains within its own category. It is an honest choice because combining very different numbers would create a level of precision the data does not have.
How the Guide Ranks Evidence and Cost-Benefit
In the verified version, 364 of the 550 recommendations were rated level A. This level requires a quantifiable result from a meta-analysis, a large cohort study, or a randomized controlled trial. Level B covers smaller studies or benefits that are difficult to quantify. Level C includes the author's experience or consensus without enough direct evidence.
The scale is useful, but it is not an external certification. “Level A” is a classification maintained by the project itself, and even strong studies may not apply to everyone. A trial can have excellent methodology yet study a population unlike yours; an association can be real without proving causation; a large relative benefit can represent a small absolute difference.
Cost-benefit is a second dimension. The system considers the money, time, and effort required, along with the magnitude of the benefit within each category. The repository listed 103 actions as having “very high” cost-benefit: a benefit in the highest range and no significant cost across all three axes. This rating is an editorial decision by the author, not a scientific conclusion.
We can represent the separation between evidence quality and convenience in JavaScript:
const recomendacao = {
titulo: 'Manter uma reserva para imprevistos',
evidencia: 'B',
beneficio: 'medio',
custos: { dinheiro: 1, tempo: 0, esforco: 1 },
fontePrimaria: true,
}
function podeEntrarNaLista(item) {
// A source is mandatory; the author's rating does not replace verification.
if (!item.fontePrimaria) return false
const custoTotal = Object.values(item.custos).reduce((soma, valor) => soma + valor, 0)
return item.beneficio !== 'baixo' && custoTotal <= 2
}
console.log(podeEntrarNaLista(recomendacao)) // trueThe code does not make life decisions for you. It simply prevents a common mistake: treating evidence, benefit size, and implementation difficulty as though they were the same thing.
Three Filters Before Accepting Any Recommendation
The first filter is the source. Open the study or official document, confirm that the link works, and look for the population, date, outcome, and limitations. A summary that says something “reduces risk by 20%” needs to explain which risk, over what period, and compared with what. If the source is in another language, translate the text, but preserve terms and figures that let you find the original again.
The second filter is jurisdiction. Large parts of the legal, employment, social security, and travel sections were written for residents of China. Deadlines, public phone numbers, benefits, crimes, and procedures do not automatically transfer to Brazil, the United States, Spain, or France. For these topics, use the guide to discover which questions to ask, then consult the laws and official agencies in your country.
The third filter is the individual's circumstances. Health recommendations may depend on age, pregnancy, diagnosis, medication, disability, or cardiovascular risk. The guide acknowledges contraindications in several items, but no public page knows your medical history. Educational content helps you have better conversations with professionals; it does not replace personalized medical, legal, or financial advice.
A simple function can automatically flag items that require local validation:
type Tema = 'saude' | 'financas' | 'direito' | 'habito'
type Conselho = {
tema: Tema
paisDaFonte: string
paisDoLeitor: string
envolveDiagnostico: boolean
}
function exigeEspecialista(item: Conselho): boolean {
// Law changes by jurisdiction; diagnosis changes by person.
if (item.tema === 'direito' && item.paisDaFonte !== item.paisDoLeitor) return true
if (item.tema === 'saude' && item.envolveDiagnostico) return true
return false
}The same gate should exist in human reading: the greater the harm of being wrong, the less room there is to act solely on an internet summary.
Compare the Claim With Independent Official Sources
The value of HowToLiveBetter increases when it points to a source you can confirm outside the repository. The World Health Organization, for example, maintains evidence-based guidelines on the frequency, intensity, and duration of physical activity for children, adults, older adults, pregnant people, and people with chronic conditions or disabilities.
The CDC states that adults ages 18 to 60 need seven or more hours of sleep per day. The same page distinguishes recommendations by age and suggests a regular routine, a quiet and cool environment, turning off electronic devices at least 30 minutes before bedtime, and avoiding caffeine late in the day. That is far more specific than “sleep better.”
For finance, the Consumer Financial Protection Bureau defines an emergency fund as a cash reserve for unexpected expenses such as repairs, a medical bill, or a loss of income. The agency does not impose a universal figure: it recommends thinking about the unexpected costs most common in your own life, starting even with a small amount, and automating contributions when possible.
These examples also show why context matters. The official U.S. recommendation about where to keep emergency savings mentions local banks and credit unions; the appropriate product in Brazil will be different and must account for liquidity, risk, taxes, and protections currently in effect. The principle travels better than the account name.
Before saving an action, create a minimal verifiable record:
function criarRegistro({ acao, fonte, verificadoEm, aplicacaoLocal }) {
if (!/^https:\/\//.test(fonte)) throw new Error('Fonte inválida')
if (!aplicacaoLocal) throw new Error('Falta validar a aplicação local')
return {
acao,
fonte,
verificadoEm,
aplicacaoLocal,
status: 'candidata', // This does not mean personalized advice yet.
}
}
const item = criarRegistro({
acao: 'Definir um valor inicial para emergências',
fonte: 'https://www.consumerfinance.gov/an-essential-guide-to-building-an-emergency-fund/',
verificadoEm: '2026-09-20',
aplicacaoLocal: 'Comparar opções líquidas e seguras disponíveis no Brasil',
})
console.log(item)Turn 550 Ideas Into a Seven-Day Experiment
A vast library only produces results when the action fits into a real week. Start by choosing a domain with visible friction: irregular sleep, no emergency fund, constant notifications, or disorganized emergency documents. Then select one small, reversible, and measurable change. Do not try to optimize health, money, and productivity at the same time.
Establish a baseline before intervening. If the goal is sleep, record your bedtime and wake-up time for three days. If it is attention, count interruptions during a block of work. If it is savings, record the starting balance and a sustainable transfer. The metric should serve the decision, not create obsessive surveillance.
Next, write down a stop condition. Pain, loss, anxiety, unexpected cost, or conflict with professional guidance are reasons to stop and reassess. A responsible personal experiment does not try to prove the study right; it tries to discover whether a safe action is feasible in your routine.
This small generator creates a seven-day plan without accepting more than one action:
function planoSemanal(acao, metrica) {
const hoje = new Date()
return Array.from({ length: 7 }, (_, indice) => {
const data = new Date(hoje)
data.setDate(hoje.getDate() + indice)
return {
data: data.toISOString().slice(0, 10),
acao,
metrica,
concluida: false,
observacao: '',
}
})
}
console.table(planoSemanal('Desligar telas 30 minutos antes de dormir', 'hora de dormir'))On the seventh day, ask three questions: did I perform the action on at least four days? Did the metric move in a useful direction? Does the perceived cost allow me to continue? If the answer is no, adjust or abandon it without guilt. Advice can have good evidence and still be a poor choice right now.
The Project Is Also an Open Source Product Lesson
HowToLiveBetter is more than text. It offers search by keyword, chapter, evidence level, and three cost dimensions. There are versions for online reading, PDF, EPUB, and a single offline HTML file. The page is static, with no database or backend; the Markdown content acts as the source for its interfaces and generated formats.
This architecture reduces dependencies and makes auditing easier. Anyone can open the files, follow the links, propose corrections, and keep a copy. According to the documentation, the book had to be split because the complete Markdown exceeded GitHub's 512 KB rendering limit. The scaling problem was solved without turning the reading experience into an opaque service.
There are maintenance risks too. Evidence changes, links break, laws are repealed, and translations fall behind the original. The English translation referenced by the project is unofficial and may not track updates. Downloading the HTML or EPUB helps with offline access, but it freezes the content at that moment. For important decisions, return to the original and check the date.
That caution applies to any system that summarizes knowledge. We have already seen how AI health answers can sound convincing and still require verification. An attractive interface improves access; it does not turn a weak source into truth.
How to Create Your Own List Without Copying Blindly
Start with no more than ten candidates. For each one, record its title, primary source, verification date, study population, absolute benefit when available, cost, contraindications, and jurisdiction. Rate the strength of the evidence and your willingness to act separately. Do not increase the scientific rating just because the action seems easy.
Next, eliminate anything that could cause harm without supervision. Medication changes, restrictive diets, complex investments, legal decisions, and emergency procedures require local sources and qualified professionals. The point of a personal list is not to remove experts, but to arrive at the conversation with better questions.
Rank the remaining items by the smallest first step, not the most dramatic promise. Preparing a contact list, automating a small amount, turning off a notification, or organizing documents can be more useful than chasing a huge and vague improvement. Consistency comes from reducing friction, not increasing pressure.
Finally, include a review date. A source verified today should not live indefinitely in an unmaintained file. The format below sets a 90-day deadline:
function agendarRevisao(item, dias = 90) {
const proxima = new Date(item.verificadoEm)
proxima.setDate(proxima.getDate() + dias)
// Store the date in a stable format so it can be sorted and filtered.
return { ...item, revisarEm: proxima.toISOString().slice(0, 10) }
}
console.log(
agendarRevisao({
titulo: 'Recomendação de sono para adultos',
verificadoEm: '2026-09-20',
fonte: 'CDC',
}),
)A short, reviewable, source-linked list is less impressive than 550 pieces of advice. It is also far more likely to change something.
Outlook: Evidence Is a Starting Point, Not Autopilot
The success of HowToLiveBetter reveals a legitimate demand: people want concrete decisions, visible costs, and sources they can check. The project gets it right by showing evidence levels, controversies, and limitations instead of hiding complexity behind motivational phrases.
At the same time, no universal ranking knows your health, income, family, country, or risk tolerance. The best way to use the guide is as a map of questions. It helps locate a possibility; official sources, professionals, and controlled experience show whether that path works for you.
In 2026, the advantage is not collecting more advice. It is knowing how to trace where it came from, distinguish association from causation, adapt the rule to the jurisdiction, and choose an action small enough to test. If the result does not justify the cost, discard it. If it works safely, keep it and review it when the evidence changes.
Let's go! 🦅
📚 Want to Keep Up With What Is Coming?
This article covered HowToLiveBetter and evidence-based decisions, but the ecosystem changes every week, and not everything becomes an article here.
On X, I share what I am testing, behind-the-scenes updates from my projects, and news before it becomes a post.
Follow Me There
💡 Daily content about development, careers, and the tools I actually use