The Mr ROI operating system: the exact order I would optimize a life

The resource stack: money is only one input

A useful life does not run on cash alone. Every decision consumes some combination of money, time, attention, energy, health, optionality, and social capital. Treating only the cash price as cost is how a cheap purchase becomes an expensive obligation, or a higher salary becomes a worse life.

A lower price can come with a bill you don't pay in dollars. Saving $100 is a bad deal if it burns a Saturday. A reliable tool may be worth more because it saves setup and repairs for years. I apply the same test to work: a larger paycheck loses its shine if the commute, stress, and hours consume the rest of life. I want the complete tradeoff, not the easiest number to put in a spreadsheet.

Triage the decision before researching it

I now classify decisions before I open twenty tabs. Five questions usually tell me how much rigor the decision deserves.

1. Can failure create a large downside? Water, electrical, health, debt, legal exposure, and career reputation deserve more evidence than decorative purchases.
2. Is the cost recurring? A $20 monthly leak can matter more than a one-time $200 mistake because it keeps charging rent.
3. Is the decision hard to reverse? Housing location, career specialization, a vehicle purchase, construction behind walls, and long contracts deserve more work than an easily returned accessory.
4. Will the result be used repeatedly? A daily friction remover can justify more spending and more research than a once-a-year novelty.
5. Does the decision change a structural lever or only a setting? Income, housing, transportation, and recurring systems usually outrank minor configuration choices.

A compact version looks like this:

This is the part of the framework that protects me from my own tendency to over-research. A $400 car subwoofer driver did not deserve twenty-plus hours of analysis. A $30,000-to-$50,000 vehicle decision deserved much more, but even there a hundred hours eventually stopped changing the answer.

Use expected value, but do not fake precision

Expected value is useful when there are real probabilities and outcomes to compare. It becomes nonsense when the inputs are invented to three decimal places.

A practical version is:

Expected value of an intervention = probability it changes the outcome × value of that change - complete cost of the intervention.

The complete cost includes money, time, maintenance, risk, and opportunity cost. The probability can be a range rather than one fake number. For a large decision I would rather model a good, middle, and bad case than pretend one forecast is certain.

This is especially useful for warranties, insurance deductibles, home repairs, career moves, and preventive maintenance. If a $200 intervention has a modest chance of preventing a $10,000 failure and has little downside, the risk-adjusted case can be strong. If a $2,000 upgrade has a high chance of producing a benefit I barely notice, the arithmetic is telling me what the marketing will not.

Asymmetric downside changes the threshold

Not every decision should maximize average return. Some failures are unacceptable even if they are unlikely. A small expected savings is not enough reason to accept a meaningful chance of flood damage, electrical fire, data loss, or a major uninsured liability.

That is why I am willing to pay for redundancy and protection in some categories while rejecting extended warranties in others. Risk capacity matters. Avoid Ruin First applies that threshold to decisions capable of derailing years of progress. A loss that is trivial to one household can be catastrophic to another.

Reversibility is one of the most underrated forms of value

I assign value to the ability to change my mind.

A used product with a deep resale market can be more attractive than a cheaper new product with no exit. Renting before buying specialty equipment can be rational even when rent is expensive per day. A job that builds transferable skills can be worth more than a slightly higher-paying niche role with weak portability. A house in the right location can preserve more options than a nicer house that locks every daily activity behind a long drive.

Don't Avoid Long Odds. Avoid Bad Games. tests whether skill changes your odds and a failed attempt leaves useful value. Optionality is not an excuse to avoid commitment forever. It is a reminder to charge irreversible decisions a higher burden of proof.

Recurring friction compounds too

Compounding is usually discussed in finance, but friction compounds in ordinary life as well. A task that costs ten minutes once is irrelevant. A task that costs ten minutes twice a day is more than 120 hours per year.

That is why I care about:

• home charging instead of repeated fuel stops;
• a home gym when it removes travel and waiting from a habit I already keep;
• automation for recurring admin instead of manually remembering the same obligation;
• reliable tools that start and work instead of repeatedly being repaired;
• good location because every trip inherits the location decision;
• simple financial systems because complexity charges attention every month.

The multiplier is roughly:

annual friction = minutes per occurrence × occurrences per year ÷ 60.

Then value the hours conservatively. The point is not to claim every saved minute can be sold for wages. The point is to recognize when a tiny annoyance has become a permanent operating cost.

Build around minimum viable systems, then earn complexity

I prefer systems with a small core that survives bad weeks.

• finances: bills paid automatically, savings/investing automated, renewal dates visible;
• fitness: a minimum training week, protein/calorie structure, and enough movement to prevent a complete collapse;
• career: one recurring block that produces visible work, not just learning consumption;
• home: maintenance calendar, accessible shutoffs, documented equipment, known contractors;
• technology: backups, MFA, update discipline, simple recovery paths;
• relationships: actual calendar space rather than assuming the leftovers will be enough.

Complexity should be earned by a repeated problem. I do not want twelve productivity apps because productivity matters. I want the smallest system that reliably closes the loops I actually have.

The seven failure modes of optimization

Any system I use has to survive the fact that I'm exactly the kind of person who can overdo it.

Optimizing the measurable proxy instead of the real objective

Salary can become the proxy for career quality. Body weight can become the proxy for health. speaker price can become the proxy for sound. Net worth can become the proxy for life progress. The metric is useful only while it tracks the thing I actually care about.

Analysis paralysis

More information feels productive because it delays the risk of choosing. My research-time stopping rule exists because I can spend twenty hours improving a $400 decision by an amount that is mathematically incapable of repaying the time.

Local optimization

Saving on one product can create more expense elsewhere. A cheap floor can require replacement. A distant house can add commute cost. An inexpensive tool can create project delays. A cheaper subscription can require three other services to replace missing functions.

False precision

A spreadsheet is not evidence just because it contains decimals. When resale, repair probability, future salary, or use frequency are uncertain, ranges are more honest than single-point forecasts.

Status denial

Status is a real human preference. Pretending it does not exist makes the analysis worse. If the brand, design, rarity, or social signal is part of the pleasure, label it. The problem begins when a status benefit is disguised as engineering necessity or financial return.

Sunk-cost defense

A bad system does not become good because I already paid for it. I want an explicit exit rule for subscriptions, projects, purchases, and even career paths. The relevant question is what the next dollar or hour buys from today forward.

Turning optimization into identity

If every meal, hobby, friendship, trip, purchase, and hour requires a return calculation, the tool has eaten the objective. The point is to remove waste so more life remains, not to justify living only where the spreadsheet approves.

What to leave alone

One of the most useful optimization decisions is refusing to optimize something that's already good enough.

I would generally leave a decision alone when:

• the downside is small;
• the current solution is reliable;
• the annual savings are trivial relative to the time required;
• switching creates setup or learning cost that exceeds the gain;
• the category is intentionally ordinary;
• the purchase has already been made and no remedy remains;
• the difference is preference rather than performance;
• the decision is primarily relational or value-based rather than economic.

The last point matters most. Money can buy capacity for relationships, generosity, family, community, experiences, and recovery. It cannot tell me which of those should matter most.

The annual Mr ROI audit

Once a year, I would review the system in this order rather than constantly reopening it.

That last row is the control system for the control system. If the answer changes, the rest of the model may need to change with it.