Don't Avoid Long Odds. Avoid Bad Games.

The base rate is only the beginning

Suppose 1,000 people try something and 50 become highly successful.

The obvious statistic is a 5% success rate.

But imagine that 800 never seriously develop the required skills, barely execute, quit early or entered because the barrier to entry was almost nonexistent.

Suppose 200 actually meet the prerequisites and 50 succeed.

Now the same opportunity has:

• a 5% success rate among everyone who enters;
• a 25% success rate among serious qualified participants.

Those tell completely different stories.

This is hypothetical. Define the qualified group before observing the winners, or the better-looking rate may simply exclude people who failed. It illustrates the question I care about:

Why do people fail?

If they fail primarily because they do not execute behaviors I can realistically control, a low overall success rate may not bother me very much.

If serious, capable participants still usually lose because the game itself is unfavorable, the base rate matters much more.

And there is a third possibility:

The strategy works exactly as advertised, but winning still is not worth what you had to give up.

Success is not the same as a good decision

A strategy can successfully produce its intended result and still have negative total ROI.

Steroids can increase muscle and strength.

A tanning bed can darken your skin.

An extreme crash diet can make the scale fall quickly.

A reckless performance tune can make a car faster.

Working 90 hours per week can increase output.

Putting your entire portfolio into one winning stock can make you rich.

The fact that something worked does not prove it was a good bet.

I care about the entire outcome:

Desired benefit

minus

health cost + financial cost + time cost + opportunity cost + irreversible downside + added fragility

plus

whatever remains valuable if the attempt fails.

That is much closer to the real return.

My test for a good hard game

Before taking an unusual risk, I now want answers to nine questions.

The stronger the opportunity scores across all nine, the more comfortable I am taking long odds.

Choose attempts that leave useful value

Fitness is a good hard game

Building an excellent physique is a useful counterexample to simplistic base-rate thinking.

Most gym members will never come close to maximizing their muscular potential.

That does not make resistance training a lottery.

The causal mechanism is unusually controllable.

Train productively.

Progress over time.

Eat enough protein.

Manage calories.

Recover.

Stay consistent for years.

Adjust when something stops working.

Genetics strongly influence the ceiling, but behavior strongly influences how close someone gets to their personal ceiling.

More importantly, you do not have to “win” for the process to pay.

Someone who spends five years training seriously but never looks like an elite bodybuilder may still become:

• dramatically stronger;
• more muscular;
• leaner;
• more physically capable;
• more knowledgeable;
• and better positioned for healthy aging.

The losing branch still has value.

That is an excellent characteristic.

Fitness shortcuts can change the trade

Natural training can improve health and capability even when the physique falls short of the original ambition. A method that adds muscle while imposing persistent health risk has a different losing branch.

For my recreational goals, I PASS on nonmedical steroids, SARMs, unapproved performance peptides and intentional UV tanning. The Physique Shortcut Tax owns the evidence, uncertainty and medical-treatment distinctions behind that judgment.

High-income careers are another good hard game

The same framework applies to money.

A selective career can have a low acceptance rate while still being an excellent target.

Software development, data science, technical sales and management analysis all have national median wages above six figures in May 2025, according to the BLS occupation pages linked in Sources.

These jobs are competitive precisely because the underlying skill has market value.

If someone spends 1,000 hours becoming much better at:

• analytics;
• SQL;
• software;
• technical communication;
• product thinking;
• financial analysis;
• sales;
• or another scarce commercial skill

and then gets rejected from their dream employer, the 1,000 hours do not disappear.

Other employers value many of the same capabilities.

The attempt creates career capital. My White-Collar Job Playbook shows how I turned that into a career move.

That makes an ambitious career target fundamentally different from an activity where failure leaves almost nothing transferable.

Real estate is conditional, not automatically bad

Real estate illustrates why I do not want this framework to become an excuse for risk aversion.

Real-estate sales agents had a national median annual wage of $52,830 in May 2025, while the highest 10% earned more than $123,590. These BLS wage figures exclude self-employed workers, so they do not describe every agent or owner.

Clearly, high-income agents exist.

The profession is not inherently a bad game.

But getting a license does not create customers.

The economically important capability is customer acquisition.

Someone who:

• enjoys sales;
• has a large local network;
• follows up relentlessly;
• understands marketing;
• answers leads immediately;
• builds referrals well;
• can tolerate irregular income;
• and has enough financial runway

may have dramatically better odds than the average entrant.

Someone who hears that houses sell for $500,000 and assumes a commission percentage automatically translates into easy income has misunderstood the business.

My verdict is therefore not PASS.

It is:

TEST if you have reason to believe you can acquire customers.

Validate that before reorganizing your life around the upside.

Real entrepreneurship also belongs in TEST

“Most businesses fail” is about as useful as saying “most people who buy dumbbells never become muscular.”

What business?

What market?

What founder?

What capital requirements?

What customer-acquisition economics?

A skilled plumber opening a local plumbing company is playing a different game from someone ordering generic products from Alibaba because a course said dropshipping creates passive income.

I want to reduce entrepreneurship into testable questions.

Can I find someone with the problem?

Will that person pay me?

Can I deliver profitably?

Can I acquire the next customer?

Do customers return or refer others?

Does demand survive outside my personal network?

Only after those answers improve should the commitment become larger.

The dangerous process works backward: buy equipment and inventory, form the LLC, build the website, buy a course and quit the job, then look for demand.

That can feel like entrepreneurship while producing virtually no enterprise value.

Activity is not traction.

Where confidence needs much stronger evidence

Active investing requires proof of edge

Buying individual stocks is not automatically gambling.

A person can intelligently analyze:

• valuation;
• cash flows;
• competitive position;
• management;
• industry structure;
• balance-sheet strength;
• and expected returns.

The harder question is whether that analysis produces an advantage over the market.

Public markets are competitive.

In 2025, 79% of active U.S. large-cap funds underperformed the S&P 500 according to SPIVA.

For an individual investor, the next question is whether there is a repeatable reason to expect better results.

It proves that confidence is not evidence of an edge.

If someone claims to have one, I would want:

• the complete portfolio;
• losers as well as winners;
• several years of results;
• an appropriate benchmark;
• risk and drawdown information;
• taxes and costs;
• consistent position sizing;
• and an explanation of the mispricing that existed before the outcome became obvious.

If someone demonstrates persistent risk-adjusted outperformance, update the conclusion.

Until then, broad low-cost indexing remains a very strong default for core wealth building.

If stock picking is intellectually interesting, a bounded active sleeve can test the hypothesis.

Do not make the retirement portfolio the experiment.

Day trading is a different proposition

FINRA's required day-trading disclosure tells prospective participants to be prepared to lose all funds committed to day trading and explicitly notes that they are competing with professional traders.

That should raise the burden of proof.

The question is not whether profitable traders exist.

They obviously do.

It is:

Why should I expect to possess a repeatable advantage against professional counterparties after costs, mistakes and taxes?

“I'm disciplined.”

“I'm intelligent.”

“I will learn.”

“I'll work harder.”

None of those is a unique edge.

For almost everyone whose goal is simply wealth rather than mastering trading itself:

PASS.

MLMs fail yet another part of the test

The FTC reviewed income disclosures from 70 MLMs and found that most participants made $1,000 or less annually, often before expenses, while many made nothing.

Look beyond the highest earner to the distribution of outcomes and the costs of reaching them.

Some are probably extremely good at sales, recruiting, audience building and persuasion.

But then another opportunity-cost question appears:

If you genuinely possess exceptional sales ability, why deploy it in a structure with such poor participant economics?

Those same skills can potentially create much more value in conventional sales, business development, recruiting, entrepreneurship or ownership.

That makes the visible winner a particularly bad advertisement for the marginal participant.

For almost everyone:

PASS.

Judge the whole system

Home improvement can contain the same trap

DIY has excellent ROI when:

• the task is understandable;
• mistakes are visible;
• failure is cheap;
• tools and skills transfer;
• and you can stop before causing serious damage.

Painting a wall poorly is annoying.

Redo it.

A mistake involving structural work, water management, high-voltage electrical work or another high-consequence system can have a completely different failure mode.

Saving $2,000 is not impressive if the unlikely losing branch is a fire, hidden water intrusion or a $20,000 repair.

The correct comparison is not:

contractor cost versus DIY material cost.

It is:

contractor cost versus expected DIY cost after accounting for time, tools, probability of failure, severity of failure and learning value.

The more irreversible the consequence, the less interested I am in proving that I can do everything myself.

Cars and technology have bad games too

The same logic applies to enthusiast hobbies.

A modification might genuinely improve:

• horsepower;
• acceleration;
• bass output;
• display brightness;
• benchmark performance;
• or some other headline metric.

But those gains can create second-order costs.

More power may reduce drivetrain margin.

Extreme suspension changes may make a daily driver worse.

An audio upgrade can create electrical or packaging problems.

A complicated home-server stack may technically save subscription fees while consuming dozens of hours maintaining infrastructure that an inexpensive service handled automatically.

Again:

Did it work?

is different from:

Did it improve the entire system?

Mr ROI cares about the second question.

Find the actual goal before choosing the strategy

A lot of bad games survive because people become attached to the method instead of the objective. Once the real objective is identified, alternatives can compete.

That is a much more useful decision than debating whether a category is universally “good” or “bad.”

The alternative does not have to be identical. It needs to preserve enough of what I actually value that the remaining incremental benefit no longer pays for the risk.

Put a stop-loss on ambition

Before starting a speculative project, I want to know: What result would tell me this hypothesis is wrong?

You cannot always choose a perfect deadline. You can choose not to create an experiment that is unfalsifiable.

The research-time stopping rule explains how to set a bounded test and decide when more information is worth buying. Avoid Ruin First handles commitments whose downside could derail years of progress.

Watch who gets paid even when you lose

Incentives often expose a bad game before statistics do.

Ask:

Who profits if I participate but never succeed?

The course seller may already have been paid.

The MLM may sell inventory.

The broker gets transaction revenue.

The supplement seller sells another bottle.

The financing company collects interest.

The influencer earns affiliate revenue.

Check the product on its merits, then account for what the presenter earns when you buy.

Most legitimate companies also get paid before knowing whether every customer will succeed.

But when the seller's economics are excellent because participants continually fail and replace themselves, I become much more skeptical.

Especially when exceptional winners are the primary marketing material.

A winner does not validate the game

You know a Realtor making $300,000.

A trader became a millionaire.

An influencer built a huge business.

A bodybuilder looks incredible on PEDs.

Someone tans constantly and has not developed skin cancer.

None of those observations settles the decision.

Every result is some combination of:

skill + effort + resources + structural advantage + exposure to risk + timing + luck

The question is how reproducible the important pieces are.

And then:

What did the winner have to risk to produce the result?

Survivorship bias does not disappear just because the winner genuinely worked hard.

Both things can be true.

A life optimized exclusively around avoiding downside would be its own bad game. The highest-return decisions often require uncertainty. Choose the difficult things where becoming exceptional changes the economics, and be skeptical when a temporary reward can leave consequences lasting years.