Learn it → Apply it → Save as Skill
Before a deal is signed, the people most eager to say yes are often the ones you'd least want saying yes. When one side knows more than the other about the risk being traded, the terms you offer don't draw a random sample of takers. They draw a skewed one.
The Dr. Wallace's Favorite Mental Models is designed to turn useful ways of thinking into reusable tools. Each model gives you a fast visual overview, deeper interactive learning, practical application exercises and an AI-ready Markdown skill you can save for future use.
Instead of merely reading about mental models, you can gradually build your own AI skills library: a collection of portable reasoning skills that can help ChatGPT, Claude, Grok and other AI assistants apply the mental models you value.
When you cannot verify quality or risk directly, the price or terms you set do the selecting for you. A generous flat rate is a magnet for exactly the applicants who know they are a worse-than-average bet.
Each bar is the share of low-risk participants left in the pool as flat terms repeat. The good risks leave first because they have better alternatives elsewhere.
Adverse selection matters most when one side of a transaction holds information the other side cannot cheaply check, and a single price or set of terms has to serve everyone.
The method is about anticipating who your own terms will attract, not about judging people after the fact.
Name the trait the other side knows and you don't: true health, true repayment risk, real condition of the item, real skill level.
Ask who is most motivated to accept your current price or terms given that hidden trait.
Build a mechanism that makes the hidden trait visible or costly to fake: underwriting, warranties, trial periods, credentials, deductibles.
If you can't screen, adjust price, terms or risk-sharing to survive the skew — or decide the segment isn't worth serving this way.
Open a use case to see what the model changes.
When coverage is optional and priced on the average person, people who know they are likely to file a claim are the most eager buyers. Premiums rise to cover the skewed pool, which pushes healthier people out and skews the pool further — a spiral insurers call a "death spiral" without mandates, medical underwriting or risk pooling.
A borrower usually knows their own repayment risk better than the lender does. One "average" rate looks generous to a risky borrower and unattractive to a safe one, who can get a better rate elsewhere. Tiered pricing, collateral and credit history checks exist to close that gap.
Strong candidates with several offers can afford to be selective; a role with vague requirements, low pay or a rushed process disproportionately attracts people with fewer alternatives. The applicant pool itself is evidence about how the role is positioned.
In used goods and freelance platforms, the seller usually knows the item's or their own true quality better than the buyer. If buyers can't verify quality, they offer an average price; owners of above-average items find that price too low and exit, and the average quality — and price buyers are willing to pay — falls again. This is Akerlof's original "lemons" case.
Owners often know more about a company's hidden liabilities, customer churn or key-person risk than a buyer does. Unusual eagerness to sell, especially at an attractive price, is itself information. Due diligence, warranties, escrow and earn-outs exist to close the gap.
If a counterparty accepts your first offer unusually fast, ask what they know about the deal's value that made it an easy call for them. Adverse selection reframes a quick agreement as a question rather than a victory.
This is an illustrative example designed to teach the method.
Expand for the origin of the term and how it relates to two neighboring ideas.
"Adverse selection" was used in the insurance industry through the 19th and early 20th centuries to describe how voluntary coverage disproportionately attracts higher-risk buyers. Economist George Akerlof gave the idea its modern, general form in "The Market for 'Lemons': Quality Uncertainty and the Market Mechanism," published in the Quarterly Journal of Economics in 1970, using the used-car market as the illustrating case.
In 2001, the Sveriges Riksbank Prize in Economic Sciences in Memory of Alfred Nobel was awarded jointly to George Akerlof, Michael Spence and Joseph Stiglitz "for their analyses of markets with asymmetric information." Akerlof's work described the adverse-selection problem; Spence's job-market signaling model and Stiglitz's work on screening described two different responses to it.
Adverse selection is about hidden information before a contract: who chooses to enter the deal given the terms offered. Moral hazard, covered elsewhere in this library, is about hidden action after a contract: how a party's behavior changes once risk has been transferred to someone else. The two often appear together — an insurer worries about who buys the policy (adverse selection) and how insured people then behave (moral hazard) — but the diagnosis and the fix for each are different.
Your own terms are often exactly what changed who is in the pool. Treating applicants as a random sample hides the skew.
One is about who selects in before the deal; the other is about how behavior changes after. Screening fixes the first; monitoring and deductibles address the second.
A signal only works if it is genuinely costly for the type you want to exclude. A self-reported survey question anyone can answer favorably filters nothing.
Repricing upward in response to a skewed pool can make the next round's skew worse, not better, if low-risk participants keep leaving.
An unusually fast "yes" can be information about what the other side knows, not a compliment to your offer.
Screening that is too costly, slow or invasive drives away the low-risk people you most wanted, leaving a smaller version of the same skewed pool.
information is roughly symmetric — both sides can see the same specs, history or track record, as with many standardized commodities.
the concern is how a party's behavior might change after risk is transferred, not who chose to enter the deal in the first place.
you can check the hidden trait cheaply and reliably — a credit bureau report, a factory inspection, a standardized test — which removes the information gap rather than pricing around it.
This browser tool structures your thinking. It does not send your information anywhere or perform AI analysis.
Save as Skill converts the model into portable Markdown instructions that teach an AI when and how to apply adverse selection, what questions to ask, what evidence to seek, what mistakes to avoid and how to structure its analysis.
Save the skills you find useful and organize them into your own reusable library. The Markdown is intentionally model-neutral, so it can be used wherever an AI system accepts uploaded files, project knowledge, custom instructions or reusable text context. Platform capabilities differ, so this is a portable skill file rather than a claim of one-click installation.
No. Insurance is the classic case because the term originated there, but the same structure appears anywhere one side sets terms without being able to verify a trait the other side already knows: lending, hiring, used-goods and freelance marketplaces, M&A, and ordinary negotiation.
In Akerlof's model, if buyers cannot distinguish good cars from lemons, they will only pay an average price reflecting the mix. Owners of above-average cars find that price too low and exit the market; the average quality — and the price buyers are willing to pay — falls again. Repeated enough times, the process can unravel a market for quality almost entirely.
Through institutions that close the information gap: warranties and return policies, third-party certification and inspection, reputation and review systems, mandatory disclosure requirements, and risk pooling through mandates (as with some insurance markets). Each is a different way of making the hidden trait visible, verifiable or costly to misrepresent.