Delusion Calculator Data Sources
The exact government datasets behind each filter, what each one measures, how current it is, and its known limits.
Every number in the delusion calculator traces back to a public US government dataset, not to opinion, dating-app profiles, or made-up statistics. When you set a filter for height, income, age, or education, the tool compares that setting against real population data collected by federal agencies that survey and measure Americans on a fixed schedule. This page documents each source: what it measures, why it can be trusted, how often it refreshes, and one honest limitation you should keep in mind.
The calculator works with the roughly 128 million adult men in the United States, and a comparable count of adult women, drawn from these datasets. It is a US-only tool. The numbers describe the American adult population and do not carry over to other countries, which run their own surveys with different methods and results.
US Census American Community Survey (ACS)
The American Community Survey supplies the age, education, and marital status figures. Run by the US Census Bureau, the ACS reaches about 3.5 million addresses every year, which makes it one of the largest continuous household surveys anywhere. Because it samples the whole country every single year, it captures how the population shifts over time rather than freezing it once a decade like the full census does.
The ACS is trustworthy for a plain reason: participation is mandatory by law, the sample is huge, and the Census Bureau publishes detailed documentation of how it weights responses to match the known population. When you pick an age band or an education level such as a bachelor's degree or higher, the share you see reflects millions of real responses, not a small poll.
It updates on a yearly cycle. One-year estimates cover the most recent complete year for larger geographies, and five-year estimates pool data for smaller areas where a single year would be too thin to be reliable. For a national tool like this one, the one-year estimates are the relevant set, since the whole US population is more than large enough to report with confidence every year.
The three ACS filters behave a little differently from each other. Age is close to an exact count, because people report their own age accurately and the survey covers the full adult range. Education is nearly as clean at the broad bracket level, such as high school, some college, or a bachelor's degree and above. Marital status is the softest of the three, since it changes often and people describe an in-between situation such as separated or divorced in inconsistent ways.
The honest limitation: the ACS is a sample, so every figure carries a margin of error. For broad national counts that margin is small, but it never disappears. Education and marital status also depend on self-report, and people occasionally round up a credential or describe their status loosely. For age and the big education brackets this matters little, yet it is why a single percentage should be read as a close estimate rather than an exact headcount.
CDC National Health and Nutrition Examination Survey (NHANES)
Height and body composition come from NHANES, run by the CDC's National Center for Health Statistics. NHANES is different in one important way: it does not simply ask people how tall they are. It physically measures them. Participants visit a mobile examination center where trained staff record height on a stadiometer, a fixed vertical ruler with a sliding headpiece, along with weight and other body measurements.
That measurement step is the whole reason NHANES matters for a height filter. People are unreliable narrators of their own bodies. In self-reported surveys, roughly a third of adult men claim to stand exactly 6 feet tall. When the same trait is measured on a stadiometer, only about 14.5% of adult men reach 6 feet or taller. The gap between about 33% claimed and 14.5% measured is not a rounding quirk; it is a systematic tendency to round up and to aspire upward. A calculator built on self-reported height would tell you that clearing 6 feet is roughly twice as common as it really is.
NHANES is trustworthy because measurement removes that bias at the source. There is no way to exaggerate a number a technician reads off a ruler. The protocol is standardized, the equipment is calibrated, and the same methods are applied across the country.
It updates in continuous cycles, with data released in multi-year waves rather than annually. The tradeoff for its precision is a smaller sample than the giant household surveys, because examining someone in person costs far more than mailing a form.
The honest limitation: that smaller sample means finer slices, such as a specific height at a specific age, carry wider uncertainty than the headline figures. The waves also span several years, so NHANES describes recent adults well but is slower to reflect any very recent change. For a stable trait like adult height across the whole population, these are minor costs against the benefit of real measurement.
You can read more about the height numbers in what percentage of men are 6 feet tall.
Census Current Population Survey (CPS) and Bureau of Labor Statistics (BLS)
Income figures rest on the Current Population Survey and Bureau of Labor Statistics data. The CPS is a monthly survey of about 60,000 households, jointly run by the Census Bureau and the BLS, and its annual supplement is the main federal source for personal and household income. The BLS is the same agency behind the official unemployment rate and wage statistics, so its income work sits inside a long, audited tradition of labor measurement.
These sources are trustworthy because they are the government's own instruments for tracking what Americans earn, refreshed constantly and cross-checked against tax and program records. When the calculator shows how rare a given income is, it is drawing on the same data used to set poverty thresholds and report national earnings. About 18% of adult men earn 100,000 dollars or more per year, and figures like that come straight from these surveys.
The income supplement updates every year, and monthly CPS releases keep the underlying labor picture current between those annual income reports.
The honest limitation: published income data arrives in brackets, not as one continuous line. An agency might report the share of people earning up to 75,000 dollars and the share up to 100,000 dollars, but not every value in between. To answer a filter set to an in-between number, the calculator interpolates between the published brackets, estimating the point along the curve. That estimate is sound for smooth parts of the income distribution but is less exact at the very top, where high earners are fewer and the brackets grow wide. Income is also self-reported and tends to be understated at the extremes. For more on the six-figure line, see what percentage of men make 100k.
Which dataset powers which filter
The table below maps each filter to its source, what that source covers, and how the data is collected. The collection method is the part worth reading twice, because measured data and self-reported data are not the same kind of number.
| Filter | Dataset | What it covers | Collection method |
|---|---|---|---|
| Height | NHANES (CDC) | Adult height distribution | Physically measured on a stadiometer |
| Body composition | NHANES (CDC) | Weight and related body data | Physically measured in a mobile center |
| Income | CPS and BLS | Personal annual earnings | Self-reported, published in brackets |
| Age | ACS (Census) | Adult age distribution | Self-reported household survey |
| Education | ACS (Census) | Highest level completed | Self-reported household survey |
| Marital status | ACS (Census) | Married, single, and other | Self-reported household survey |
Why the source behind each number matters
Two filters can look identical on screen and still rest on very different ground. A height filter built on measured NHANES data will always disagree with one built on self-reported height, and the measured version is the honest one. That single design choice is the difference between telling you that 6 feet clears about one man in seven versus about one in three. When a result feels surprising, the collection method is usually the reason, and it is usually the more careful answer.
Grounding the tool in federal data also means the numbers are checkable. Anyone can look up ACS tables, NHANES documentation, or BLS income releases and see the same figures the calculator uses. Nothing here depends on a private database or an app's user base, both of which skew toward whoever chose to sign up. For the full account of how these inputs are combined into a single percentile, see the methodology page and the guide on how it works.
None of this makes any single percentage a perfect measurement of one real person. Surveys have margins of error, income tables are interpolated, and self-report adds noise wherever measurement is not possible. What the sourcing does guarantee is that the calculator starts from the best public evidence available rather than from a guess. You can run the numbers yourself on the female delusion calculator or the male delusion calculator, and read how it works and criteria explained for how far each figure should be trusted.