How Than Average (thanaverage.xyz) Works and How Results Are Calculated

Than Average is a small, self-described "unscientific investigation" that asks you to compare yourself to an imagined group of peers and then shows how your answers stack up against everyone else's. You answer self-reflection questions, your response is added to an aggregate, and results appear as percentages — but only for questions that have crossed a 500-response threshold. It is a curiosity project, not a research instrument, so treat the numbers as a snapshot of what visitors think about themselves rather than a measured finding.

What the site actually asks you to do

The core setup is a thought experiment. As the site puts it, you imagine you are in a room with 100 strangers who are similar to your peers and neighbours, and you answer based solely on your own instinct, perceptions, and self-reflection. Then you compare yourself to them.

Each question is a self-assessment framed as a comparison to that imagined group. The example shown on the site is:

• "Yes, I am beautiful"
• "No, I am ugly"

You pick one, and the site records it. The framing matters: you are not being asked to rate yourself on a scale, but to place yourself on one side of a comparison against people you are imagining.

How results are generated

Results are aggregated across all participants and displayed as percentages. The site's own example phrasing is "of participants also think they than average" — in other words, it shows you what share of other respondents gave the same answer you did.

Two mechanics drive what you see:

1. Aggregation, not scoring. There is no right answer and no individual score. Your response joins a pool, and the pool is summarized as a percentage.
2. A 500-response floor. Results are only shown for questions with more than 500 responses. Below that threshold, there is not enough data to display, so a question you answer may not yet have a visible result.

The site labels this openly as unscientific. That label is accurate in a specific way: the sample is self-selected visitors, the questions are subjective, and the "100 strangers" framing is a device to prompt a gut answer rather than a controlled comparison group.

Browsing and contributing

Beyond answering, the site gives you a few ways to move through the data:

Sorting by agreement is the most useful view if you want to find questions where opinion is lopsided versus genuinely split — a near-50/50 result is a different kind of signal than a 95/5 one.

What the results can and cannot tell you

They can show you:

• How a self-selected group of visitors answers the same subjective question you just answered
• Which questions produce consensus and which produce division
• A quick, low-stakes prompt for reflecting on how you compare yourself to others

They cannot show you:

• How a representative population thinks — the sample is whoever visits the site
• Whether any answer is accurate, since every question is a matter of self-perception
• Trends over time, unless the site exposes that; the visible interface is a current snapshot

The site's own framing — "how we value and compare ourselves to each other" — is the honest description of what it measures. It is a mirror for self-comparison, not a benchmark.

Practical notes before you use it

• Expect gaps in results. If a question has fewer than 500 responses, you may answer it without seeing an outcome.
• Your answer is part of the aggregate. Answering contributes to the pool that later visitors see.
• The site is small and hand-built. The page credits it as "built badly by Fred, fixed by Jae," which sets expectations for polish and scale.
• No account or payment details are indicated on the page. The visible interface is answer-and-view; nothing in the provided material describes pricing, login requirements, or data handling, so do not assume any particular policy on those points.

If you want a fast, informal read on how people rate themselves against an imagined peer group, the site delivers exactly that. If you need data you can generalize from, the 500-response floor and self-selected sample mean you should look elsewhere.