What is a good Rice Purity Test score?

There is no good Rice Purity Test score. The test has no pass mark, no answer key and no norming sample — it is a checklist of one hundred experiences, and your number is simply one hundred minus the ones you have had. A lower score does not mean worse and a higher score does not mean better; it mostly means younger.

The question still gets typed into search boxes every day, and it is worth being honest about why, because the real question underneath it is usually not about the number.

Why people ask anyway

Almost nobody takes this test alone. It arrives in a group chat, a dorm room, a car, a comment section, and the score gets said out loud or screenshotted within minutes. That makes "is my score good?" a social question with a precise meaning: am I an outlier in the room I am about to show this to?

Both tails are uncomfortable. A 96 in a group averaging in the 60s reads as sheltered; a 31 in a group averaging in the 80s reads as an overshare you cannot take back. The anxiety is symmetrical, which is the clearest sign that the number is not measuring quality — it is measuring distance from whoever happens to be nearby.

The screenshot makes it worse. A spoken answer fades; an image of a number sits in a chat log for years, attached to your name, without the context of the conversation it was posted in. Before you post a score, the useful question is not whether it is good but whether you would be comfortable with that image surfacing in eighteen months.

What each range actually implies

The only defensible reading of a score is mechanical. Your score is 100 minus the number of boxes you checked, and the list is ordered — so a total tells you roughly how far down the sheet someone got. The ranges below are read straight off the question order on the classic list, counting from the top: a 70 is thirty checks, which reaches question 30, and question 30 is masturbation. Not alcohol — the drinking block does not start until question 53, which is why the range tables you will find elsewhere put first drinks twenty or thirty points too high.

Nobody checks in strict list order, so treat the table as the shape of the list rather than a reading of a person. Two people can land on exactly the same number by checking almost entirely different questions: a 62 built from a long relationship and a 62 built from three years of nightlife share a total and nothing else. Every value has its own breakdown page under score meanings if you want the detail for yours.

Age is the dominant variable, and it is not close

The test is cumulative and irreversible. Nothing on the list can be un-done, so your score can only fall or stay flat for the rest of your life. That single property makes it, structurally, a counter of elapsed exposure — and the biggest input to elapsed exposure is how long you have been alive and out of the house.

The consequence is worth stating bluntly: comparing scores across ages is mostly comparing ages. A 15-year-old at 88 and a 24-year-old at 88 are not similar people; the second one is unusual and the first one is typical. If you want a comparison that means anything, it has to be against people at roughly your stage, and the honest version of that comparison is your own score at different points in time — which is what your saved history is for.

After age, the strongest inputs are environment rather than personality: living away from home versus commuting, a city versus a small town, a social scene built around bars versus one that is not. None of that is character, and all of it changes which blocks of the list you get the chance to check at all.

Four things the score is bad at

Weighting. Every item is one point. Being arrested and stealing a road sign cost the same. A total that treats those as equal units cannot be read as a severity measure.

Context. The list records whether something happened, never how, with whom, or whether you wanted it to. Some questions can be answered yes by someone describing an experience that was not their choice. A number that cannot tell those apart should not be treated as a description of a person.

Honesty. Nothing is verified, and nothing can be. A score is inflated by embarrassment and deflated by bravado, and there is no way to tell a filled-in column from a lived one — which is why nobody, this site included, can tell you how common either is.

Resolution. A three-point difference is noise. The gap between 64 and 67 is one ambiguous question you might answer differently on a different day.

What about a perfect 100, or a 0?

A 100 means you answered no to every question, including the first few, which are mild enough that most adults check them without thinking. It is a real result for younger teenagers and rare enough among adults to be worth a second look at the sheet rather than a celebration or an apology.

A 0 is the opposite edge, and answering yes to literally every item — including being convicted of a crime and injecting a drug — is a combination few lives contain. It is also the easiest score to produce by dragging down the column, and the page cannot distinguish the two. Both ends of the scale are where the data is least trustworthy, which is the practical reason not to build an identity on either one.

How to read your own number

Read the clusters, not the total. Which blocks did you check, and which did you leave empty? That is the specific thing the number knows about you, and the total flattens it into a digit. Every score has its own page with the questions its count reaches.

On comparing — against a friend, against the site average, against your own past score — how to read your percentile covers what each of those comparisons can and cannot support, and it is a shorter answer than it looks.

Finally, treat the result as what it is. The test began as an icebreaker for people who had just met (see where the Rice Purity Test came from), and it is still best at that job: it produces specific stories fast. It is not a ranking, and it has never been an evaluation.

Ready to get the number itself? Take the 100-question test, or pick a version suited to your situation from all four tests.