The lowest rice purity score ever recorded in the Thresher dataset is 0, reached by only 336 of 124,952 takers. Learn why verified 0 scores are so rare and how to judge viral low score claims.
The lowest rice purity score ever recorded in the best documented dataset is 0. In the Rice Thresher dataset, covering 124,952 test takers before 2018, only 336 people scored 0, while 2,719 scored a perfect 100. A score of 0 is therefore real in the data, but extremely rare, far rarer than a perfect score.
That fact answers the question, but it raises a second one. How many of the 0 scores shared online are genuine? This guide separates recorded facts from viral claims, and shows what counts as a believable very low score.
The Rice Purity Test has 100 questions, and your score starts at 100. Every question you check removes one point. To score 0, you must check all 100 questions, from ordinary items such as holding hands, to the hardest items at the end of the list, including items that involve serious legal trouble and items most people never encounter.
A score of 20 means 80 checked items, a score of 10 means 90, and a score of 0 means everything. Each step down gets harder, because the last few items are the ones very few people can honestly check. See our guide to what a rice purity score of 0 means for the full picture.
Almost every number quoted about rice purity scores traces back to one of two sources. The first is the Rice Thresher dataset, from the student newspaper at Rice University, covering 124,952 takers before 2018. The second is arealme, which reports an overall weighted average of about 63.85 and a most common single score of 93. This article uses the Thresher numbers for records.
The Thresher figures give the record its context. The average score was 61.46. At the top end, 2,719 people scored 100. At the bottom end, only 336 scored 0. About 2.2 percent scored a perfect 100, while only about 0.27 percent scored 0. A perfect score was roughly eight times more common, which shows how much harder the bottom of the scale is to reach.
| Thresher result | Number of takers | Share of all takers |
|---|---|---|
| Total takers in the dataset | 124,952 | 100 percent |
| Average score | 61.46 | Not applicable |
| Scored a perfect 100 | 2,719 | About 2.2 percent |
| Scored 0 | 336 | About 0.27 percent |
Those percentages explain the curiosity. A result that only about one taker in 370 reaches feels like a record, yet real cases exist in a large dataset. Do not treat every claimed 0 online as one of those 336 cases.

The Thresher dataset is useful because it was collected in one place, from one version of the test, over a defined period. A screenshot posted online has none of those controls. You cannot see who took the test, which version they used, whether they answered honestly, or whether they edited the result. A score page can be faked in seconds, and a test page can be retaken until the desired number appears.
Version differences make verification harder still. Some sites run a weighted variant, in which some questions count for more than others, or a shortened variant with fewer questions. A 0 on either variant is not the same result as a 0 on the classic 100 question test, even though the screen shows the same number.
There is also a privacy problem with proof. A genuine 0 would require publishing a full list of 100 checked experiences, including the most private items on the test. Almost nobody will do that, and nobody should be asked to. So the strongest public evidence for a 0 is usually a single number on a screen, which is also the weakest evidence.
Believability is a spectrum. A score in the 20s is unusual but believable for an older adult with a full life. A score in the teens means 80 or more checked items across almost every category. A score in single digits means 91 or more checked items, nearly everything on the list. A score of exactly 0 requires literally everything.
Age matters a great deal. The average score falls as people get older, because each year adds chances to check new items. A teenager claiming a single digit score has had far fewer years to accumulate those experiences than a person in their thirties, so the same number is much harder to believe at 16 than at 36.
Our own site data points the same way. Our homepage counter shows 170 tests taken so far, with an average score of 64, close to the Thresher average of 61.46 and the arealme average of about 63.85. In a group of that size, you would expect no 0 scores, because the Thresher rate of about one in 370 means 170 takers may well contain none.

Low score claims go viral because they are shocking, and shock travels faster than verification. A streamer who announces a very low score gets clips and comments, while a quiet correction gets almost nothing. That rewards exaggeration: checking extra boxes, or typing a lower number into a thumbnail, costs nothing.
There is no reliable public list of verified celebrity rice purity scores. A genuinely verified 0 would require full answers, exposing private details no public figure would reasonably share. What circulates instead are screenshots and jokes told on camera, none of which can be checked. Treat viral low score claims as entertainment, not data.
First, ask which version was used. A classic 100 question result and a weighted variant result are different numbers sharing a scale. Second, ask whether the full answers were shown, not just the final number. A number alone proves nothing. Third, weigh the age of the claimant against the experiences the score requires. Fourth, look for a reason to exaggerate, such as a video or an audience. Fifth, compare the claim with the Thresher rate. A 0 claim from a young claimant or a creator chasing views starts with long odds against it.
None of these checks can prove a claim true, but together they show when doubt is warranted. As our history of the rice purity test explains, the test began as a paper checklist in the 1920s, long before screenshots or social media. Its charm was always private self reflection, not public proof.
A very low score does not mean a person is good or bad, trustworthy or untrustworthy, interesting or boring. It means they checked many items on one list, at one point in time, as they interpreted the questions. Two people with the same score may have lived very different lives, because the test counts experiences without weighing context or timing. A score also freezes a moment. On the classic version, a later retake can only keep the score the same or lower it. What a very low score does show, reliably, is rarity. That rarity is worth understanding, but not worth chasing, because dishonest checking defeats the test's only purpose: an honest private snapshot.
Yes, in the Thresher dataset 336 of 124,952 takers scored 0. That is the best documented evidence for a 0 score, and it shows a genuine 0 is possible but extremely rare.
The lowest rice purity score possible on the classic test is 0, which requires checking all 100 questions. No score below 0 exists on the classic scale, although weighted variants can display different results.
Yes. In the Thresher dataset, 2,719 people scored 100 against only 336 who scored 0, so a perfect score was roughly eight times more common than a zero score.
Treat it as unverified. Screenshots can be edited, tests can be retaken, and weighted variants can produce the same number for different reasons. Without full answers, a viral 0 is entertainment rather than evidence.
There is no fixed floor, but single digit scores are very hard to believe for a teenager, because the experiences required take years to accumulate.