With genealogical evidence, you should not trust any one source to be accurate. I contend that even a preponderance of the evidence is not necessarily accurate. Researchers often look for corroborating evidence from other sources before accepting a fact as true. Various sources have different weights as to their trustworthiness and accuracy. But it can still be a mistake to draw a conclusion based on a given set of sources.
For instance someone’s date of birth taken from their death certificate has a greater chance of being incorrect since their birth happened so many years beforehand. The person filling out the death certificate may only be guessing the deceased’s birth date. The birth date is seldom verified with other official records when the death certificate is filed.
Birth certificates are considered more accurate because they are recorded soon after birth, when everyone involved is sure when it happened. However a clerk generally recorded births in a ledger book. Sometimes these ledgers were themselves re-copied several years later. So even a “birth certificate” is subject to transcription errors and recording mistakes.
Even if a birth record and a death record each point to the same birth date, that date is not necessarily accurate. Two or more inaccurate records do not make an accurate record.
That is why citing a source in genealogical research is so important. You, or a subsequent researcher may happen upon another source in the future that corroborates or refutes a given fact. All sources will again need to be weighed for accuracy before another conclusion can be drawn. Any genealogical fact has an inherent degree of accuracy that is never 100%.