Finding the same awazon onion link on eight pages feels like confirmation. Usually it is one fact repeated eight times. Directory sites copy each other, scrapers copy directories, and a mistake made once propagates outward without anyone re-deriving anything. Volume of agreement tells you how well something spread, not whether it was right when it started.
What you are looking for is two sources that could not have got it from each other. That is a much rarer thing and it is worth far more than a long list.
Signs that two sources are really one
- The same typo, the same spacing, the same odd capitalisation. Text that travelled together carries its scars.
- The same ordering of a list. Independent collections almost never agree on order.
- The same surrounding sentences. Descriptions get copied along with the data they describe.
- One cites the other, or both cite a third page that neither of them checked.
- Both appeared at once. Simultaneous publication usually means a shared upstream rather than two people doing the same work on the same afternoon.
This site is a source like any other and the same test applies to it. It publishes addresses as they were given and checks nothing. If it agrees with a page that got its list from the same place, that agreement is worth nothing at all, and you should treat it that way.
What genuine independence looks like
A signed statement from the market and a directory listing are independent in the way that matters, because one is derived from a key and the other is derived from someone typing. A string you copied off a forum post from last year and the same string in your own dated file are independent in a weaker but still useful sense, because they were captured at different times by different processes.
The strongest version is a signature you can check against a fingerprint you already held. At that point you are not comparing two rumours, you are checking one claim against material in your possession, which is why collecting that fingerprint early is worth the trouble it takes.
Doing it without spending an evening
The whole exercise takes a few minutes if you are honest about scope. Two places, chosen because they are unlikely to share a source. Write both strings down separately rather than pasting one and glancing at the other, because glancing is how a person confirms whatever they already have.
Compare from both ends inwards. Human eyes are good at the start of a string and lazy in the middle, and a difference planted in the middle of a base32 address is the one nobody catches by reading normally.
When the sources disagree
Do not average them and do not pick the prettier one. Note both, date both, and treat the item as unresolved until something with a signature settles it. An unresolved item you know about is safe. A resolved item you guessed at is the dangerous kind, because you will stop thinking about it.
Disagreement is also information about the sources themselves. A place that turns out to be wrong once has told you something useful about the rest of its list, and that is worth remembering the next time you need to collect anything.
Questions that come up
How many sources is enough?
Two that are genuinely separate beats ten that are not, and this is not a close call. Once you have two independent captures that agree, more copies of the same lineage add nothing you did not already have.
Is a well known site more trustworthy than an obscure one?
Not for this. Reach and accuracy are unrelated. A large site with a stale list is confidently wrong in front of more people, and its size makes other pages copy it, which is how a single stale entry becomes an apparent consensus.