two notes with the same content published at the same time, one has gotten 10x more attention than the other. can you spot the difference? https://blob.satellite.earth/53b3eec9ffaada20b7c27dee4fa7a935adedcc337b9332b619c782b030eb5226
Discussion
Yep I can!
Yes . Broadcasted to more/different relays .
nostr.wine ?
@fiatjaf is a Coracle.social user ??
One has zaps the other one doesn’t
You used nostr.wine relay on the one with more engagement.
I see. What's the problem？
Freerse is working normally
One went to nostr.wine, which tells me people who pay for relays are more likely to know who you are and engage with your content.
I think this is a valid case,
Yes, maybe.. is it: 🍷?
I saw them both at the same time and just randomly chose to engage with the top one.
Eyy all wise Fiatja, are you trying to teach us the recency bias now
No discernable difference from the image.
People engaged w/ the first one, then assumed the second was a dual-post, as is common.
The relays published to are different.
Easy: first fiatjaf is a chick, second fiatjaf is a dude.
wow that Web of Trust scoring is so high for you!
How do we know people simply didn't feel the need to respond to the 2nd note because they already saw / responded to the first? honest question.
You seem like someone who understands what a null hypothesis is
What my have been more effective is to say "Please respond to this note" on the 1st, and "please respond to this note as well, it's not a duplicate"
The fact that they look identical is enough reason to ignore the 2nd note.
Challenge accepted.
This one, not:
The former is visible ALSO on relay.nostr.band, the latter is not.
Further:
- The former is 7 seconds ahead of the latter.
Aside from that, .id, .sig and .created_at are different, as expected. So, the only true difference is: relay.nostr.band.
That is actually mindblowing.
You are too late. At this point they both got rebroadcasted into many relays. In the beginning one was just in my pyramid relay and the other in the 3 top most popular relays I found. In 5 minutes one in the top relays had 10 times more interactions than the other. After that the test is skewed by many other factors like rebroadcasting and the fact that Jack reposted it and so on.
I've repeated the test in my next note after that though, and the results were the same.
Interesting. This is giving me quite food for thought as to how this could be improved - but it also shows that nostr.band certainly does what it was ment to do (as far as I understand).
I wonder what would happen if nostr.band were to go under for whatever reason (spontaneous implosion of the infra or something) - how much of the nostr apps out there would outright break?