Realizing multi-party bets are still a hard thing to pull off in a non-custiodal way in bitcoin, especially at scale.
One could do a DLC, but then you have a large amount of data has to be communicted between each participant before the transaction is broadcasted. Not to mention the fees would make small bets (think less than a dollar) not a viable possibility. The amount of data grows nearly exponentially as the number of participants grows, there are some cool math tricks there to try to shink the amount of data needed for determining the payout path, but this doesn't seem like it would work for say a 1000 person competition style bet.
One could go the hodl invoice route of having each participant pay invoices for all possible payout routes then have a central node where all the payments go through funnel where the payments go at bet completetion. The node would cancel the payout routes that lost and allow the winning ones to pass through. But this brings in a lot of issues with lightning state and if you're bet is a long window of time there's a high likelihood of things going sideways (think force closures along payout routes).
It seems there might be hope here to do this in a less custiodal way via fedimint? The prediction market module seems interesting, but not sure how it handles orcale type data, ie an attestation about the outcome of whatever is being predicted/bet on.
Going to keep thinking on this, if anyone has ideas around this I'm all ears.
tee8z
npub13t…rhwpt
2024-01-20 15:16:39
Author Public Key
npub13tn44ckhtf76rmg32us9njawye6wa0rlkt975c9c8kv0rstzy9xs9rhwptSeen on
wss://relay.damus.ioPublished at
2024-01-20 15:16:39Event JSON
{
"id": "d4b8443abbe9a2d77deba8818e1984aeeab2ce3f9874843f8462a6cb17b4a101",
"pubkey": "8ae75ae2d75a7da1ed11572059cbae2674eebc7fb2cbea60b83d98f1c162214d",
"created_at": 1705763799,
"kind": 1,
"tags": [],
"content": "Realizing multi-party bets are still a hard thing to pull off in a non-custiodal way in bitcoin, especially at scale.\n\nOne could do a DLC, but then you have a large amount of data has to be communicted between each participant before the transaction is broadcasted. Not to mention the fees would make small bets (think less than a dollar) not a viable possibility. The amount of data grows nearly exponentially as the number of participants grows, there are some cool math tricks there to try to shink the amount of data needed for determining the payout path, but this doesn't seem like it would work for say a 1000 person competition style bet. \n\nOne could go the hodl invoice route of having each participant pay invoices for all possible payout routes then have a central node where all the payments go through funnel where the payments go at bet completetion. The node would cancel the payout routes that lost and allow the winning ones to pass through. But this brings in a lot of issues with lightning state and if you're bet is a long window of time there's a high likelihood of things going sideways (think force closures along payout routes).\n\nIt seems there might be hope here to do this in a less custiodal way via fedimint? The prediction market module seems interesting, but not sure how it handles orcale type data, ie an attestation about the outcome of whatever is being predicted/bet on. \n\nGoing to keep thinking on this, if anyone has ideas around this I'm all ears.",
"sig": "b46bc6ecfc04cc05d0523b7af5dff813a8016f99f41446fb19927da2ba9fea9dc27b5d352ac37610f46b4682cde523f330c41802f0a9e731ba98d36368bde07c"
}