0 sats \ 5 replies \ @SwapMarket 7h \ on: Bet on the US election with sats on bitcoinprediction.market bitcoin
What are the formulae to calculate the odds and the potential return %?
We're running a parimutuel market setup at the moment (winners split the losers money in proportion to how much they bet), but we want to move towards a modified version of LMSR. LMSR is a better forecasting tool, as long as you set the liquidity parameters correctly.
When a user bets we basically calculate how much their proportion of the winnings are if they won without any changes to the liquidity. So like
odds = (sats bet yes)/(sats bet yes + sats bet no)
potential return = (your sats bet yes)/(total sats bet on yes) * (total sats bet on yes + no) or (your sats bet no)/(total sats bet on no) * (total sats bet on yes + no)
It's not the best system, LMSR is better, but it's good enough for a short-term auction (akin to horse racing).
The reason we want to transition to an ecash system rather than just custodial lightning is because that would enable people to sell their shares before market close (currently we cannot facilitate that with just Lightning payment & address). Then people could bet -> receive ecash tokens (representing shares) -> download or save those ecash tokens. Then whenever they want to sell or redeem the value of their shares, they can upload those tokens and enter in a receive address (lightning or on-chain), and then the market maker could send sats to them.
Another reason we're going for ecash is because it is effectively an authentication mechanism for people without needing to give an email or npub or signing up at all. We'd much rather avoid the UX of signing up to place a bet, and rather enable anyone to be anonymously, as long as they have access to a receive address in the future.
Then we could list the bets with just bet amount and position, and we don't even need to reveal the lightning addresses while keeping everything publicly auditable.
reply
Do you not deduct any fees?
It's just weird that odds for Trump and Kamala add up to 108%. But it does not mean that betting on all outcomes will make a profit for this guy:
Did you remove the link to the stats?
reply
Stats are listed below the topics. Click on the title of the topic to go there. We thought it would be better to show all the stats below the topics, rather than on a separate page.
We show the bets placed, but the fees are deducted when we send the payouts (1% fee at the moment). The total odds add up to 108% because it is basically
Odds_total = Odds_Trump + Odds_Kamala = Y_T/(Y_T+N_T) + Y_K/(Y_K + N_K).
Both markets are independent (betting on Trump does not change Kamala's odds, it only changes Trump's odds), so they are not necessarily supposed to add up to 100%. For a clever user like yourself there's an arbitrage opportunity there to balance the odds. Few people have spotted that so far.
reply
from ChatGPT:
How Arbitrage Works:
For arbitrage to exist in a betting market, the combined implied probabilities of all outcomes must be less than 100%, allowing you to bet on both outcomes and lock in a profit regardless of who wins.
For example, if Trump were at 60% and Kamala were at 35%, you would have a combined 95% probability (less than 100%), and you could profit by betting on both sides. However, in your case, the total is over 100%, so no arbitrage is present.
reply
deleted by author
reply