At time of writing, these are the top 10 markets on a mix of near-certainty and liquidity. How many of them will resolve as expected?
/MartinRandall/will-russia-invade-finland-or-swede (expected: NO)
/Tripping/will-mike-pence-win-at-least-one-st (NO)
/TheSkeward/will-google-acquire-twitter-by-eoy (NO)
/TenShino/will-ukraine-be-invaded-by-poland-b (NO)
/Tripping/will-the-little-mermaid-2023-gross-f862668cf378 (NO)
/Tripping/will-indiana-jones-and-the-dial-of-739c6fcaee5f (NO)
/LivInTheLookingGlass/this-question-will-resolve-positive-8a54be80b7e4 (YES)
/Quate/will-kevin-mccarthy-serve-as-us-pre (NO)
/Odoacre/will-proof-emerge-that-the-world-is (NO)
/chrisjbillington/will-the-sun-explode-before-2025 (NO)
This is a derivative market; it will resolve exactly according to underlying market resolutions. N/A is not expected for any of these markets.
One of the neat things that some bookies and odds-makers offer is long-shot bets at useful liquidity. You can get 1000:1 returns on weird stuff if you want; it's fun and informative. Fundamentally, they do this by offering lots of those bets, and assuming that not too many of them will go wrong for the house.
Presumably one could write an arb bot that, given adequate liquidity, arbitraged this question against the underlying ones. This might allow normal users to bet on questions like this, thereby allowing normal users to bet on all of the underlying questions at the same time. This would allow a bet on "nothing surprising happens" to have a slightly more reasonable rate of return. Hopefully these questions aren't so correlated as to cause problems.
If the consistency of the resolution was logically enforced, perhaps it could even be a house bot that didn't have to supply all the capital for its bets, since some of the arb is risk-free.
Related questions
@Eliza Where should the target for the "all" option be? More like 80%? Or anywhere between 10% and 90%?
@Eliza Excellent question! Seeing as I'm the one with the limit order at 97%, I think it's fair to say this was the expected result. (I didn't put it there right away because of a combination of being short on available balance and curiosity as to what other people would do.)
I'm going to make a couple more of these markets and then write up some thoughts on how I think they should / can work in more detail. But the real answer is going to be "try a couple options and see".
I guess my main issue with seeing 97% on this one, is it is tough for me to tell at a glance if Manifold thinks this market is equally likely to resolve to the 10 option, as other individual normal Manifold markets of 97-99% are.
Some of these are probably actually 97-98%, and others seem to be actually above 99%:
/NuñoSempere/this-question-will-resolve-positive-5c753f5a33e1
/ACXBot/27-will-elon-musk-remain-owner-of-t
/ahalekelly/will-vladimir-putin-still-be-the-le
/Ernie/will-the-sbf-trial-last-into-2024
If the percentage for the most popular option was more like 90%, I would have some more confidence that it wasn't a case of "This rounds to 100% but Manifold has trouble showing that". As it stands, I'm still thinking "Hmmm, which of these has a percent or two chance of being No, or is it just hard to push the one option high enough"
(If every Manifold market that was 99.5% actually sat at 99.5%, I would not have this issue.)