Inspired by the varying probabilites of different markets on the superconductor questions.
Resolution criteria: resolves to MKT unless people who aren't trustowrthy make bets that are too large or suspicious in which case it'll resolve to a poll or consensus among commenters
1,000
3.00edit: never mind I misunderstood the new rating system
@jacksonpolack I wouldn’t care too much about a 5.0 rather than 4.9 average rating. You seem to be doing fine!
I trust Manifold slightly more than Metaculus on all markets without significant counter party risk (and >20ish traders). I trust it more than Polymarket at all times.
I trust Metaculus more than Polymarket, and more than Manifold where there is significant counter party risk. It does better though at displaying (or not displaying) how confident it is, and there are less perverse incentives around questions with significant counterparty risk (AI doom, etc).
I trust Polymarket the least as it has a historically mediocre record, and mechanistically its transaction fees are just too high.
@RobertCousineau Mostly agreed, but I don't think Polymarket is that bad haha.
Re Polymarket fees - when they rolled out the central limit order book, that basically removed transaction fees. I think now there are just extremely tiny transaction fees for Polygon operations (unless I've missed something)
@jack Haha, I did bet them up from 13% - I don't totally hate them. It sound like you have more up to date info about fees than me too - it has been a year plus since I last seriously looked. If they are better now, that is super awesome.
@jack Polymarket has zero fees on its markets currently (they sometimes add a 2% fee to a market - so you need to pay attention and/or join the discord).
Closed a bit early to prevent people from whaling in at the last minute. Does anyone object to resolving Manifold 23% / Polymarket 17% / Metaculus 60%? The movement towards metaculus highest seemed pretty organic to me.
Not objecting, but just my personal opinion: If I had noticed that movement I would have bet them closer to even, I think they are all very close (as per the analysis I posted in https://firstsigma.substack.com/p/midterm-elections-forecast-comparison-analysis) - I would have put them within like 20% or so of each other.
@jacksonpolack As a heads up, I was not aware you were resolving to market - I figured it would be winner take all (which is why I only placed a limit order).
It's a bug - I edited that in 6 days ago, and as proof I referenced the edit in discord here https://discord.com/channels/915138780216823849/938171760237477998/1134396665797804042
the mesage is dated jul 28
@jacksonpolack I do object. The initial whale was done by an arbitrage bot betting 100 mana. The bot voted in favour of your personal preference.
Resolve to Market bets are all momentum bets and it would've unfairly given momentum to people who already aligned with you and dissuaded people from voting against the momentum because who likes losing? Even fake money?
I'd say this should resolve as N/A.
Alternatively, if you really want to compare, please make three markets. "Do you trust X" , "do you trust Y" and "do you trust Z", and use the percentages in the end to compare across these. Don't setup a meta market that ensembles these individual markets.
This allows people to individually commit to the trustworthiness of the market without gambling on the meta outcome.
This will get you the results that you want. Certainly better than now.
The initial whale was done by an arbitrage bot betting 100 mana.
The way that bot works is it bets back against big market moves. You bet a lot of M$ on an option, so the bot bet a bit against you. So if anything you were the initial whale lol
Resolve to Market bets are all momentum bets and it would've unfairly given momentum to people who already aligned with you and dissuaded people from voting against the momentum because who likes losing? Even fake money?
Eh, there are a bunch of small bets on metaculus, enough to suggest that people do like it. If people didn't have a preference, it's more profitable to bet everything to 33%
One data point: analysis of the 2022 midterm election forecasting
https://firstsigma.substack.com/p/midterm-elections-forecast-comparison-analysis
