If they disable it, that counts as NO resolution.
Disable means: People can not give ratings for more than 3 days and this is not due to the site being down.
🏅 Top traders
# | Name | Total profit |
---|---|---|
1 | Ṁ177 | |
2 | Ṁ167 | |
3 | Ṁ89 | |
4 | Ṁ86 | |
5 | Ṁ79 |
@firstuserhere Manifold search has foiled me again. In its defense using the emoji instead of star is not very SEO friendly.
Current plans for the feature:
https://manifoldmarkets.notion.site/Star-reviews-484a46d632474bd79f1c2b401b6b2ad7
There are also plans to disable the feature until it is better:
A dev called it 'worse than useless now'.
But since there are no resolution criteria on this market I don't know which way to bet even though there is all this information. I would rate this market 2 stars at most.
@firstuserhere Yeah I think that's a good idea, I posted 3 ideas in discord and then later made this market:
@Gen My main concern is that by having a 1-5 system, it's basically always going to be 1s and 5s. Even if someone runs a market "imperfectly", I can't see myself giving them a 3 or a 4 without feeling bad, especially if I know someone else has given them a 1 because of the "imperfections" already.
@Gen FWIW I'll definitely use the full range. There's a comment box so you can tell them what to improve, so it can be constructive.
@firstuserhere @chrisjbillington Are you guys disagreeing with me? 1 star
I don't know much about review systems and what works, but I imagine letting people go back to old markets and review them retroactively is probably a bad feature - even past a few days. Amazon most likely have the best research so they're a good model to copy (e.g. only let people review something they've actually put mana in, make them rate it immediately)
It might also be worth considering weighing reviews from people who profited from markets heavier than people who lost (or at least, displaying if the reviewer profited or lost), if the goal is to demonstrate "This guy can be trusted to resolve correctly" or not. I can imagine people getting 50 1 star reviews salting their stats if someone is mad about being resolved against.
@Gen i mean, on amazon, a display of what % of people voted 5 stars, 4 stars, 3 stars and so on is what is helpful to me. I'm able to see that oh 75% people voted 5 stars and only 4% voted 1 or 2 stars, with the rest divided evenly in 3 and 4 stars. I like that product's ratings in this case. More helpful to read 4 and 3 stars reviews than 1 or 5 stars
@firstuserhere this simple bar graph (which is very easy to add, just hover or click, and you'll see) of rating tells me more than the 4.4 out of 5 does. Also, the reviews are currently sorted so that the latest reviews are at the bottom, that should change - latest ones on top
@firstuserhere and i dont even see the reviews immediately, a simpler UX. I see reviews only if i want to
@firstuserhere Yep, I can see that. Great point. Encourages ratings of 2, 3, 4 and gives a clear picture straight away if someone can be trusted. A lot more information is displayed from the same dataset.
@firstuserhere @Gen I think its helpful to think of a market creator as a "bookie". You go to the bookie, he offers you odds, and you bet. This makes it easier to think about what markets you should invest in aka, which markets should you invest in based on what the bookie tells you about the odds. On subjective markets about bookie whose odds are sold to you by the bookie -> not a great idea to go all in, for example.
@Gen "It might also be worth considering weighing reviews from people who profited from markets heavier than people who lost (or at least, displaying if the reviewer profited or lost), if the goal is to demonstrate "This guy can be trusted to resolve correctly" or not. I can imagine people getting 50 1 star reviews salting their stats if someone is mad about being resolved against."
I'm unsure if that would actually make the ratings less-biased. Just like salty losers, people who made mana will tend to feel pleased with a controversial resolution. I do agree there's probably some asymmetry, and losers probably react more negatively than winners react positively.
But OTOH, if even people who lost mana agree with the resolution method, that seems like a more credible signal about the market maker's handling, than a mana-winner saying it was great. Similarly, if somebody won mana but gives the handling a worse rating, that also seems notable.
Not that it's necessarily worth it for Manifold to try factoring for this, I'm kind of just spitballing here. But anyway, I'm just not confident that putting higher weight on winners' ratings is necessarily better. That sounds like it would lend itself to incestuous / rent-seekish behavior.
@ScroogeMcDuck Yeah, I should have said, weigh negative reviews heavier from people who won, and vice versa. Idk how it would work. I think in retrospect it is a ridiculous concept / total waste of time.
The three more serious suggestions I gave I have noted here:
@firstuserhere oh okay. i thought it looked weird. so once it completes, opens up for review i assume?