🏅 Top traders

In principle I support a path to unban, but it'd have to happen after a period of good behavior (i.e. not using the website / not using the website outside of a single account you're authorized to use), and that period should be longer than a month or two

Probably shouldn't be unbanned quite yet given they've continued to use the alts in this market

release Relay

Making alts for the free mana is against Manifold rules. It would've been okay if he purchased the mana. I think we wouldn't have banned him if he had borrowed the mana from another user either.

Relay legend

Y'know, he turned out to be right. Could be a valuable asset to manifold's calibration. 😅

I swear i didn't plan this but apparently it is allowed

@Joshua 😂

After discussing in standup, we've decided Aella's market will not be marked nonpredictive.
We will not be doing a partial N/A as such functionality does not exist and that would require too much dev resources to create.
We will do our best to make sure that Relay does not benefit and is punished for their alts.

We do not approve of sockpuppeting and fraud. It is important for community trust to stop it. Ideally we would prevent it from happening in the first place, for instance via better detection of automatically generated accounts. That said, we need to balance that against dev effort and against the new user experience for real users.

Fining/banning someone who just signed up for the site doesn't really do anything, does it? He's stopped making alts, but he could continue to do so if he wanted. The biggest damage he did is to the leaderboard and league standings which this does not address.

@Joshua I don't see any other good options, do you?

the ideal option would be to N/A the last 16/however many hours of the big market. Absent the code for that, I'd manually zero out the profit + shares of all big YES holders and (if substantial) the profit of whales who market-made, and then try to slap together a technical solution to stop more alts from being created

(note that this was my position from the start in the disc, and isn't a response to new worries about how it resolves)

When all of the profit is coming from alt accounts that are breaking the community guidelines, I think that the least that should be done is marking the market non-predictive so League standings are not affected.

I think the biggest issue is to discourage future action of this type, and the way to do that is to minimize the impact it has. Which means removing the large amount of free mana (nonpredictive isn't enough) and removing the ability to do it again

@jacksonpolack The users who are abusing the site aren't getting the free mana tho afaict, unless the market resolves NO.

giving non-abusers free mana is bad too! it encourages future abuse where the "non-abuser" coordinates it, and it contributes to mana inflation. (I think it's going to resolve yes). This was the opinion of most of the YES holders i spoke to in discord earlier. It's not satisfying to compete over 'dirty mana', and we don't want to do that, but kind of have to to stay competitive

this is also what would happen on a 'real market' if someone injected a bunch of dirty money into it, i think

If someone used an infinite mana exploit to print a million mana and then put it all in a limit order at 1%, the first person to see it would buy the limit order and then have a million mana.

That person buying the limit order has done nothing wrong, but they shouldn't get to keep the million mana and it should't disrupt all of the markets about Leagues and Leaderboard standings.

@jacksonpolack but the non-abusers aren't the ones perpetrating the abuse, and I don't consider them particularly suspect for abusing the site in the future just because they got some mana out of the abuse.

@Joshua This is very far from 1 million mana

If I were a low-trust person, I would get on a vpn, pick another market, make 100 google accounts and do that for a second market, and then 'participate in' eating up the surplus again. I would pretend to be the same person as the first guy. It's important to take action against potential actions like this before they happen, people notice when you act in exploitable ways and will take advantage of it.

It's not a million mana, but it's enough to distort league standings that people spend a lot of time on. We've been playing a game of chess and this guy has come in and knocked the table and a bunch of pieces fell off.

The right thing to do is set the pieces back up where they were, not keep playing with the pieces on the ground out of the game. And this is especially true when people are betting on the results of the chess game.

And to be clear, I'm profiting from all this. I'm arguing against my own profits because I want a fair game.

that's just the market working as intended. lots of people shot up to top of leaderboard after Isaac's pump in whales v minnows. sometimes people will just be really wrong, and that's that.

you could say it's a fundamental weakness of markets relative to polls at determining "who is the best at predicting". but the disproportionate reward is necessary to keep markets good at predicting the truth, because the incentive is a correction mechanism

it is not the market 'working as intended' when someone prints free money and dumps it into a market not intending to profit off of it. Especially when you're manipulating the price to influence an offsite market.

@jacksonpolack no but the people buying YES are making the market work as intended.

All this moralizing about "dirty money" actually makes the attack more attractive to the attacker. If you just bought YES and corrected the price then there wouldn't be much of an affect on Polymarket's market

@Sinclair I agree that the all-time leaderboards are already skewed by Whales vs Minnows, but Leagues isn't. Marking the market non-predictive solves the short-term problem of all the markets about leagues, and then you can decide what to do about the leaderboard as a whole later.

All this moralizing about "dirty money" actually makes the attack more attractive to the attacker

With my proposal, the price goes back to 85 as soon as someone notices. The price is currently 64. People (including me, admittedly) were actively coordinating to keep the limit price down to increase profit, and would start doing that again if he came back.

@jacksonpolack I think your proposal is heavy-handed, and the actual way to get back to probability accuracy is to ... buy in the accurate direction

Long-term, I'll point out that the biggest lesson here for me is that new accounts should definitely not be able to bet their full 500 mana on one question. Hopefully fixing that would go a long way towards preventing incidents like this.

Long-term, I'll point out that the biggest lesson here for me is that new accounts should definitely not be able to bet their full 500 mana on one question

They can just bet 100 each on 5 liquid markets, sell it all, and then reinvest. Or wait a few days to get all the mana, etc. That's something of a bandaid

@Joshua Restricting new accounts to making a maybe max 20 mana bet in a single market for maybe a week would work fine, I think. It also has the upside of teaching new users to diversify their portfolio and could also help with user retention ("I put all my money in a single market and lost -> I leave" vs "I lost some money in this market, but I'm still waiting on some others to resolve -> I stay for now").
@jacksonpolack I think if the initial amount that can be invested in a single market is low enough there won't be enough of an incentive to create new accounts to manipulate a market. But yeah, it's not a perfect solution.

Prediction markets are inherently "unfair" in that some predictions count more than others. One market might have M$100,000 in volume and another M$1000; is it "fair" for your record in one to be scored with more weight than the other? No, but that's core to how prediction markets work.

@IsaacKing agreed. it's not as if the spanking market had a TON of traders before all of this. we're showing up for the liquidity and market making, which is always happening in bigger markets. it's an expectation that people do that around here. imagine marking LK-99 non-predictive because people were doing live trading