Is manifold fair? (Closes on March 1, 2024)
Basic
150
Ṁ11k
resolved Feb 29
Resolved
NO

The most bet option by March 1 wins.

Each user is encouraged to share your own arguments why Manifold is or is not fair.

I'll start with a couple of arguments against it:

  1. Users accumulate mana over time. This liquidity gives them excessive weight when betting.

  2. The outcome of a market should be measured by the number of individual users who voted for a specific outcome, not by how much they bet.

This market resolves TRUE if the percentage at close is greater than 50%

This market resolves FALSE if the percentage at close is lower than 50%

This market resolves N/A if the percentage at close is exactly 50%

Get
Ṁ1,000
and
S3.00
Sort by:

🚨 Warning 🚨

This market is whalebait, not based on the result of a poll. Trade at your own risk.

Ignore everything it says about "votes" -- there are no votes.

(The description has been updated to help clarify the situation.)

opened a Ṁ3,000 NO at 30% order

@SimoneRomeo Resolves no

"The outcome of a market should be measured by the number of individual users who voted for a specific outcome, not by how much they bet."


The outcome of markets isn't measured by how much people bet, unless you happen to create a self-resolving market that explicitly says it will resolve based on percentage at close. As you have done here, but that was your choice to make, not manifold's.

Self-resolving markets are bad but that's not manifold being unfair, that's people choosing to make bad markets.

Ignoring the fact that this is just whalebait...

  1. Users accumulate mana over time. This liquidity gives them excessive weight when betting.

^ That's the point. It's quite fair to assume that people who make good predictions should have more weight on future predictions. It would be unfair if people who consistently made bad predictions had the same weight towards a market as people who consistently made good predictions.

  1. "The outcome of a market should be measured by the number of individual users who voted for a specific outcome, not by how much they bet."

^ This would (1) completely defeat the purpose of prediction markets and (2) only be useful for predicting a binary outcome without giving a probability, so I'm not even sure this needs rebutting.

🚨 Warning 🚨

This market is whalebait, not based on the result of a poll. Trade at your own risk.

Ignore everything it says about "votes" -- there are no votes.

(The description has been updated to help clarify the situation.)

predictedYES

@Eliza what are you saying? The description clearly states how the market is going to be resolved

@SimoneRomeo "the most voted option by march 1 wins" makes it sound like the market will resolve to the option that has the most people holding shares of that option at market close, imo. i think that's what eliza was referring to

Where is the poll?

Please add the link to the poll to the main description, I can't find it.

predictedNO

@Eliza It's whalebait

@nikki It says "votes" multiple times. If there is no poll I'm tempted to unlist it because people won't understand.

predictedNO

@Eliza

This market resolves TRUE if the percentage at close is greater than 50%

This market resolves FALSE if the percentage at close is lower than 50%

This market resolves N/A if the percentage at close is exactly 50%

@nikki Sigh. I thought it was talking about the result of a vote.

predictedYES

@Eliza you can vote by betting yes or no here

@SimoneRomeo Betting is not voting. After nikki pointed out the criteria was referring to this market and not to a non-existent poll, it's obvious this is just a self-resolving market.

If there's no poll and the market resolves only based on the closing value of the market, it doesn't matter what the question is -- it may as well be asking "Am I Donald Trump?'.

Read this for more info:

https://manifold.markets/old-posts/selfresolving-markets-why-they-dont

I have to warn the participants because unwitting users will walk right into the whalebait without realizing it and be sad when they lose their mana.

predictedYES

@Eliza you could as well advise them to read the description. The resolution criteria are clearly stated there. It's indeed a self resolving market. And indeed I don't think it's fair. It was kind of the point of this whole market to make people realize that.

predictedYES

@Eliza I see your point regarding betting and voting. I changed the description to avoid being misleading

Offering liquidity on this market for up to 30% YES and 70% NO

“The outcome of a market should be measured by the number of individual users who voted for a specific outcome, not by how much they bet.”

Wait, you could fix that for this specific market. You could look at the number of people betting on Yes vs No, instead of the final percentage.

predictedYES

@SantiagoRomeroBrufau that would indeed make it more fair. But I already reply to this question and agreed the final percentage counts

predictedYES

What is fairness? I can think of a few conventional meanings of the term.

There's what you might call "rules fairness" - do all participants play under the same rules, or do some get different rules or get away with cheating?

Manifold is clearly ~fair on this basis, as opposed to e.g. an election in a dictatorship.

There's what you might call "ergodic fairness" - does the system converge over time to a result that is not particularly path dependent? Monopoly is the canonical anti-example: it's rules-fair but not ergodically fair, an early winner tends to snowball.

It's a contentious point but I'd argue manifold is mostly ergodically fair. Some advantages exist for tenured traders - they can kelly-bet bigger, take more small wins that require a lot of capital, etc. I think these are outweighed by the risk/reward ratio: a small trader can double their net worth in a few good trades, but a large trader cannot, so it's not too difficult to catch up somewhat.

The last sense I can think of is "closed system fairness" - is your success or failure determined by a well-specified and demarcated set of inputs that all participants can equally access? Pay to win games are unfair on this basis. So, deliberately, is manifold, by allowing insider trading. (I don't think buying mana matters to this point, because of the ergodic point above).

predictedYES

There is a reason I bet over 600 on yes

The fact that some thing make this place a good (convergent) prediction place do not make it fair. "Fair" is not the same as "good market".

A person with more mana has more opportunities, wider portfolio, bigger absolute gains. Thus anyone who started earlier has an advantage.

Any game which makes strong players stronger is not fair. A good (most saturated) example to this is counter strike wcs: with each thousand of kills you get a new superpower (vampirism, +100 armor, +health, flight, xray vision, teleport, meganade and so on). Imagine coming into the server when everybody has already 53k health and you, with your 100 health have to somehow do 1000 kills to progress. Even if you are s1mple, your progress will be much slower.

"Of you are better predictor, then bet where your mouth stands" - newcommer, even if he sees some market is wrong by 75%, cannot bet everything he has, because it will mean a risk of balance falling back to 0. Diversification is needed even for perfect predictors, and that will slow them down.

Let's assume that in such game people start simultaneously. "Two apples in a gravity field" problem appears: initial fluctuation in time of start will lead to steady growth of distance between apples, when they fall. Secong apple, experiencing the same acceleration, will never catch up, even if it later invents a way to give itself a slight speedboost. Value of such speedboosts decreses over time.

There is no yearly rank+mana reset either, which would at least regularly start a new race to see whether somebody of newcomers is actually better than olds.

(Metaculus recently implemented totally independed time frames for medals, which makes it fair for new people. Mana on manifold is not independed, it accumulates through all seasons.)

There are bonuses. Longer you play, more bonuses you've accumulated.

Fair games do not give advantage to stronger players.

Good fair games even use handicaps in non-rated matches to make the result of a match a 50/50 so you could see even the slightliest increase in sb's skill.

If you play chess against Magnus, nothing can be told about your or his skill improvement over time. You will lose all of matches (a whale will get bigger average absolute gains from all of markets). But if the starting position and pieceset is set up, that you win 50% of matches against Magnus, then when that % changes with statistical significance, you gain info that either you improved, or he fell weaker. Thus it truly shows your dynamic (after you have several such comparisons to others).

Out of mechinics of manifold only calibration shows you your dynamics (but it is hidden behind additional click). And even the calibration is weighted by mana and is accumulative stat, thus is easier correctable over time.

Nothing here is fair in the game-design sense.

It is only fair in juridical sense: you know the rules and you agree to play by the rules.

God only knows why such silly, low-effort questions get so many traders.

Fair? For who?

As a prediction market, I think it would be much more accurate if people betting on questions didn't know the percentages created by previous betters

Manifold is fair, this market isn't.

It's fair, but corrupt. Everyone starts at M500.

© Manifold Markets, Inc.Terms + Mana-only TermsPrivacyRules