Repost: Under what conditions is it ok (morally/philosophically) to bet on my own markets with an expectation/hope of profit?

Basic

Ṁ275

resolved Jun 26

35%29%

If you are explicit about what your intents are (example: "self-calibration: bet at your own risk", "self-incentive: buy NO to movitate me", "I will not bet on this market") and respect them

31%42%

As long as the resolution criteria are clear

31%16%

If the outcome is not under your control and you don't have insider knowledge of the outcome and the resolution criteria are objective and you resolve honorably; OR if you are transparent about your intentions

2%6%

All the time

7%Other

0.0%

“All things are lawful,” but not all things are helpful. “All things are lawful,” but not all things build up.

(See my offer on the previous market to refund my mess-up for anyone who made a bet there) Reposting without “poll” in the title this time. I’ll resolve to answers depending on how compelling I find them. I will not be betting, although I may add answers and immediately sell my shares so that they’re “on the board”. Jun 18, 9:35pm: previous market here https://manifold.markets/Angela/poll-under-what-conditions-is-it-ok

This question is managed and resolved by Manifold.

Get

1,000

and

3.00

15 Comments

10 Holders

16 Trades

Sort by:

Another aspect to consider that I just thought of: the author is a liquidity provider, and being a liquidity provider also means you have a stake in the market, just like being a trader. If a market trades up to 90%, then you as a liquidity provider end up with a substantial number of NO shares, and you'd profit by resolving NO while you'd lose M$ by resolving YES. So the author always has some amount of stake in the market, creating some amount of potential bias. I'm not sure if there's any strategies to mitigate that. Normally this is all kind of hidden which helps, but if you inject additional liquidity and then withdraw it, it becomes more obvious - you end up holding extra shares in your portfolio just like a trader does.

This is mainly about cases where e.g. the resolution criteria are ambiguous and the author wishes to avoid biasing themselves.

Some possible solutions: - Allowing for the author to create the market without providing liquidity (either it's provided by someone else afterwards, or two users collaborate to create the market) - If the author can choose to give up ownership of their liquidity pool shares. Then they put in M100 initially, and its ownership is transferred away to the platform or something. That leaves them with no stake in the market. - If you can delegate resolution to someone else. I think this would be a useful feature in general. Doesn't work for markets on personal questions though.

Also, a lot of those issues don't just apply to the author, but potentially to all traders. Ambiguous resolution criteria only applies to the author of course. The insider trading / manipulation of outcome ones can apply to other traders too, but often the author is the most likely to be in the position of an insider.

I think these are the main potential issues with betting on your own market - all of them are ok as long as your intentions are sufficiently clear: - The resolution criteria are ambiguous - your trading could bias your resolution. This becomes a tradeoff: your betting can improve the prediction accuracy, but create bias in case of ambiguous resolution. If you are transparent about it, then traders can bet at their own risk. - You have insider knowledge of the outcome. Personal markets are fine as long as that's clear. But if I created a market "guess what number I'm thinking of" I think it's implicit that I shouldn't bet on it myself unless I explicitly state otherwise. - The outcome is more under your control than people should reasonably expect. Motivational/commitment markets are fine - everyone knows that the outcome is highly under your control. If I made a market "How much will someone donate to Manifold for Good", I think it's fine for me to influence the outcome by donating and also bet on that, because people are aware of this possibility. But if it's possible for me to influence or abuse the resolution in a way that seems dishonest (e.g. hacking the number displayed for resolution), and I bet to profit off of that, that's probably bad (unless I'm up front it, e.g. if I say it's anything goes as far as manipulation of the outcome). - Also, for free response (parimutuel) markets, I think the author probably shouldn't bet on the outcomes they're choosing before resolving (unless other traders could have equally well done the same), because that actually takes away the profits of traders. But ultimately Manifold should fix this problem by changing them to fixed-payout markets (as yes/no markets are).

From a combination of @JoyVoid and @MartinRandall 's answers on the previous market

(Edit: On self betting, I allow myself to bet and sell a dollar to bump the market)

@flatmap I would say all the time, except the author benefitting from being the only one to know how they're going to draw the line on a vague resolution rubs me wrong.

My only potential quibble is if "if you're explicit" is read as "iff you're explicit". Because I think it's typically self-evident that it's fine for the creator to have carte blanche on betting. But can't hurt to be explicit and, again, your policy ontology sounds awesome to me.

@JoyVoid those are great! nice work!

@JoyVoid agree entirely. Doesn't make sense to have a blanket policy on self-betting when there are so many different kinds of markets. The important thing is clear creator communication with potential bettors, and honesty in that communication. At that point, it's up to any individual bettor whether or not they want to participate.

Since we don't have maniblogs yet, I'll use this comment to detail my new policy when opening a market so that I can link to it on them. When opening a market, I'll append to it at the end "Author betting policy: [Policy]" to say what my intentions are. Those are: - Betting as usual. Will mostly use those for market that are external to my condition and where it would be impossible for me to inside trade. I'll just bet on it as normal as if it were any other markets - Not betting. I will not bet on this market at all. If the criteria are external to me, I might provide updates as usual. If they are internal (like: "What game will I enjoy"), expect little updates before resolution - Self-calibration: Bet at your own risk. This will be used for markets I want to know the answer of, and want to try to predict as best as I can, but which are personal to me and so I have way more information. Expect frequent updates whenever I change my mind. Will probably first post an update to let some time for people to adjust their bet before betting on it myself (I'll tentatively try 12 hours for now, might change it it doesn't work great) - Self-incentive: Can only buy up in the prefered direction (So like if the question is "Will I finish this project", I can only buy YES). Cannot sell it and cannot buy the other share. This is used to motivate myself, you can encourage me by buying NO or just bet as usual. Also expect frequent updates. I might add more categories if at market opening I feel that I missed one

@dreev a case where that does not apply i think is in the case of self-motivating market. For instance if you ask "will i exercise 30 minutes a day on average", I'd be doubtful if you start betting no. Though I can understand the desire for self calibration, but I think you need to commit ahead of time about whether you intend to buy NO shares at all

As I said in the other market, I'm in favor of a norm for authors to declare intention of betting. I think Manifold could facilitate this through its design

I think this is basically correct. The market probability being accurate trumps all other concerns! Just be nice and transparent in the description and in the comments. And it may sometimes be more fair to let other traders know why the market probability is wrong and give them the chance to move it first, but a few minutes is enough. Then go ahead and move it yourself. I'd say that's the first rule of Manifold: the market probability must be correct. Monkey wrenches are possible if the resolution criteria are subjective or if there's market manipulation. But insider trading is mostly fair game.

Related questions

Related questions