Goal: Discussion
What's the best thing for a creator to do while they deliberate something about their market? How can the site help them do this while maintaining a fair trading environment?
Application
Any time a creator is faced with making a choice about their market. It can be: choosing whether to resolve it, whether something 'counts' in the market, tweaks to the rules, or dealing with surprising things that are not in the criteria.
Problem
If the creator leaves trading open while they are unsure how to rule, traders often start trading on "What will the creator decide" rather than "What will happen in the future". This type of trading can sometimes eclipse trading on the actual question at hand, and it can also influence the creator's decision-making process.
Current options
There are several strategies creators have used over the years, they seem to all have some upside and some downside.
Leave trading open: Make multiple comments and be open about your thought process. Traders respond to every comment. You do get the benefit of instant response to your ideas, but traders may feel misled if your initial comments do not line up with your final decision. The most common pitfall is creators frequently issue contradictory clarifications.
Leave trading open but don't comment: Just let everyone battle it out while you deliberate, and don't post about it until you've come to a decision. This prevents the creator from making multiple contradictory responses, but can leave traders just as upset if the final ruling is not in line with the way they read the criteria. For "Will X happen before Y"-type markets, if the creator does not respond to questions about whether event Z 'counts', the traders often eventually decide that no, Z did not count, even if the creator never issued a ruling about it.
Trade on it yourself: Sometimes creators who do trade on their own markets start trading when things are in question. Maybe that reveals their own thoughts?
Close trading pending decision: It is not obvious to a new creator, but they have the power to close trading at any time using the same window where they can change the closing time. By closing trading before initiating discussion, the creator can avoid disgruntled traders who were influenced by thinking-in-progress. There are some downsides: Reopening the market is tricky because there is often a price delta pre- and post-closing. Who gets to 'steal' that liquidity? What if the creator is not around fast enough to close trading and a bunch of people already made trades about "What will the creator decide"? There is also an awkward interplay with league scoring here once every month.
Discussion topic
So, is this really a problem that needs solving? Is it okay how it is? Should we be looking at better options here?
Here are a few possible 'helps' for the situation, not necessarily solutions:
Better UI/guidance for creators in this situation -- the site could give them advice on avoiding common problems.
More sophisticated close/reopen mechanism -- there have been several suggestions for a more fair market reopening procedure, including 'auctions'
Provisional trading mode -- let people trade while the market is in limbo, or some new 'not-quite-closed' mode, but those trades will all be N/A'd if the creator decides that it resolved at the time they closed it earlier
Yank the liquidity -- maybe "closing a market" is actually just removing the creator's liquidity, but traders can still trade with limit orders. The creator is usually the one putting up all the juice, let them withdraw all but 1 mana or something, until they've ruled, and then let them put the market back live with fresh liquidity at the price they think it should be at. Traders can still trade in the interim if they wish.
Got any additional good ideas? Maybe you think this is a huge nothingburger and not a big deal. I don't think this is the #1 problem facing the site, but could be a fun discussion.
Gödel's second incompleteness theorem.
No consistent formal system can prove its own consistency using only the rules within that system.
Applying this knowledge to markets context: No level of rules elaboration, policies, conventions will be enough to prevent the appearance of grey events, which were not covered by original version of the market description and/or its appendices.
If anyone feels really strongly about any specific way to help creators when they need to deliberate about a market, share here!
@Eliza Good to see this being discussed. Certainly not top priority, but it's indeed a problem.
I want to mainly point out that your "solution helps" might help those creators who are aware of the problem and willing to handle this in a better way. That's maybe 10% of creators.
Let me gesture at a wild idea:
Add a checkbox to the comment form stating "Cruxial for resolution, needs clarification" where traders can e.g. ask "If X happens, will this suffice for Yes?".
Allow traders to agree/disagree vote on the cruxial-ness.
If deemed cruxial, duplicate the market twice: Once with the added clarification "If X happens, will resolve Yes", once with "X happening does not suffice to resolve Yes".
All shares, liquidity, comments and orders get duplicated as well, except trader bonuses.
Immediatelly notify traders that there are now two succcessor markets.
Original market is N/Aed.
Creator must choose which of the two new markets to keep. The other is N/Aed.
As soon as the choice is made, the creator recieves trader bonuses for all future trades in the single remaining market.
Of course this would all need to happen automatically. I'm pretty sure this would facilitate new ways to cheat the system, also it's probably not technically feasible. Thus "gesture" and "wild idea".
@Eliza Market resolutions should have an “resolved as of” date field.
Bets before the “resolved as of” field are fulfilled, bets made after the “resolved as of” field are automatically refunded.
Moots this entire issue.
@Primer duplicating the market duplicates all bets? so if I get 5M in a market I’m suddenly in the hole for 20M if the market splits into four?
no way, the etherium DAO was hacked exactly like this in 2016
@KJW_01294 A duplicated market is just a copy of the original market's description, tile and close date.
your comments says all shares, liquidity, and orders are duplicated, so what do you mean? (Edit: sorry for misunderstanding)
@KJW_01294 It wouldn't make sense to duplicate all bets (that's essential adding in leverage systems)
@KJW_01294 Yeah, duplicate everything except trader bonuses. Yes, this creates Mana out of thin air, but only one of the markets survives, all others end up N/A.
@Primer oh I actually like that a lot, assuming traders don’t have arbitrage opportunities
@100Anonymous I'll call these "Quantum Markets". We turn 1 market into 2 virtual markets and decide upon observation of the creator's judgement which of them is real, the other ceases to have ever existed.
@KJW_01294 I don't see how there would be arbitrage opportunities, but if someone can find them, it's Manifolders.
It's important, though, not to duplicate the trader bonuses, as we
don't want markets needing those clarifications at all
don't want to incentivice creators profiting from intentionally setting up unclear markets
I'll call these "Quantum Markets"
I have certainly considered ideas along this axis in the past. As you noted, it seems like a leap to get there. Worth putting into the discussion though.
That's maybe 10% of creators.
That's why my #1 suggestion was improving the things creators see on the site to help them understand what they are doing. I think one of the biggest problems here is creators just don't have a good idea what their options are or what their 'responsibilities' are for running a market. They don't know that saying the wrong thing could 'ruin' their market, etc.
@Eliza Well, right now creators are not even told to give resolution criteria at all:
@100Anonymous Sure, but
Description (optional)
does not help at all. The least Manifold could do, for starters, is tell creators to provide criteria.
there are ways to avoid it
Manifold could automatically link a poll to each market asking "Does this market provide clear resolution criteria which minimize the potential for controversial resolution?", keep the poll's results secret, generate a "no drama"-factor for each market out of the poll data and scale trader bonuses and market visibility (and maybe even liquidity) according to the "no drama"-factor.