A core tension in the Manifold community is "power users" vs. "normies". Manifold is trying to appeal to normal people who don't spend much time interacting with forecasting or markets, and as such often neglects features that the power users would find useful. Two of these features that have been frequently requested are a robust search function and improved trading dashboard. So, I built one. Right now it's pretty bare-bones, but I'd like to make it better.
I think this is an ideal division of responsibility; Manifold focuses on its core business model of appealing to a broad audience and not confusing them with complex feature sets, and the power users can design their own more sophisticated tools to cater to their needs. For example, I've been watching the feature suggestion channel on the Manifold Discord for any features relating to search/dashboard functionality, and any that are not implemented by Manifold I plan to add to my tool.
Immediate goals: Fix the lag and other small frustrations that make the site annoying to use, and add a more user-friendly interface. (This is almost done already on my dev copy, I just need to work out the last few kinks before pushing it live.)
Medium-term goals: Expand the features to include more things that power users would find useful and aren't easily available elsewhere, such as visually-compact sparklines on a list of markets, concise indicators of recent trends like movement direction and trader count, automatic Kelly bet sizing, automatic arbitrage across multiple-choice answers, etc. For a longer list of features I'm considering, see here.
Long-term goals: Turn this into fully-featured prediction market trading and discovery software, adding integration with other prediction markets like Polymarket, letting users manage a global portfolio and transfer funds between different platforms, create their own trading bots without needing to interface directly with each platform's API, etc.
I don't have a traditional day job, my income comes from various forms of gig work and self-employment. Any money I get from this project is time that I don't have to spend making that money some other way, and is therefore additional time I can put into Manifold.
If this project is fully-funded, I believe I can complete at least 80% of my short and medium-term goals within 3 months.
Primarily me, following suggestions from the community. Ernie has expressed interest in helping out on the UI and designed their own visual wrapper for my underlying code; we may end up collaborating on it.
Track record: My existing search engine already works well enough that several Manifold users choose to use it over the native Manifold one for various tasks. I have previously completed much larger software projects, the biggest being this one.
Manifold chooses to improve their own search function and dashboards to the point where mine becomes redundant. (If Manifold plans to do this, I'd have no objections whatsoever, and just ask that they inform me before I put a bunch of work into mine.)
Technical challenges preventing me from adding things as quickly as I hope. If this occurs, I will just add them more slowly. (I enjoy programming and want this to be a thing that exists.)
None.
Manifold Markets
almost 2 years ago
Valuations: 600, 1400 Median Valuation: 1000.0
Tod Waddington: I personally don't use it, and I think it appeals to very few people. Count this 0 as a "I don't know", rather than zero
Rachel: No update, no change.
Austin: No real updates, but a slight bump in valuation for the presumed ability for mods to continue using it for the additional month. Thanks again for making this public tool for our community's power users!
Manifold Markets
about 2 years ago
Valuations: 500, 700, 3000, 600, 500, 1200
Weighted Median Valuation: 600
Eliza: The bulk actions toolbar for regular moderators to be able to do group actions without manually using the API is awesome.
Manifold should built this into their main product.
cc6: Very useful for people on Manifold, but not for non-Manifolders. Glad that you have been continually improving the search
wasabipesto: Isaac's search tool is invaluable for how I use Manifold. The onsite search bar has come a long way - and it's fine for a lot of casual searches - but when I want to do something serious Isaac's search is 1000x better. I have a folder of jupyter notebooks I used to use for searching and filtering markets around the end of the year but they have been completely replaced by this super-convenient tool.
Rachel: Increasing my valuation slightly from last time. Letting moderators improve group sorting easily does seem nice for everyone on Manifold.
SG: No comment
Austin: The updates are fairly minor, but I think this is one of those projects that continue to deliver value to our users and thus I’m bumping up my scoring here.
Isaac King
about 2 years ago
The main change this month was adding the bulk add/remove groups option, which moderators have been using to get markets categorized better. I also added an omnibox, though its functionality is minimal at the moment. Probably also some other minor changes that I'm forgetting. I've been brainstorming with Will Howard about how to best integrate Manifolio, but nothing concrete yet.
Isaac King
about 2 years ago
@Isaac-King Oh, please mention any specific features that you liked/didn't like for the purposes of resolving this market: https://manifold.markets/IsaacKing/what-changes-to-my-manifold-dashboa-5151c7c05ae0
Manifold Markets
about 2 years ago
Valuations: 500, 400, 500, 1000, 800
Median Valuation: 500
SG: Impactful largely through getting us to improve our own search page. There are a few other features here we might want to copy as well.
Rachel: People like this, give the people what they want. It’s pretty hard for me to judge the value of the work done so far, since the product doesn’t mean much to me and idk what it was like before, so I’m mostly just going with my gut for the number.
Inga Wei: If power users like better it then I'm for it.
Ian Philips: It sounds like users in the community love your search, are already paying you for it! Seems useful now and should continue to be in the future! It hurts to look at, though 😅. Would love to hear what features your users appreciate the most that we could add in an 'advanced' search mode
Austin: I think “a better search interface for Manifold” was not one of the things that seemed like an obvious good fit for an API integration (compared to something we do ourselves)… but it turns out that there was a lot of demand! I’m a fan of projects that solve a pain point that the creator themselves have, because they have the potential of solving problems for a lot of others; in this case, Isaac’s search/dashboard clearly found a lot of happy users, and for that reason I’m happy to award prize funding to this project.
I do think it suffers a bit from “designed by an engineer” vibes, aka the interface is somewhat cluttered. But maybe that’s the interface that power users want!
Isaac King
about 2 years ago
@FoldyOfficial The focus on function over form is intentional. Manifold already looks really good, and people who care about that can use the native search. It seems like a waste of effort for me to duplicate aesthetic appeal; I'd rather provide a project that's tailored for a different group of people.
(I'm not opposed to making it look good, it just doesn't seem worth spending lots of time on when that time could instead be spent on useful features.)
Isaac King
about 2 years ago
Current project update: I've added several quality-of-life improvements and fixed various bugs; you can find a full list of features on the documentation page.
The site is a little broken right now; I'm having trouble figuring out how to eliminate the lag while the markets are loading in, and I had to turn the database updating script off as it seemed to be eating up all my server bandwidth and preventing people from reading other pages on my blog. So new markets will not currently be reflected in the search, but searching old markets will still work fine. I'm going to be busy this weekend, but will try to fix this on Monday or Tuesday. Apologies.
I'm pretty hopeful about the project overall; quite a few people in the Manifold ecosystem have made suggestions or said they want tools much like what I have in mind for this project, so I'm confident that the demand is there. It's just a matter of finding the time to do everything.
(If any Manifold employee is reading this and wants to help out, there are some simple changes to your API that would significantly help me out.)
After the current bugs are fixed, my focus for the next month is going to be on the trading dashboard side of things. People clearly want a better interface for rapid betting on multiple related markets, and I think I can give them that within a few weeks. I also plan to integrate my rate of return calculator and either Manifolio or my own Kelly bet calculator, with the end goal of letting people input only their personal credence, and then the site will handle choosing a bet amount for them. (This makes betting on current events much faster, since traders will only need to enter one number each time, and won't need to check their current investment amount.)
Isaac King
about 2 years ago
I've created a market on what features will result in this project receiving a Community Fund payout.
https://manifold.markets/IsaacKing/what-changes-to-my-manifold-dashboa
Robert Cousineau
about 2 years ago
I have found Isaac's work on the search function to make my time on manifold much more valuable (I've even paypaled him for it prior to this). Discoverability for older markets is so much better with it than without, allowing us to have accurate probabilities for lower interest questions while still taking less new user attention for them. It gives a much more information dense view than the normal Manifold search, which is beautiful. I have faith the further improvements he makes will be wonderful!
