Final Evaluation Results

Valuations: 2500, 3000, 2500

Median Valuation: 2500.0

Tod Waddington: I honestly cannot believe bro agreed to do this for 20k mana initially. It really is a fantastic way to source high quality markets, and jskf doesn't waste time or overcomplicate things, he's easy to communicate with and created exactly what I was looking for, but better - and solved all of the technical challenges.

Rachel: No update, no change.

Austin: No change in valuation given not much additional usage of the bot. I do appreciate that jskf took time to meet with Christian of Metaculus to talk through how to not spoil Metaculus tournaments -- thanks for helping clarify that.

January 15th Evaluation Results

Valuations: 1400, 2000, 2000, 3000, 1500, 2500 Weighted Median Valuation: 2500

cc6: Good to bring together the prediction market platforms and add more high-quality questions to Manifold

Isaac King: Very useful. High quality questions are a necessity for Manifold, and why not outsource the writing to other platforms? $2000.

wasabipesto: In my opinion Mirrorbot should have existed since Day 2 of Manifold. It plays to Manifold's strengths (create any market!) and alleviates its weaknesses (not being taken seriously) while also allowing people to see a direct comparison of the same question on different platforms. Often I see someone nitpick differences in resolution criteria between platforms but Mirrorbot solves that problem entirely.

Rachel: I gave a high valuation for this last time and still think it's great! I've updated upwards slightly because of the ACX 2024 markets, which make awesome posts like Zvi's recent prediction one possible and which I witnessed a new user being very interested in.

SG: No comment

Austin: One major awesome thing that Mirrorbot enabled was mirroring the ACX questions from Metaculus, so that people could trade them on Manifold as well. These markets have been pretty popular, reflecting the perennial demand for high-quality questions to forecast on. This also led Zvi to create a really great post breaking down each of his own forecasts for the ACX questions and where he would buy the market to — Manifold is extremely interested in having these kinds of quality writeups get produced and shared.

December 15th Evaluation Results

Valuations: 1500, 2000, 1000, 600

Median Valuation: 1250

SG: Adding more high quality content to the site is good!

Rachel: This seems like it should obviously exist. It’s already created a bunch of markets, many of which have resolved, and has 4.92 stars on Manifold as a market creator.

Inga Wei: Cool! Think this would be a great way to be able to directly compare prediction markets

Austin: It’s cool that now anyone can mirror any market via managrams! This is an example of a project that I’m quite happy exists and am happy to award retrospectively, even though it was mostly built before MCF kicked off. If usage continues along these lines, I expect to continue awarding prize funding to it.

I’d love to see more work along the lines of popularizing this bot, eg some kind of way to get it in front of more Manifold users and encourage them to take the first step of mirroring a market. One idea would be to run a tournament for whoever mirrors the market that gets the most traders, or something like that.

(Also nit: I’d include the simplest example in the bot profile description like “mirror <https://www.metaculus.com/questions/17280/frontier-ai-lab-in-china/> with 60 mana” so people don’t have to find the github link. I’d also link to GitHub from the bot profile description)

This seems like a great initiative for a couple reasons:

• A lot of potential to bring new users to the site, if the markets get posted in the comments of the originals.

• Making Manifold more accurate in the long run: If some people trade based on the most accurate numbers from other platforms, Manifold will gravitate towards the better numbers as mana flows towards them.

require some additions to the Manifold API

I would be happy to help with adding/extending the Manifold API to support you doing Multiple Choice!

This sounds like it'll be a huge contribution to Manifold, especially when paired with arbitrage bots that keep the market prices on Manifold in line with those on real-money platforms or any other platform that anyone might think of as a more authoritative source of predictions.

My thinking is that you start with arb bots that force the market prices on Manifold to match the mirrored markets. If those bots make mana, great! Free mana. As soon as those bots start losing mana, that's even better. You can pull the plug on the bots and also shout it from the rooftops that Manifold's market probabilities are provably more accurate than where the markets are mirrored from.

MirrorBot is excellent, and I look forward to seeing jskf make it more excellent!