Amazing Austin!
I'll find a time in your Calendly to talk through your questions
1
Base Rate Times
Not fundedImpact certificate
$0raised
$20,000valuation
Project description
What is it?
The Base Rate Times is a global news site built around prediction markets, not media narratives - like The Economist meets Metaculus
How does it work?
The Base Rate Times creates:
Pretty charts, showing how different prediction markets are assessing each news story
Original headlines, re-contextualising news around forecasts from prediction markets (examples below)
Imagine a simple website with charts and news headlines.
We match news stories to existing, liquid forecasts from multiple platforms - the plan is NOT to launch new markets.
Example #1
Conventional news: "Taiwan braces for China's fury over Pelosi visit"
--> Base Rate Times: "China-Taiwan conflict risk up ~2x after Pelosi visit"
Chart similar to this: https://imgur.com/a/yStsfJ9
Example #2
Conventional news: "William Ruto declared winner of Kenya presidential election amid dispute"
--> Base Rate Times: "2:1 underdog Ruto wins Kenya presidential election"
With additional context bubble: "9 out of the last 10 election results have been disputed, a predictable feature of Kenyan politics"
Example #3
Similar to this, except easier to read: https://ourworldinprediction.net/wp-content/uploads/2022/03/Screenshot-2022-03-20-at-08.32.54-1024x697.png
(Taken from the ACX Ukraine warcasting post)
Example #4
Conventional news: "Ukraine won't retake Donetsk by winter because of new logistical problems"
--> Base Rate Times: "12% chance Ukraine retakes Donetsk by winter, down 9 p.p. since new logistical problems"
(Taken from ACX prediction markets FAQ post)
Product roadmap
The first version prioritises prediction markets, but v2 will also integrate betting markets and expert surveys (e.g. IGM Forum https://www.igmchicago.org/).
Context bubbles with key background information will be added where needed (see example above on Kenyan elections).
What is your track record on similar projects?
The draft version of the website will be launched on 31st March for community feedback, so I will have a track record before September. Otherwise, this is my first time launching a media project.
I have been following/dabbling in 'superforecasting' for ~7 years.
My background
~5 years experience at blue chip multinationals, where I owned large P&Ls, led several priority projects and gained experience across multiple business functions
Graduated from a top 50 (globally) university with an honours degree in economics, including studies abroad in Asia and Europe
I have recently quit my job to work on entrepreneurial projects
How will you spend your funding?
I estimate that $20,000 will deliver ~4.5 months runway
Key cost drivers include:
Labour (~25 hours per week running & promoting website, plus ~150 hours building website)
Subscriptions for back-end tools, website hosting, domain, etc. (Webflow, Flourish, etc.)
Admin (incorporation, taxes etc.)
Automation project (automatically scraping latest data from prediction markets)
--> Total ~$4,400 per month
Austin Chen
over 1 year ago
I saw from your EA Forum post (https://forum.effectivealtruism.org/posts/hChXEPPkDpiufCE4E/i-made-a-news-site-based-on-prediction-markets) that you were looking for grants to work on this. As it happens, we're working on a regranting program through Manifund, and I might be interested in providing some funding for your work!
A few questions I had:
- How much time do you plan on investing on Base Rate Times over the next few months?
- What has traffic looked like (eg daily pageviews over the last month or so?)
- How do you get qualitative feedback from people who view your site?
Also happy to find time to chat: https://calendly.com/austinchen/manifold
🌽
Marcel van Diemen
over 1 year ago
Amazing Austin!
I'll find a time in your Calendly to talk through your questions
🌽
Marcel van Diemen
over 1 year ago
Hi Brian, my email is marcelvandiemen@protonmail.com
Crystal Ballin' and Free Basing are brilliant names, would love to talk.
Brian T. Edwards
over 1 year ago
This is a cool project! I was thinking about including a segment in the Crystal Ballin' podcast called Free Basing, where we would take a few markets each episode and maybe bring on an expert to explain how to find the base rates. Seems like it fits well with what you are building and could be good cross promotional opportunity. Would love to talk about partnering up if you are interested!
🌽
Marcel van Diemen
over 1 year ago
Hi Dima, my email is marcelvandiemen@protonmail.com
GradeFact looks fascinating, let's chat
(Just to be transparent, my focus for March is launching the current Base Rate Times MVP)
🌽
Marcel van Diemen
over 1 year ago
What are your plans if you receive less than $20,000 in funding?
I will launch the MVP on March 31st regardless of funding and see what kind of traction it gets.
What kind of visitor numbers would you find acceptable...?
As long as visitor numbers are growing at a better than linear rate, I will continue the project.
My initial goal is ~20k-25k visitors per month within ~18 months (i.e. ~1/10th of current Metaculus traffic). Alternatively, something like ~1000 patreons within a similar timeframe would also be acceptable.
What are your marketing plans?
The initial target audience is superforecasters and adjacent communities like rationalists, Effective Altruists, ACX and MR readers, etc. My marketing plans are:
1) Twitter launch
I have been collecting tweets from big(ish) accounts expressing a desire for a product like Base Rate Times. Example: "this is how the news should look" (https://twitter.com/ben_j_todd/status/1504439141924278278).
I aim to get engagement from these accounts by i) directly replying to their original tweets, and ii) quote-tweeting them in a launch thread. I think this could generate ~30k impressions based on rough data from my tiny personal account (way more upside if it genuinely goes viral) --> ~250 followers to start with.
2) 'Key online influencers' (for lack of a better term)
I will reach out to influential people with an interest in 'epistemic institutions' and a track record of promoting ~similar projects. I will ask Philip Tetlock to tweet about it, Tyler Cowen to blog about it, etc.
If Scott Alexander posts about it (I think >80% chance, conditional on project being funded), the rest will follow on the back of his credibility.
3) Posting to Reddit, Hacker News, etc.
I will post it to the various forums where my target audience hangs out (e.g. LessWrong), as well as typical launch sites (e.g. Product Hunt, BetaList).
Copy in progress, but general theme will be "here's something I made that you might find useful, would appreciate your feedback." I suspect this will take a few iterations to get right.
Not an immediate priority, but eventually I would like to build something like https://www.tiktok.com/@base_rate or https://www.instagram.com/thedailyaus/.
🥨
Dima Gershenzon
over 1 year ago
Great idea, we are also building something in close proximity to your idea, and have it listed here: https://manifund.org/projects/gradefact
Could i get your contact details and maybe we can chat about a collab.
Domenic Denicola
over 1 year ago
What are your marketing plans?
What kind of visitor numbers would you find acceptable and would make you continue to invest in the project?
What are your plans if you receive less than $20,000 in funding?