Subscribe
Sign in
Forecast — A community for crowdsourced predictions. Made by Facebook.
Forecast is a community for crowdsourced predictions, starting with a focus on COVID-19 and its impact on the world. To join the community, please request access to the Forecast Beta Testers group (facebook.com/groups/forecastbeta/) and download the app.
23
Replies
Best
@misbahspeaks We will be working through the backlog here probably over the next few days! https://www.facebook.com/groups/...
Looks really interesting. Can't wait to try it out and SO HAPPY TO SEE @taylorhughes SHIPPINGGGGG!!!!!!!! Your screenshots look great!
?makers How will you deal with forecaster bias?
this is not a market but wanted to pose the question.
"One way the prediction market gathers information is through James Surowiecki's phrase, "The Wisdom of Crowds", in which a group of people with a sufficiently broad range of opinions can collectively be cleverer than any individual. However, this information gathering technique can also lead to the failure of the prediction market. Oftentimes, the people in these crowds are skewed in their independent judgments due to peer pressure, panic, bias, and other breakdowns developed out of a lack of diversity of opinion."
https://en.wikipedia.org/wiki/Pr...
@giannidalerta Good question! This is something we'll be monitoring closely, especially in early stages. :)
@taylorhughes are you guys hand selecting your forecasters? How will you know their bias? seems like a difficult task on yes/no answers, without seeing their past. I guess overtime you can see their bais across multiple topics.
@taylorhughes - interested in how this compares to the Good Judgment Project (https://goodjudgment.com/) (also referenced by @charles_martin )
also - is the goal to build reputation as a forecaster?
@charles_martin @stephenweiss1 The goal in the app is to climb the leaderboard, which is based on points that you earn from buying positions in forecasts, and then those positions gaining value (by either resolving correctly, or by the crowd buying in the same direction). In the short term this can come from betting the way the crowd bets, which feels a lot like polling — but long term this should align with predicting outcomes correctly. You can imagine this leading to accurate predictors winning the market overall and having outsize influence in future predictions, etc.
That's super interesting. Do people predict either Yes or No, or are they giving the weights?
What I'm trying to ask is: if you bring a roulette and ask us "Will it hit red?", it makes sense to respond "no", because the chance is <50%.
Q: Will it poll result in the probability of red to zero?
@visualpharm You can see some of the questions/responses on https://forecastapp.net — there are some yes/no questions and some "linked markets" where you have multiple choices to choose from. (The implementation of linked markets is actually to use several underlying yes/no questions.) You buy in on a certain outcome at the current odds, which are based on what people have previously bought — so you're essentially betting that your chosen outcome's likelihood (and thus, value) increases by more people buying that outcome. It's all fairly confusing. :)
Interesting @taylorhughes. Using the bottomless pit of FB users to find super-forecasters. Competing or collaborating with https://www.gjopen.com/?
ความเสี่ยงซึ่งนำมาถึงความสำเร็จในความแน่นอนและอาจจะได้ผลลับไม่บวกก็ลบเป็นเพียงเป้าหมายที่ชี้นำทางได้เป็นบางส่วน
No thank you. We don't support Facebook Inc. in any way whatsoever in this house.
@edisonjoao6871 very different type of prediction :) you could imagine this more as a way to structure public conversations around specific provable outcomes, rather than trying to predict the future per se.
Forecast