Anki is a popular flashcard software people use to study and remember facts about whatever they want. Shameless plug, I recently wrote an add-on that adds the challenge of predicting how likely you are to get the correct answer before it is revealed and helps the user get correctly calibrated (similar to the recently removed calibration chart Manifold had). It’s available here: https://ankiweb.net/shared/info/665919307

This market will resolve at the start of October to the number of upvotes minus the number of downvotes at the above link. As of writing it has three upvotes and zero downvotes, so this would resolve to 3-0=3. (As this market is equally easy to manipulate with up- or downvotes, all I can ask is to vote your conscience 😃)

Here is a detailed description of the add-on. I’d be very thankful for any experiences, feedback or suggestions.

What is 'calibration'?

“I'm 97% confident that the capital of Azerbaijan is Baku.” - Someone after learning capitals using Anki Calibrate

Calibration in this context means that if you are, say, 90% sure you know the answer, then you should actually end up being correct 90% of the time. The overconfidence effect is a well-established bias in psychology according to which people in a wide range of scenarios are wildly too certain they have the right answer. In one study, 93% of US respondents stated they are better drivers than the median.

How does the add-on work?

Usually with Anki, you get shown the front of your flashcard, click "Show Answer", see the backside and then score yourself on whether your answer was correct. This add-on replaces the "Show Answer" button with a slider from 0% to 100% for you to predict how sure you are to have the right answer:

Then you score yourself as usual. After doing this a few hundred times, you can check if your sense of your own knowledge is accurate: