This game was inspired by a similar-themed game, functionality of which was far insufficient.

Its primary goal is to evaluate one's calibration of credence - how sure you are of something - and help improve it.

It offers a series of questions with answers, and you choose how sure you are of each one. Then you can view how well your credence correlates with actual answers, and correct correspondingly.

Let's say that you plan on doing something and there's a small probability of something going wrong. Then you either have an urge to dissmiss that possibility, or, if it's a very bad scenario, your attention focuses on it too much and it seems like a much bigger problem than it acually is.

If, however, you know that it is a 3% probability, you can actually evaluate how impactful it is.

E.g. let's say failure will cost you 10 times worth of success - then it's a 30% deduction from your total evaluation, take this into account and carry on.

What if the chance is 10%? Then 10% * 10 = 1 weight of failure, whereas 90% * 1 = 0.9 weight of success. It's not (on average) worth doing at all, whatever it is!

Relative worths of various outcomes we know instinctively and weigh them well. With probabilities, it's much, much worse. Not only 2% and 20% feel kinda similar unless you've trained, we tend to assign systematically faulty probabilities. Seeing on exactly which intervals you're over- and underconfident will help you negate that.

Hence, this game.

• Evaluate the existing level of confidence
• Improve confidence calibration
• Introduce no additional bias to the player - questions must be thought out carefully

These are optional and will not be optimized for if it interferes with primary goals.

• Train memory via various algorithms
• Help you show off in an aspect that matters! Publish your results on facebook, twitter, instagram, vine, pinterest, igmur, reddit, linkedin, blogpost! But don't limit yourself - print it and put in a frame next to your awards and certificates!

• Cross-platform app, accessible via https link, and possibly as android/desktop app later
• Full offline functionality
• Fast enough - no lag during q&a (main part of the app)
• User-contributed questions, because just devs can never fill large enough bank

• Themes
• Localization - may be tough and not easily implementable