The goal is to finish your backlog / due cards by bringing forth more engaging cards.

What sucks is doing all the easy cards and then there's a bunch of hard cards at the end to slog through. Slogging through the hard cards first sucks just as much too.

The K/E Ratio

There is a tradeoff between knowledge gain and engagement. Focusing more on one can mean sacrificing some of the other, so the K/E ratio lets you tune that balance. We try to slog through as many important high knowledge valued cards as early as possible and fit in some momentum cards in-between when a user's strain gets high, to decrease it back and hopefully keep the momentum going.

The algorithm:

Engagement score.

$E=\sqrt{T}=\sqrt{r(1-d)}$

Higher retrievability (r) and lower difficulty (d) means easier to gain momentum from.

Track session strain

$M=(1-r)^{1-σ}r^σ$

When you fail a card, strain (σ) increases. Whereas passing a card, strain decreases. High strain will tend to easier cards to reduce strain. Low strain will tend to harder cards taking advantage of the low strain.

Track Hype:

Hype is based on hated cards. If a user passes a card they hate, that signals euphoria and hype goes up. Each card holds a euphoric potential that determines the hype's impact (A). Hype begets more hype. So while a user is hyped up, we try to see if they can continue to take advantage of the hype and go through even more hated cards in hopes of hitting euphoria when passing it.

$H'=1-(1-H)e^{-A}$

Hype will decrease when failing euphoric potential cards.

$H'=He^{-A}$

You can configure the K/E ratio to make it more engaging or towards more knowledge gain.

Notes

• Initially made this for the ADHD folks out there. The current anki sorting algorithms are simple and benchmarks prioritize efficiency in time taken to memorize a deck. However the real issue is the motivation to even review the cards. Barely even had the willpower to open up Anki. Needed engagement to be a driving factor.
• This add-on was inspired by the No Scheduler add-on, which created the Discounted Knowledge Score. Our add-on utilizes that to stay efficient. Get as much knowledge as we can get from the backlog and at the same time keep the user engaged.