Hi,
as far as I understand ease hell doesn’t exist it’s a byproduct of the commonly heard advice “don’t use hard or easy”. When you see your cards often because you are in the so called “ease hell” you should just simply press easy because in fact the card did become easy for you, which would get you out of ease hell naturally. So “only use good and bad” is in my opinion bad advice.
Is this reasonable or did I get something wrong about it?
Ease Hell is a situation where many cards are interval too short and the learning burden is too high, and learners are more likely to burn out. It happens in these situations:
For students, exams are one of the causes. e.g. If the next exam is in 2 weeks and the interval of Easy on the card is 1 month, the student cannot press Easy. Because if the student does not remember the card at the time of the exam, they will lose their score. So if students are worried about the exam, they avoid Easy and intentionally press Hard or Again.
They cannot press Easy because just all those cards are too difficult, plus the cards are important and cannot be skipped. This often occurs with medical students and difficult language learners. (or the cards are made improperly.)
So Ease Hell is the result of the proper working of the algorithm. The learner is learning difficult cards and is intentionally pressing the button to shorten the interval to get a higher score on the exam. It is only learners with time to spare who can solve it by pushing the Easy button or editing cards.
Ease Hell does not occur in FSRS because FSRS adjusts them automatically, but if all cards are too difficult the intervals will be shortened. In this case, perhaps learners can adjust their workload with the FSRS option.
That Ease is determined in SM-2 by what grades you give your answer – is not a myth. That’s just how the algorithm works.
That Ease is incorrectly or unfairly determined by SM-2 – is an opinion some hold. That opinion can often be found on one side of a fundamental disagreement about what Ease means and how it should be used – for instance, whether there is a naturally correct “equilibrium point” for Ease on any given card, or whether Ease is some sort of a value-judgement that a learner should always be allowed to recover from.
So there are those that decided the unfair determination of Ease was a form of punishment being inflicted by SM-2, a so-called “Ease hell.”
That led some to develop full-fledged methods to combat that unfairness by rendering the “unfair” parts of the algorithm ineffective. Like anything that circulates on the internet, those methods eventually became somewhat distorted.
And that in turn led others to develop half-baked shortcuts to combat that unfairness or otherwise trick the algorithm into doing what they wanted. [An example is exactly what you said – 2-button grading alone doesn’t do anything to combat “Ease hell,” it gives in to it instead.]
[If it’s not clear, I personally think that “Ease hell” is great term for marketing, but based on some flawed (albeit well-intentioned) ideas about what was good or bad about SM-2. I think that even the best and most thorough of the methods to fight back against “Ease hell” were a bit shortsighted and made learning less efficient for folks. But since they were still using spaced repetition, it was still fantastically more efficient than anything else! And then I think that everything after that was the uninformed (but still mostly well-intentioned) leading the uninformed.]
Video explaining: https://www.youtube.com/watch?v=roR8S9zjUh8