How to come back from a gap in language studies

Should I just start over from scratch?

Staring at the 4,000 flashcards I had due, that seemed like the most rational option. It's been 2 years since I studied any Japanese. I've forgotten practically everything. I should just start fresh. Do it right this time.

I made this mistake more times than I'd like to admit, but luckily this one time I didn't start over.

I did the scary but wise thing: Just spammed "Good / Remembered / :)" on all 4,000 cards.

To my surprise, it worked out just fine. It was my first of many lessons about how spaced repetition systems (SRS) should be used.

The algorithms are wrong

SRS is amazing in theory. You tell the system how well you remember a specific flashcard, and then it shows you that card again when you are likely to be in danger of forgetting it.

There are two problems with such a system:

People self-rate how well they remembered something differently
Brains and memory are too complex for the system to accurately guess when content should be shown again

Way back when I had those 4,000 flashcards, my SRS program would have liked me to see anything I'd "forgotten" the next day, while stuff I "remembered" could be shown again at all kinds of intervals — a few days, weeks, months, or years.

By spamming "Good," I was ignoring every single "forgotten" card and effectively defeating the whole (supposed) purpose of SRS. Now I was going to see cards that I "should" see tomorrow at much later intervals.

And yet, that didn't matter.

Because I would still see those forgotten cards eventually.

This is the true benefit of SRS is not that it shows you cards again at the right time, but that it shows you them again eventually.

Blasphemy! Lies! Heretic!

Some SRS fanatics (like my former self) might be yelling at their screens as they read this. "Spacing is scientifically proving to increase retention!" To quote Wikipedia:

...active recall with increasing time intervals reduces the probability of forgetting information. This robust finding has been supported by studies of many explicit memory tasks such as free recall, recognition, cued-recall, and frequency estimation.

Spacing effect

I agree. This is true. There are a couple of other things to take into account, however:

Even if the algorithm were accurate, it would break once you miss your reviews for an extended period of time, like I did when I had 4,000 cards due.
Spaced repetition is inherently built into a language. So if you're continuing to expose yourself to the language, you will experience this effect anyway, especially if you're studying with a resource that uses contextual iterative immersion (e.g. NativShark).

Escaping a flashcard avalanche

I was surprised by what happened after I cheated on my flashcards and just selected "Good" on all 4,000 of them without putting much thought at all into their contents.

I hadn't forgotten everything.

According to the SRS algorithms, there wasn't a single card in that entire deck that I should have remembered. But I remembered a lot of them. I wish I'd had the forethought to record exactly what percentage, but suffice it to say that I'm very glad I did not start over from scratch.

I also tried to re-learn thousands of words and concepts that I hadn't even forgotten.

What a waste of time that would have been.

While I didn't particularly enjoy cheating on my flashcards, and I had a fair bit of anxiety about selecting "Good" on cards that I no doubt had forgotten, it turned out to be the right play.

If you've forgotten a card, does it really matter if the SRS helps you re-learn it today as opposed to two or three months from now? Because when those forgotten cards show up again after the wrongly assigned interval you assigned them, you can still re-learn them at that time. And it will be on a day when you don't have an intimidating or seemingly impossible number of reviews due.

Long story short, by spamming "Good," I didn't re-learn things I'd never forgotten, and I still eventually re-learned the things I had forgotten.

In other words, it all worked out.

The purpose of flashcards

Flashcards — and in particular SRS flashcards — are beneficial for two reasons.

They help you avoid forgetting stuff
They reassure you that you are in fact making progress

Usually when people recommend flashcards, they do so for the first reason. This makes sense. It is the logical reason to do flashcards. Don't forget stuff!

But the second reason is more important.

It's more important because failing to remember words and concepts is not the reason people fail to learn things. We fail to learn things when we stop showing up.

For years and years, I had some less-than-friendly voices in my head.

You're not making any progress.

You're wasting your time.

You're not any better at Japanese than you were a whole year ago.

You're forgetting everything.

Flashcards shined a light on the flaws in these arguments. It's easy to feel like you're not making any progress, but when you can look at number of words and concepts learned, and it's higher than the number was last month or last year, you know that you are in fact moving forward. Or at the very least, not moving backwards, which is why review-only days are still so valuable.