“If you can't answer it in about ten seconds, that's not a flashcard, it's a slide.”
You've got 180 cards from one lecture, you review them every day for a week, and in the exam you still can't answer the question. I did this for a whole semester before I worked out the cards were the problem, not me.
Almost every bad deck I've made had the same cause. I made the cards while reading the slides, one card per bullet point, without ever asking whether the card was worth having.
What's actually wrong with a bullet-point card
Three things go wrong, and they're easy to spot once you know them.
The card doesn't ask anything. Front says "Osmosis". Back is the whole slide. There's no question there, so there's nothing to get right or wrong. You flip it, you read it, you feel like you studied.
You learn the card instead of the idea. If the answer is always the third bullet, or the card with the long back is always the one about active transport, you start recognising the card itself. That works perfectly until the exam asks the same thing in different words.
One card is holding a whole list. "Name the four stages of mitosis." You get three, you flip it, you go "yeah, close enough" and move on. You've just marked something correct that you didn't know.
The rule I use now
One question, one answer, answerable in about ten seconds.
That's it. If it takes longer than that, the card is doing too much and you should split it. If there's no question, it isn't a card yet.
The ten seconds part matters more than it sounds. It's the difference between a card you can actually be wrong on and a card you skim.
Make them after the first pass, not during
This is the thing that changed my decks most.
Making cards while you read feels productive because you're producing something. But at that point you don't know yet which bits are hard, so you make a card for everything, and most of them are for things you already knew.
Do a recall pass first. Read a chunk, close it, write down what you remember, see what you missed. Now you know exactly which things keep slipping, and those are the cards. Everything else can stay in the slides.
I wrote about that pass in more detail in the first post. Day 5 is where the cards come in, and that's deliberate.
Writing a card that's actually worth reviewing
Turn statements into questions. "The mitochondrion produces ATP" becomes "what does the mitochondrion produce, and from what?"
Ask why and how, not only what. "What is active transport" is a definition you can recite without understanding. "Why does active transport need energy when diffusion doesn't" is the thing you'll actually be marked on.
Write the answer in your own words. If you paste the slide's wording, you're testing whether you can recognise that sentence. If you write it yourself, you have to have understood it first, which is most of the value already.
Put context on the front. "1815" as an answer is useless if you can't remember which of six dates it was.
You need way fewer than you think
A lecture that produces 180 cards will usually produce about 25 that matter. The rest are things you already know, things you'll never be asked, and three different cards all testing the same fact.
A deck you dread opening doesn't get reviewed, so a smaller honest deck beats a big thorough one every time.
Prune as you go. If you've got a card right three times in a row, take it out. It isn't teaching you anything anymore, it's just making the deck feel bigger.
What flashcards can't do
Cards are good at facts, definitions, formulas, dates, vocabulary, the stuff with a right answer.
They're bad at understanding. If the exam asks you to explain why something happens, or to compare two processes, no card is going to save you, because you can know all the pieces and still not be able to put them together out loud.
For that you have to explain it, properly, in full sentences, to someone or to yourself. Cards and explaining are two different tools and you need both.
Where a tool helps and where it doesn't
The copying is the boring bit. Reading a slide and typing the same thing into a card adds nothing to your understanding, and that's exactly the part software can do for you. Leurons makes cards from your slides in one go, so the hour of typing is gone.
What it can't do for you is decide what matters. You still have to go through the deck, cut the cards for things you already know, and rewrite the ones where the answer is a whole paragraph. That's ten minutes of work, and it's the ten minutes that makes the deck worth reviewing.
The short version
Don't make a card for every bullet. Make cards after you know what you keep getting wrong. One question, one answer, ten seconds. Cut anything you get right three times in a row.
A deck of 25 cards you'll actually review beats 180 you won't.