How to Stop Regressing While You Read (With a Piece of Cardboard)

You Went Back Four Times on the Same Line and Never Noticed

You were reading the second sentence of the paragraph. Without deciding anything, your eye jumped back to the first.

It wasn't a choice. It took under a second, it happened a few more times on this page, and all that's left is the feeling you're not moving forward.

If the question that brought you here is "why do I keep going back", half the answer already lives elsewhere. Not every regression is a flaw: some of them repair exactly what didn't land on the first pass, and blocking them on hard text costs you comprehension. Treat that answer as settled.

This is about the other half: you have an involuntary movement, and you need something to physically block it so you can actually see it happening.

The tool costs about as much as an empty cereal box. A business card works just as well, and in a pinch, so does your palm. How you hold it matters more than what it's made of.

The Cardboard Covers What You've Read, Not What's Coming

Almost everyone who picks up a card for the first time slides it under the line and pushes it down, like a ruler. That's how your grade-school teacher held it too.

The ruler-under-the-line trick does one useful thing: it holds your place. Against regressions it does nothing, because everything you've read stays right there, intact. Your eye has somewhere to go back to, so it goes.

The position that actually changes something is the opposite one. The card sits above the line, its bottom edge glued to the line you're reading, and it moves down over the lines you've finished. The text you've covered disappears underneath it, so there's nowhere left to return to. This isn't a theory about the brain. It's geometry.

Direction isn't a matter of taste: the two setups don't cost the same.

The zone your eye pulls useful information from isn't symmetrical. Rayner (1998), the classic review of eye movements in reading, describes it roughly like this: three or four characters to the left of the point you're fixating on, fourteen or fifteen to the right.

The right side is where the work happens: Schotter, Angele, and Rayner (2012) pulled together two decades of experiments on how much you pick up from the next word before you get to it. Rayner and Bertera (1979) showed the reverse effect too: mask a single letter right at the point you're looking at, and reading speed drops by half.

The practical conclusion draws itself. A card that covers downward, over what's coming, takes away the region that's doing the work. One that covers upward, over what's finished, takes away three characters you weren't leaning on anyway.

Worth saying plainly: this is an inference, not a lab result. No one has tested the two positions against each other. But the inference rests on real numbers, and those numbers aren't symmetrical.

It's a Metronome, Not a Lid

If you leave the card still until you finish the line and only then move it down, you've built a lid. It blocks the regression and that's about it.

The part that matters shows up when the edge moves down on its own, continuously, no matter where you've gotten to. At that point it stops being an obstacle and becomes a pace.

Calibrating it isn't complicated:

• Pick an easy book. Read for one minute normally, nothing in your hand, and count the lines you covered, not the words. Say it comes to twenty-six.
• Divide sixty by that number. You get a little over two seconds per line. That's your natural pace, and the card has no business moving faster than that.
• Start there for the first minute, then cut ten percent. No more than that: a thirty percent jump doesn't make you fast, it makes you watch the card instead of the text.

The movement needs to be a glide, not a series of jerks. The difference looks cosmetic, and it isn't: when the card jumps from line to line, a small pause is left, and the regression sneaks in through it.

Hold it for five minutes, ten at the outside. Past that, your hand starts leading independently of your eye, and you're moving cardboard, not reading.

And here's the clarification that takes the magic out of the story: what's doing the work is the imposed pace, not the piece of cardboard. A finger does the same job at the same pace, and almost everything that applies to a finger applies here too. The card adds one thing on top: it erases the past.

The First Few Minutes Feel Bad, and That Doesn't Mean It's Not Working

The first thirty seconds are strange, purely physically. Your hand feels in the way, the edge intrudes on your field of view, it annoys you.

Around the two-minute mark comes the part that stops most people: the feeling that you're losing information behind you, that the text is slipping away and you're not allowed to check.

The feeling is half true. You have lost the repair mechanism, and on long sentences it shows in what you retain. The other half is the discomfort of a reflex that's just had its target taken away: you can't confirm you haven't lost anything, which is different from actually losing it.

By the end of the first session, the end of the line stops being a cliff edge. By the second or third session on the same kind of text, you usually stop noticing the edge at all.

I don't have numbers for these timeframes, because no one has published an adaptation curve for this tool. Take them as a shape, not a schedule.

One sign is worth watching, though: if the discomfort hasn't changed after three sessions on easy text, either the text is too hard or the pace is too fast. Cut ten percent and start again.

When to Set It Aside

On hard text, immediately. A sentence whose subject only becomes clear at the end needs a regression to settle, and that's what Schotter, Tran, and Rayner (2014) measured, masking each word after the eye had passed over it. On ambiguous sentences, comprehension dropped close to chance level.

The card is a drill for easy material, not a way of reading. You don't read an anatomy textbook with it.

Within a single session, you drop it when the end-of-session check stops holding up. The check isn't the stopwatch: you close the book and say out loud, in two sentences, what the passage was about. If you can't, the speed you gained is speed running on empty.

In the long run, you drop it when you stop reaching for it. The goal was never a lifetime of holding cardboard, but catching the regression while it's happening. Once you notice it, you get to decide what to do with it.

And if regression isn't your main brake, the card is solving a problem you don't have. Taking inventory of your own brakes comes before any drill, not after.

The On-Screen Version, and Why It's Weaker

Reading masks exist in browser extensions, in reading apps, and in "line focus" modes. In theory, they do the same thing. They have three limits.

The first and most important one: most of them don't cover, they fade. The rest of the page stays legible, just dimmer. A reflex whose target is still there isn't blocked, it's discouraged, and the difference shows in the first thirty seconds.

The second: masks that scroll on their own move at a speed you set once. Your hand doesn't work that way. It slows down instinctively on a dense paragraph and speeds up on dialogue, without asking you.

The third is about the screen itself. On a phone, a line runs five or six words, so the regression being blocked is short anyway, and the page scrolls, so you can always swipe back with your thumb. On a wide monitor, though, the mask has something to block.

None of these differences has been tested in a study. What's certain is only that a mask that fades and a card that covers aren't the same intervention.

What Can Honestly Be Said About This Tool

There are no controlled studies on the cardboard card as a reading method. You search, and you find courses, not papers.

The closest thing in the literature was built for the opposite purpose. The 2014 experiment is the laboratory version of the card, except it was designed to measure what you lose, not what you gain.

What remains is a small, solid claim: the tool erases the target of a movement you'd otherwise make involuntarily. On top of that, the imposed pace is the part that produces the number, and pace isn't a property of the cardboard.

From here, two practical things follow. Don't buy anything: a scrap cut from a cardboard box does everything a laminated card sold at a course does. And don't judge a session by how fast it went, judge it by what you can say afterward.

To see whether it's doing anything for you, you need two numbers, not one: how fast you're reading and how much you retained. The first one on its own is flattering. The adult average sits at 238 words per minute for non-fiction and 260 for fiction, from a meta-analysis of 190 studies (Brysbaert, 2019).

Measure once before your first session and once after a week, on the same type of text. If speed goes up and comprehension holds steady, the tool is working for you. If speed goes up and comprehension drops, you went too fast. That's the easiest mistake here to fix.

Measure your speed and comprehension