Three Passes Over the Same Text: The Speed Drill and Its Trap

Round one, at your normal pace: 240 words per minute. Round two, pushed: 350. Round three, pushed harder still: 470.

Ten minutes, and you've nearly doubled the number. It's hard not to write that down somewhere.

Except the explanation isn't the one you're thinking of, and it has nothing to do with how hard you worked. It comes down to one detail baked into the exercise itself: the text was the same, three times over.

How the Drill Works: A Stopwatch and Two Pages

You don't need anything special. A book in plain prose and the stopwatch on your phone.

Pick a passage two or three pages long and mark the start.

Round 1, two minutes, your pace. Read the way you normally would, without rushing at all. When the timer goes off, mark where you got to and count roughly: lines covered times average words per line, divided by two. That's your session baseline.

Round 2, two minutes, from the same mark. The target is about half again as much text as in the first round. You'll feel yourself skating over things. That's fine — it's part of the exercise.

Round 3, two minutes, starting over again. The target is double round one. Here you drop any concern with understanding entirely and focus only on movement: the eye moves forward, doesn't go back, doesn't stop to confirm.

To hold an imposed pace you need something outside yourself, because left alone you slow down without noticing. A finger or the edge of a card under the line does the job. Not because it speeds up the eye — the evidence for that simply isn't there — but because a marker moving at a steady rate shows you instantly when you've fallen behind.

So far this is a reasonable exercise: cheap and easy to repeat. The problem shows up the moment you take the number from round three and call it your reading speed.

Why the Second Pass Is Cheaper Than the First

The eye doesn't glide along the line — it stops at points and jumps between them. And how long it stays at any point isn't constant: Rayner gathered the measurements in 1998 showing that fixation duration depends on how frequent the word is and how predictable it is in context. Predictable words get shorter fixations, and some of them are skipped entirely, with the eye never landing on them at all.

The model built by Reichle, Pollatsek, Fisher, and Rayner that same year explains why. The signal that moves the eye forward isn't an internal clock — it's the moment the current word has been recognized. Recognition finished earlier means departure earlier.

Add to that what you pick up out of the corner of your eye. Schotter, Angele, and Rayner synthesized, in 2012, two decades of experiments on the information pulled from the area next to the fixation point: the first letters of the next word reach the brain before your gaze gets there, and the gain is larger the more predictable the word is.

Now put all three together on a second pass over the exact same paragraph.

Every word sits at maximum predictability, because you just read it. Recognizing it is as cheap as it will ever get. And predictability plus recognition speed are exactly the two variables that shorten fixations and raise the skip rate.

The number climbs, then, without anything having changed in you. Kuperman and Van Dyke showed in 2011, across a full battery of verbal skills, that word-recognition speed predicts reading speed better than anything else measured. The second pass doesn't train your recognition. It subsidizes it — once, on that one passage.

What the Round-Three Number Actually Measures

What the drill leaves out is the point of comparison. To know what 470 words per minute on the third pass actually means, you'd need someone reading that passage for the first time, at the same pace, with the same comprehension check. On your couch, you don't have that.

How much of the jump is familiarity and how much is skill — I can't give you a percentage, because nobody has measured it on your text. The direction, though, isn't ambiguous, and the familiarity share is large.

You do have a way to see the difference with your own eyes, in five minutes. After the three passes over passage A, open a passage B from the same book and read it once, at the pace you were just holding. If that speed was really yours, it carries over. If on B the pace falls somewhere between the first and second round, you've found out for yourself what the final number was worth. It's not a lab, but it's the check the exercise doesn't hand you on its own.

There's another layer. Carver worked out, in 1990, the threshold past which the process changes nature: above a certain pace you're no longer reading, you're doing something else, with a different yield. Masson had shown, in 1982, what survives when you skim a story — the broad thread, while the details fall away.

On the third pass, the broad thread is already in your head, put there by the first two. So you skim, and still get the clear sense that you covered the text. The exercise produces the very feeling that confirms its own success, which is an awkward position for any measurement to be in.

This effect shows up, more weakly, even when the second test is on a different text, just because you've taken it once before; it's worth reading separately why a second speed test almost always comes out higher. Here it's the extreme version, because the material itself repeats.

The field's synthesis, signed by Rayner, Schotter, Masson, Potter, and Treiman in 2016, is blunt on the subject: the trade-off between speed and comprehension is real and doesn't go away. A number with no comprehension percentage next to it, taken on a text you already know, isn't a measurement.

What the Three Passes Are Actually Good For

Something does remain, and it's exactly the repetition that makes it possible.

When you no longer have meaning to decode, your whole attention is free for the mechanics of movement. You feel, concretely, where you want to stop needlessly, where your inner voice finishes pronouncing a word the eye had already left behind, where the pace collapses because you've hit the end of a line. On new text you can't notice any of this, because you're busy understanding.

What gets trained is pace control, not the ability to read. They're different skills, and the first one really does learn fast.

With one important caveat. Round three bans you from going back, and going back isn't all waste: Schotter, Tran, and Rayner blocked rereading in 2014, masking each word once the eye had passed it, and comprehension dropped mainly on sentences that called for reinterpretation. A drill that forbids it is an exercise, not a way of reading.

How to Use It Without Lying to Yourself

Three passes over a passage, then you switch passages. A fourth pass measures memory, not reading.

Don't test your comprehension on the same text. You'd be checking what stuck from the earlier rounds — a different question, a perfectly legitimate one, but not this one.

The number you write down doesn't come from the drill session. It comes from a separate, cold measurement, on a text you've never seen, with questions at the end. Once a week is enough.

And don't do it on the chapter your grade depends on. There, you need the backtracking, the pauses, and the time they take.

If you want the same forced-pace effect without the measurement confusion, the version that uses fresh text every round is the push-up method, where the baseline round and the recovery round are done on different material.

The Measurement Happens on a Different Text

The three-round drill is good at what it's good at: it shows you that your reading tempo is a negotiable habit, and it gives you a place to practice pace control, without paying the comprehension tab.

The verdict just doesn't happen there. It happens on a text you don't already know, with the stopwatch and the questions running at the same time.

→ Measure your speed and comprehension, free