The 30 chord changes a minute rule, tested
Justin's rule is 30 changes between two chords in a minute before you move on. My baseline on day two was 24, with a string mark pressed into the finger.
JUN 9, 2026
THE FILM THIS IS ABOUT
Watch it on YouTubeThe timer is running and I am counting out loud into the camera. One, two, three, four. D, then A, then D again, and most of them are not ringing properly, but a change is a change and I keep counting. By the time the minute is up I have said twenty-four.
Twenty-four is not a pass. The number I need is thirty.
Then I hold my left hand up to the lens, because there is a string mark pressed into the finger and the finger hurts, and that is the other half of what day two produced.
I had wanted this for years and never done it
The reason for the guitar is in the first ten seconds of the film, and it is not a musical reason.
From a long time I really wanted to learn guitar but couldn't because of life.
What changed is that I started filming things. Planning videos gave me a reason to turn the wanting into a series with a weekly update attached, which is a cheap trick to play on yourself and it worked. I bought a guitar, put a camera in front of it, and said on day one that I would do thirty minutes a day, every day, following JustinGuitar from the beginning. The course is free and it is genuinely good, and I am following Grade 1, Module 1 of it. Nothing I do in the film is a tutorial. I am a beginner filming himself being a beginner.
The format I set up on day one is a timelapse of each day's practice, with a progress update at the end of every week, so the record is the practice itself rather than a summary of it afterwards. It is the same arrangement as everything else on the experiments page: if it is not filmed while it happens, it does not count.
Hold on to the thirty minutes a day, every day. That part does not survive the week.
Day one teaches you nothing you can hear
Day one was a letdown, and I said so at the time.
It was kind of counterintuitive, or say less exciting, because I was thinking at least I will be learning something today.
What day one actually contains: how to tune the guitar, how to hold it, and how to read a chord box. First fret, second fret, third fret, fourth fret, and the fingers numbered off from the index as the first. That is the whole lesson. There is no sound at the end of it that anybody would call playing.
The one practical thing I got out of it is that I had no pick, so I learned to strum without one. You use the thumb, and the skinny part of the thumb, which gives something closer to a melody sound. That is day one's entire output: a tuned instrument, held correctly, strummed with a thumb because I had not bought the ten rupee piece of plastic.
If I were watching somebody else's series I would have skipped this episode. It is still the day you cannot skip, and leaving it in the film is the whole point of filming at all.
Two chords, and suddenly there is a number
Day two is where it becomes an experiment rather than a hobby.
Today I got to learn two chords and now we are playing guitar.
The chords are D and A. And attached to them is the thing that made this filmable: Justin's rule that you have to be able to change between the two chords at least thirty times in a minute before you move on to the next lesson. One minute, two chords, a count.
That is a pass mark. It turns practice into a measurement with a threshold, which means there is a day on camera where I fail it and a day where I do not. Most of what people call learning has no such line in it, and so it can be narrated as progress indefinitely. Thirty a minute cannot be narrated. You either hit it or the timer runs out while you are still on twenty-four.
So I set the timer, counted the baseline, and got twenty-four. I also noted, because it is true and it is on the recording, that most of the chords were not playing nicely at that speed. The count measures changes. It does not measure whether what came out was worth hearing.
D is easy. A is not.
Day three was spent on nothing but switching between those two shapes, and it split cleanly.
D I can put my hand on. It lands where it should and it rings nicely, and after a day of practice it stopped being something I had to think about.
A is hard. I have to twist my hand a long way to get the fingers into that space, and even when it mostly works, the last small string catches on a finger and stops ringing. The twist is what hurts. My wrist hurts from holding the angle and one particular part of the finger hurts from the string, while the index and middle fingers were fine the whole time. The pain is not evenly spread, which surprised me.
Then I ran the test again and failed it. Under thirty on day three.
Justin's own number is at least thirty, preferably sixty. I set my pass mark at thirty, the minimum, and the rule I gave myself on camera was that if I could not do thirty in a minute I would repeat the same exercise the next day until I had a passing grade. No moving on. Which is the correct way to run it, and also the part I would quietly have skipped if nobody was going to see the number.
Forty
Okay, I pass with 40. That was exciting.
Day four, forty changes in a minute, ten over the line and a third faster than the day-two baseline. Two days of practice between twenty-four and forty.
I want to be careful about what that figure means, because it is the only hard number this week produced. Forty changes a minute is a mechanical result. It says the shapes are becoming automatic enough that my hand can leave one and find the other without my full attention, which is the thing the rule exists to check. It says nothing at all about whether the result sounds like music.
Four days, not seven
The week ended at four days of practice, against the thirty minutes a day, every day, that I announced on day one.
The honest framing is that Justin recommends about twenty minutes a day, four days a week, so four days is the schedule rather than a shortfall. That is true, and I said it in the film, and I notice that I only went looking for it after the week had already come in at four. Both of those things stay in.
What I do have at the end of it: Module 1 finished, D and A playable with some confidence, and the 1-2-3-4 strumming motion from the last lesson. The final take of that motion in the film is not good, and I say so immediately:
Sorry, that was really bad. I'm really sorry.
That take stays in the video. Cutting it would make the week look like a cleaner result than it was.
The number that does not agree with itself
One wrinkle, because this is supposed to be a record and the record has a crack in it.
On day four I passed with forty. At the end of the week, summarising, I said I could do around thirty-eight. I also kept calling them strums when what I was counting was chord changes. Neither error changes the conclusion, and both are the sort of thing that happens when the person running the experiment is also the person being measured and is counting out loud with a sore finger.
There is a third crack, in the captions. At 8:01 on day three I say something about my hand that neither I nor anybody transcribing it afterwards could make out, and rather than guess at a plausible word the caption file was left with the gap showing. A guessed word would have read better and would have been something I never said.
The fix for all three is to write the number down the moment the timer stops, which is what the running experiments do now.
What this was actually testing
The guitar is incidental. What I was testing is whether a number with a threshold changes how an adult beginner behaves.
My honest read after four days is that it does, and in a specific way. Without the thirty-a-minute rule I would have declared day three a good practice session, because it felt like one: D got easier, I understood what the A shape wanted, my fingers hurt in a way that reads as effort. The test said I had not earned the next lesson, and the test was right. Effort and progress are different measurements, and only one of them was being recorded.
What I do not know yet is whether this transfers to anything larger. Four days is nothing. The deeper question, the one that also runs underneath The Alchemist piece, is whether a visible, countable pass mark keeps someone at an instrument for the months where it stops improving in jumps of sixteen changes a minute. Week one is the easiest week there will ever be. I went from twenty-four to forty in two days and the next forty will take considerably longer than that.