“Don’t Learn It Wrong”


“DLIW” is the working title of how I’m attacking each exercise in a more systematic way.

The concept here is unsurprisingly… don’t learn it wrong in the first place. And the way to do this, is to play it correctly more times correctly than not. Many, many times.

What do we usually do with ad-hoc practice? Play it over and over until you get it correct, right? But all we’ve accomplished is learning it wrong – we’ve screwed up 50 times and done it right twice.

Those aren’t very good odds for performing it well.

I realize this is old news to all the real musicians in the world — “perfect practice makes perfect”. But for the life of me, I can’t recall ever being specifically instructed or reading how to really do this. So we’re left to figure how practice perfectly, and it if we can’t figure that out, well, now we’re stuck.

Simply saying “play it slower”, while true, isn’t enough. What if I can’t play it at q=40, and my lip can’t hold out the notes that long – how much slower can I go? At that point we have to explore alternate ways.

How to think of it

I think as like a growing a small plant. Call it “weed theory” of practicing.

You start with a tiny, fragile sprout. Each time you play it perfectly, at any tempo, that nice little sprout grows a little bigger. Each time you play it imperfectly, it dies a little bit.

Because of its frail initial state, it doesn’t take much to kill it. Worse, once you do kill it, a weed grows in its place. Just like the sprout, every time you play it badly, the weed grows. Now you have two problems! First, stop making the weed grow (“when you’re in a hole stop digging”). Then, kill it – play it perfectly many times. And even then you only get the sprout back.

That’s a lot of work to end up where you started.

Putting it into practice

With this mindset, now we can evaluate every action against both the short term result (did I play it right?) and the desired long-term result (am I learning it right?). After all, playing it right means nothing if you don’t learn it right.

This was a revelation, and it made the work both more enjoyable and more efficient at the same time! Less head-banging on hard stuff. Suddenly, progress through even difficult exercises took half as long as they did, if not more.

As I write this now, I realize it even applies to performance.

It should be now be obvious that after sight-reading something, there is a limited window of opportunity before we ingrain bad habits. The longer we wait, the harder they are to fix. And, now we have a concrete idea of what to prioritize, and why.

Maybe this is why all the good players I know tend to take pictures of even moderately-difficult charts on the bandstand. I always wondered why they did this. Surely the pieces was easier for them than it was me, and I didn’t have much trouble. The difference was they were perfecting it, and I would get it 95% right. They were solving their technique problems long-term by preventing bugs from creeping in, and I wasn’t. Maybe that’s why they’re so much better than me! Huh, funny that.


Now that we have the correct mindset for learning, let’s describe one specific way of Not Learning Things Wrong.

As I worked through the first LFCC exercises that were challenging (which were downward thirds in keys like F♯ and B major) and working them up slowly, I started optimizing the process to work through them quicker. It even started to be a bit of game after a while. To my surprise, not only did this save time by skipping certain things, it actually made progress faster because I made less errors going forward.

The One Commandment

Thou shalt stop immediately after a mistake. Do not power thorough the rest of the exercise! Just stop, and start again. This applies throughout this entire process. This is, essentially all you need to know. Everything hereafter is just working within that constraint.

Getting Started

  1. Decide what you’re working on. In LFCC, is it decided for you – in this phase, each exercise is a two bar pattern of diatonic notes in various permutations. The goal is ♩=160. But if you’re learning something, you’ll have to choose what subset to work with first.
  2. Decide what you’re not working on. In my case, I need the technique to be clear, but I’m not working on tone, dynamics, or expressiveness. That doesn’t mean play horribly, just don’t pay attention to what’s not on the table.

Find Your Baseline Tempo

Next is to find the fastest comfortable tempo you can start at. Don’t just start at the target tempo and work your way back, remember, that only makes you start off with a weed. But also don’t start at some too-slow tempo, you’ll just waste time and bore yourself.

Here’s one way to do it, and I admit this might be a too detailed and systematic. The important part is start either where you are reasonably sure you can do it, or the midpoint if unsure.

  1. Take your goal tempo and divide it in half. OK, ♩=160, we start at ♩=80. Set your metronome there.
  2. If it’s really tricky, trial-run just the fingerings at half-tempo. Leave the metronome set at the candidate tempo. Don’t blow into the horn yet. If the exercise is eighth notes, then run the exercise as if they were quarter notes. If your phrase has multiple note values, use the smallest one for each, regardless of the value. Aim for the cleanest possible move that you’re comfortable with, pay attention to how your fingers feel. This way, you can hear any clunking and blipping that might be masked by playing.
  3. Trial-run the fingerings at the candidate tempo. Same as #2, just at the real tempo.
  4. Play it at the candidate tempo. As soon as you make a mistake, stop. You only get one chance at this part of the game.
  5. If you played it twice correctly in a row, change the tempo halfway between the candidate and target, in this case: 80+((160-80)/2) = 120. Repeat until you find your upper failure limit. That is the tempo where you make your first mistake.
  6. When you fail, play at 75% the speed, (80*3)/4 = 60. Repeat until you find your lower limit.

For the curious, we are binary searching to find the best starting tempo while minimizing mistakes.

Next up, working that exercise up to speed.


Now the real fun begins. We’ve figured out what to play where to start. You can probably guess how to get it up to target tempo. Let’s make that sprout grow!

The “DLIW” Main Loop

Without looking at the music, play it twice in a row perfectly before increasing tempo.

What follows may look overly detailed and pedantic, but it is a very efficient and fast in practice. I dare say it gets fun, after you see quick progress.

Get A Good Foundation

  1. (Optional if difficult) Run fingerings at half-tempo, with no sound
  2. (Optional if difficult) Run fingerings at full-tempo, with no sound

Ramping Up To Target Speed

  1. (Optional) Play the music at half-tempo, with sound
  2. Play the music perfectly twice – stopping as soon as you make a single error. You can play it more than twice, but if you make a mistake, you still have to do twice again.
  3. If you play perfectly twice, increase the tempo. Depending on how you comfortable it felt, or how fast it is – this might be 10bpm, sometimes less. I tend to jump 10+ if it was easy, and use the next metronome mark if not.
  4. If you fail twice in a row, go back to step 1, but not optional this time.
  5. If the whole thing is a mess and you don’t get it after 2 or 3 tries, reduce speed to 75% and start over.
  6. Go past your your target tempo. Usually, the next increment isn’t too hard. If you feel like exploring what your upper bound is, keep repeating until you fail. It can be tempting to keep working on these. Really, just quit while you’re ahead.
  7. Finish decisively. If you explored your upper limit, the last few might feel little shaky. Go back to your target tempo or just above it, play it twice, and call it success.

There! You’re done! If you log or record your playing, you’ll find the number of times you played it correctly significantly outnumbers the times you haven’t.

Extra Hints & Tips

  • At any time, play it at half-tempo to refresh it in your head. This doesn’t count as a failure, but it also doesn’t count as doing it correct twice in a row.
  • You can stop playing before making a mistake, it is neither success or failure, as above. You can do this as many times as needed. My mind sometimes wanders, and this is helpful to get it back on track.
  • This works best when you don’t look at the music while playing. Memorize it enough to start, only looking when not playing. This develops the mind’s eye and ear and frees you from visual crutches.
  • Feel free to jump tempos by more than the standard amount, if any are just too easy. After two or three tries at a slow tempo, sometimes things just “click” and you can jump quite a bit.
  • As tempos get higher (say, around 170) it helps to start it at half-tempo every time.
  • If you look at the metronome, aim to do it twice without. We need to use our ear more than our eyes.
  • Don’t rush. Take time to reflect on the success or failure. Let it sink it, but don’t overthink it. Identify one thing that might be holding you back, but move on.

If all went well, your sprout now is a hardy plant in full bloom.