What do you do when you run into a difficult passage that you can’t play?
Do you try it again? Repeat it a few times to see if anything changes? Maybe slow it down a bit?
We all run into difficult and challenging passages, no matter how many years into our musical journeys we might be. So have you ever wondered what expert musicians do when they encounter a problem passage that they can’t play at first?
Like, how much time do they spend on it? How many repetitions do they do? How do they avoid reinforcing errors? Or keep from getting frustrated and rage quitting?
Believe it or not, there aren’t really any studies out there that analyze expert practice in a detailed repetition-to-repetition way that would help us answer questions like these.
That is, until now!
A study!
A 2025 study, authored by performer/educator/researcher Micah Killion and previous podcast guest Robert Duke, does a deep analysis of expert musicians’ practice behaviors and provides some revealing insights into how they practice challenging passages.
Six musicians
Six artist-level trumpet players were recruited for the study, each of whom were members of or held principal positions in top ensembles and orchestras in the US and abroad.
Then the researchers asked the musicians to videorecord a practice session where they practiced a passage of music that wasn’t where they wanted it to be.
Here were their instructions:
“First identify a passage in your current repertoire or exercises to play. Then record a video of yourself playing that passage warts and all, followed by whatever you do next to get the passage closer to something you’re satisfied with.”
Similarities and differences
The issues they chose to work on ranged from intonation, sound, and coordination issues on pieces that they were in the very early stages of learning, to detailed “polishing” work on fast passages in repertoire that was close to being performance-ready.
There were quite a number of differences in some aspects of their practice – like how much time they spent practicing each passage, how many repetitions they did, how much time they spent thinking between repetitions, and so on.
But there were some key similarities in their practice as well. And these common elements are what I think were most interesting, illustrative, and relevant to our practice as well.
Six key behaviors
Ultimately, there were six common practice behaviors that emerged from each of the musicians’ practice videos.
Behavior #1: Clear intentions/predictions
Each of the expert musicians were guided in their practice by a clear musical intention and expectation for how each repetition (and the passage overall) should sound and feel. Kind of like sailors navigating the seas using the North Star.
On average, the musicians spent only about 57% of their practice time actually playing. A good bit of their practice time was also spent thinking and reflecting on each repetition, or singing and conducting to help themselves clarify what they were aiming for with each phrase.
They also tended to stop whatever they were doing immediately after making an error. This is revealing, as the researchers note, because it implies that the musicians had a pretty clear idea what they were going for in each repetition and knew immediately when they fell short. Which is only possible if one has a clear goal to begin with.
Behavior #2: Strategically modified repetitions
Repetition is important in practice, of course. But not repetition for the sake of repetition. Or stubbornly repeating the same thing over and over while continuing to make errors.
Whenever the musicians failed to achieve their goal with a particular section, instead of playing the section again in the same way, they often strategically modified the section to increase their chances of playing it correctly the next time.
Meaning what exactly?
Behavior #3: Doable-izing
In the same way that there are regressions and progressions in physical therapy, strength training, mobility work, or yoga, where you use supports or more limited versions of the exercise as you work to build up to a full range of motion, if the musicians couldn’t play a section up to their standards, they modified the passage to make it simpler or easier. A process the researchers termed “doable-izing.”
This might entail playing fewer notes, slowing things down, reducing the dynamic range, modifying articulation, etc.
The idea is to reduce the difficulty of the problem passage, so that it becomes possible to play it accurately. And as the simpler version stabilizes and becomes more secure, gradually add complexity back to the passage so that it gets closer and closer to your goal.
Behavior #4: Expression always!
That said, the musicians did their best to maintain as much of the musical and expressive context in each repetition as possible!
After all, oversimplifying a passage, and making it too easy, runs the risk of cultivating bad habits too. For instance, if you slow a passage down, and take out the phrasing, vibrato, and dynamics, sure, you might be able to play it in tune, but now you’re playing a completely different version of the passage.
Importantly, the researchers found that these expert musicians included expression in every (emphasis theirs) single repetition captured on video. At no point did they play a passage without at least some expressive elements. Which again goes back to behavior #1, and how important it is to have a clear intention or goal for each repetition.
Behavior #5: Recontextualize systematically
Of course, it’s one thing to nail a difficult jump or shift or leap when you start right at the tricky spot, and another thing entirely to nail it in context.
Once the musicians were able to play a doable-ized section up to their standards consistently, the musicians gradually added back the contextual ingredients they had removed earlier. Like adding in more notes, or playing with a full dynamic range, or increasing the tempo.
However, they did so thoughtfully and systematically, making sure the passage remained doable, by adding back only as much as they could handle.
Behavior #6: Think!
In a recent podcast episode with violinist Ida Kavafian, she mentioned that her teacher, Oscar Shumsky, had signs up around his teaching studio which simply said “Think!”
Likewise, the musicians in this study spent a significant portion of their practice time thinking and reflecting between repetitions. Amusingly, the researchers put together a cut of only the pauses between repetitions for one of the participants, which you can see here.
So…what are we to do with all of this?
Take action
Well, first off, the full article is well worth reading, as there are so many nuances and details that weren’t included here, including a more detailed analysis of the musicians’ practice (which trumpet players in particular will enjoy) and excerpts from practice videos. You can check it out here.
But if you don’t have time for that right now, the researchers consolidated the core ingredients that they observed in the expert musicians’ practice, into a practice framework which they refer to as the “central strategy of music practice.”
Essentially, if you’re wondering what effective practice for musicians looks like, this is it!
NOTE: Rather than diluting the authors efforts by rewriting the steps in my own words, I’m including the framework below exactly as was described in the study.
- Formulate a vivid intention/prediction for each trial
- Perform
- Identify discrepancies quickly
- Pause to think
- Strategically repeat passages to minimize prediction error by
- adjusting the parameters of each succeeding performance trial to render momentary challenges surmountable in the near term
- maintaining contextual elements to the extent possible
- recontextualizing strategically in ways that maintain a high percentage of successful trials
References
Killion, M. F., & Duke, R. A. (2025). The central strategy of music practice: A blow-by-blow account. Journal of Expertise, 8(2–3), 85–128.
Dear Noa,
Thanks for this article — it’s interesting and succinct. I teach beginning band kids and general music. The key thing here is the very first step in which you already have an idea of what the passage could sound like. But for most kids, they don’t have the skills to audiate at that level; they are using the notation to find out what the music sounds like. Decode and repeat until it’s more or less memorized. Maybe they’ll get around to having a concept, maybe they won’t.
You can provide a recording, but then you get into the issues that Building Thinking Classrooms talks about (this is a math pedagogy system) — is mimicry thinking? Is thinking about the accuracy of your mimicry thinking? Is learning by ear not thinking? Or is it thinking because you are constantly comparing what you did to what you heard? In a way that, say, doesn’t happen when you are doing math homework that consists of drilling?
I would love an article from you that focuses on what practicing should be like to lead to this point where you are starting from conceptualization. How do you get absolute beginners to practice at a more conscious level?
cheers,
Andrea
Thanks for the question, Andrea. Yeah, it’s an interesting question, and one that several previous guests have spoken on – although not necessarily at the beginner level that you’re referring to. Although, I think their thoughts could still be relevant.
The gist, is that listening does seem to have a lot of value. At some point, yes, being able to audiate and derive a concept from the printed music is important, but being able to learn from ear, or develop one’s listening skills is hugely beneficial as well.
Here’s an article that gets at a tiny bit of the research that might be related to your question, as well as links to a few artists who brought this up in their podcast episodes: https://bulletproofmusician.com/a-simple-way-to-keep-music-in-your-fingers-when-practicing-more-isnt-an-option/
There’s also the recent podcast episode with pianist Orli Shaham, who actually surprised me when she revealed how listening fits into her learning process: https://bulletproofmusician.com/orli-shaham-on-overcoming-setbacks-and-learning-to-trust-the-process/
Now that I think of it, I think harpist Lynn Aspnes spoke to this as well and how her father had her listen to multiple recordings as part of a guessing game: https://bulletproofmusician.com/lynne-aspnes-on-singing-breathing-and-the-value-of-cultivating-a-diverse-range-of-musical-influences/
There are other research-related articles that pop up here and there on the blog that could relate to this, but hopefully this is a helpful start!
(Beginner at age 74) I have been receiving your emails for about 1.5 years, and out of all your suggestions, I have faithfully put into practice “Microbreaks” which you explained would really help with muscle memory, and it has. So, I was surprised – and disappointed (?) – that it was not part of the CORE practice habits. What’s up with that??!!
Thanks for sharing your experience with microbreaks. Always helpful to hear how things work in the “real world”. =)
There are a lot of different techniques and elements that can influence motor learning, ranging from microbreaks to different ways of structuring practice (e.g. interleaved) to increasing variability (e.g. variable practice or differential learning) or even magnifying errors (e.g. amplification of error).
Ultimately, these are all important and even essential aspects of the learning process, but the authors suggest that the basic, underlying, fundamental process involved in motor learning may be the one described in the study.
Like, if you’re thinking about cheesecake, the crust matters, the toppings are important too, but you gotta have the cream cheese, otherwise it’s not quite going to be what you want it to be (I’ve been trying to find a cheesecake for my daughter’s birthday today, as it’s not necessarily super easy to find where we are at the moment, hence cheesecake being top of mind 😆).
I’m having some fun adding “doable-izing” to my practice. For an organist this could mean (besides slow practice, which I already did) ideas like manuals only, pedals only, left hand and pedals only, right hand and pedals only. I plan to try to come up with other ideas besides these.
Thanks for sharing, John!
A game changer for me is the suggestion that I “doabilize” a section so I can hear the difference in ONE practice session. I’ve always wondered how much to modify a section I’m working on, and often have not heard improvement until a week or two or longer. Now I focus on something, which is usually very limited, very “micro”, but in which I can discern improvement in one session. My question: Is this micro-doabilizing a good strategy for an intermediate player? Or should I keep my horizons broader until I become an advanced player?
Hi John – yes, I believe the idea is to doable-ize within a practice session, regardless of level. Because the core principle here is to maximize “correct” repetitions, while maintaining as much context as possible. Which one could argue is perhaps even more important at intermediate levels, as one is working to build up solid fundamentals and habits.
This is great! Thank you, Noa, for always bringing the latest research to us and for summarizing it into “doable” chunks. We appreciate you!