It's very simple: figure out what you suck the most at, and practice that carefully until you don't suck as much. Rinse and repeat.

The things you suck the most at are usually painfully obvious.

There's a bit of a chicken-and-egg problem here.

Only things I’ve done “it” for were matters of physical exercise - first indoor rowing utilising an actual coach and a progressively overloading program, and now for running utilising ChatGPT as my coach, as sort of an experiment.

I suppose the “how” is really boring, and that’s why none of the resources talking about deliberate practice really seem to dive too deep into it. For my rowing practice it was simply getting feedback from the coach constantly regarding form and doing checkups for my pacing and current condition. I’m more interested in this running practice with ChatGPT, where I aim to keep it updated on my current progress on the program we came up with together, and ask for advice on possible adjustments.

After all the difference between deliberate practice and any old practice seems to be the progression and constant feedback from someone who knows what they’re doing.

Yes, exactly. How to design such training for some specific area, that's still an uncovered topic.

Or just some guides for some topics, even a few examples of how deliberate practice for X might look like is already much better information to start your own practice than these aloof principles. It's great to know the principles if you're also aware on how to apply them, if there's no path to application it becomes mostly noise...

This is why "self help" books are so rarely helpful. They often lack concrete, actionable steps that you can take and simply provide faux-epiphanies.

I do not doubt this author's words, but they did not provide much "concreteness" here.

Sorry for kinda hijacking this, but it is the main reason I am building trydeepwork.com, to actually put deep work or deliberate practice into actual "practice".

Do you think of deep work (as described by Cal Newport) and deliberate practice as being similar or connected? I could see deliberate practice being deep work, but deep work doesn’t necessarily have anything to do with deliberate practice.

Your overall goal is to improve your ability to do some larger task. Larger tasks tend to be able to be broken down into smaller tasks. Improving your ability on smaller task tends to make you better at the larger task.

So what you do is find a way to:

- somewhat isolate a smaller task

- then focus on doing that smaller task with ONE particular focus or improvement in mind

- have feedback on how it went

When feedback is present and you are paying attention, you will tend to fairly naturally improve.

Repetion is good, but repetition in different circumstances is much better for learning.

Deliberate practice is practice + an evaluation feedback loop. If you aren't analyzing how you are doing and using that to feed back into your practice you are just practicing. The use of a coach\teacher\oracle as the evaluator is important because of the DK effect, higher skill people generally are better at evaluating a performance making the feedback loop better.

So in general: practice activity, evaluate performance, come up with drill to improve weaknesses identified during evaluation, do drill, repeat.

Love doing something so much that noticing what you are bad at and putting attention into picking up your performance in that area comes naturally (inspired practice is more effective than 'deliberate'). Along with the intelligence to solve those issues effectively and the wisdom to ignore what's not so important. Rinse and repeat for x units of attention.

Source/flex: that's what got me into the top .01% of multiple pursuits, or into a decent percentile with not that much time invested.

What are you good at?

Peaks: Speedcubing (top 10), classical guitar (comp level in my youth), couple of major games (top .01% recently in one, top .1% a while ago).

I wish that I had learned to love programming in my teens, these things have made me happy but certainly not rich, haha. I like it but don't love it - I'm just okay at it.

Spaced repetition is useful, there's a level of nuance beyond that with interleaving and variation that hasn't really hit the mainstream of podcasters trying to hype research yet. https://bjorklab.psych.ucla.edu/research/ has some good nuggets.

The article was so long, if someone here read it top to bottom I would be very impressed.

The "how" is taught in a lot of places without those people knowing what they're referring to; immediate feedback. In golf this takes the form of a launch monitor of some kind. You have to know when you do it wrong, and the faster you know the easier it is to correct.

This is 99% of the content from fs.blog.

Remember when they were writing constant articles about mental models... that didn't contain a single mental model?

Write down the definitions of words with citations into a kind of personal glossary, drilling down until I have a concrete understanding. I typically put the definitions into a spreadsheet, one column being a direct quote from the reference and another being my own definition - this is typically more like color commentary and connects to other concepts or even just journals my own journey of exploration.

I check it first for word (or acronym) definitions and it reinforces the whole thing.

I have found it effective for doing deep dives into academic topics (so many medical research papers) and tech.

Can you share a sample?

Actually no, as it turns out, I don't have 'the good ones' anymore - or at least not conveniently searchable in some cloud storage and instead on some specific machine somewhere. But a sample row of an older version is

APDU Application Layer Protocol Data Units, the actual packet on the network. Max size specified by device Max_APDU_Length_Accepted parameter, one of two alternatives to NSDU

later versions would simply have copy and pasted the relevant text from the source into a different column and the URL into another column - I would still use zotero as a citation manager it was simply convenient if I needed to go back to the source to review context. The core features are present here though: a definition some context that I thought was important, how I understood the thing itself.