Never Be Too Pleased With Your Past Work

Out of the thousand articles I’ve written, there are few that I genuinely like. Most of those I feel are mostly correct or useful, upon reflection, are still lacking in a lot of ways. Sometimes they’re too wordy, the research is too sparse or there are obvious counterarguments I ignored.

I feel the same way about all of my books, and all of my products. Since I wrote Learn More, Study Less, several years ago, I’ve done at least five major renovations (although often as different packages, rather than a complete replacement to its predecessor). Even after five generations, I’m still not satisfied with my work, and it will probably take me another few thousand hours of work before I might be.

Looking back at when I did the MIT Challenge, I see the flaws in my design. I can think of dozens of ways that would have make the project more successful, more generalizable to others or more interesting. I’m not finished yet, but I’m sure I’ll look back on this current project with a similar eye for its shortcomings.

Because I live in a Western society, where any lack of self-praise that doesn’t border on oblivious narcissism is somehow an illness that needs to be cured, let me stress: I think this a good attitude to have about your own work.

Too Much Self-Esteem?

If you look throughout history, or across other cultures, it’s hard to see why self-esteem in your own work is currently seen as an indispensable virtue. Eastern cultures historically valued modesty and a focus on process rather than your accolades. Even Western culture’s roots recognize the danger in self-praise: pride did, after all, make the short-list of deadly sins.

Today, self-esteem seems to be the quality one can never have enough of. Almost any problem, from depression to narcissism, somehow stems from not having enough self-esteem. Every successful person is painted as someone with unwavering faith in themselves and their talents.

There’s definitely a point at which, below that, having too low an opinion of your work is crippling. You end up obsessing over details instead of going out into the real world and getting feedback. Maybe you’re below that point, in which case this entire article doesn’t apply to you. I don’t know.

But, I feel, just as there is definitely a lower-threshold where insufficient self-esteem kills your motivation, there’s definitely an upper threshold where it blinds you to feedback. When you think too highly of your ideas and your work, then you can’t see the flaws which should be improved for the next iteration.

Balancing Self-Criticism and Praise

Since knowing exactly where those limits lie is difficult, I’ve found it’s better to employ a rule of thumb: your past work, which needs no motivation since it is already complete, is optimally viewed in a more self-critical light. Your current work, which needs commitment to a plan and less wavering, needs more of your inner motivational speaker.

When I worked on the MIT Challenge, I tried to avoid criticism of the project as much as possible. Not because I knew the criticism wasn’t valid, but because I knew it probably was. My critics had a point: self-grading isn’t perfectly accurate, the value of college has a lot to do with accreditation rather than knowledge, college is about more than just book knowledge, computer science isn’t terribly important to the career of a writer. However, mid-project there’s little you can do with these criticisms other than have them suck away your zeal.

Now that the project is complete, I’m more than happy to entertain those criticisms, and often agree with them to some extent. I don’t need faith because the work is already done—I can instead view my own work with a critical eye, looking for information that can improve the next iteration.

As a blogger, I think the form of this introspection is equally important as its skew. I generally don’t rely on reader feedback (good or bad). Of course, I use it on clear-cut cases of bugs that need to be fixed or features that need to be reworked in a product. Hard data for quantifiable metrics or benchmarking against writers who you feel better you along a specific dimension work well. But the general waves of love-or-hate comments you get as a writer are a terrible proxy for the actual quality of your work.

This last step though, of going through your past work and dismantling all the conviction you built up along the way, isn’t a fun step. It aches to look through the thousands of hours that could be dismissed with a simple objection. Or that a possibly wrong idea has been etched into the thesis of a book.

If your only desire is to feel good about yourself, then, by all means, skip this step. It’s not nearly as fun as being your own biggest supporter. But if your work matters to you on a deeper level than just its emotional or material rewards, I don’t think it’s one you can afford to ignore.

  • Neil Azavedo

    Hi Scott, very true. I totally agree with you when you say “having too low an opinion of your work is crippling”. Personal introspection of ones work is also critical to grow, learn and stay humble along the journey.

    Usually when our work is a product of our true beliefs, we tend to cherish and keep building it.

    Thank you for your sincere perspective.

  • Steven Reda

    Interesting perspective, however i would argue that a high self-esteem does not prevent you from being self critical. Someone with healthy self esteem is actually able to look at their mistakes objectively, rather than ignoring them. I don’t think there is a limit to the amount of self esteem one can have.

  • adbge


    The self-esteem fad is, if not dead, dying, at least is psychology.

    For an example of too much self-esteem, there’s narcissists, but what about bipolar mania? Actually, maybe I’m more thinking of self-confidence here, and to do great work, I think one needs to be supremely confident in that work. (Although how many greats are famously depressives? So maybe I’m off track.)

    I avoid criticism during a project, too, and I think it’s mostly toxic. It kills the creative drive — worry, self-doubt, I mean, they’re part of the process, but if you have too much fear you’ll never create a damned thing. This is one of the great ills of the internet, too, I figure: armchair critics pointing out flaws instead of attempting to do something awesome.

    With that said, reflecting on the shortcomings of a project afterwards is an find idea. It’s at least as old as Polya in *How to Solve It*, and no doubt there’s a line in the Odyssey somewhere which says the same thing. I’m pretty sure Shoenfield talks about it “Learning to think mathematically,” something to do with metacognition. I’ve always thought of it as analogous with refactoring software.

  • John


    Great insights!

    I think you can make “dismantling your work” more pleasurable if you see it as a learning opportunity. With that said, I firmly agree that looking back on past projects sucks. I sometimes feel like I was literally stupid, but at the same time I think we have to be accepting of where we are. We also have to keep in mind that someday we will return to the work we did today and laugh at it.

  • Sebastian Marshall

    Was just linked to this by a friend, had been discussing this a lot over the past few weeks.

    We were analyzing, and realized that just about everyone we know with a very high degree of mastery in their craft dislikes most of their work — but it’s combined with a growth mindset that things will get better with more work done. It pushes one to do better.

    Re: praise and criticism, it reminds me of Chogyam Trungpa’s quote, but in reverse — “Know no hope and no fear, for expectations are the root of all misery.” It seems like your point is the backwards-looking version of that, “Know no praise and no criticism, for expectations are the root of all misery” — perhaps liking or disliking your work are a sneaky form of backwards-looking expectations?

  • Farhad

    What would you do differently in pursuing the 12 month MIT challenge? I am curious to know.

  • Ilham

    Hello Farhad,

    I saw your question and was motivated to answer it. I am actually making some plans to supplement my traditional degree from University (I studied Biotechnology) with new knowledge in Electronics or Electrical Engineering.

    I know you asked Scott but I think I do have some ideas as to what I would do differently. But please keep in mind, that the context I do this in is as an entrepreneur. I am trying to start up my own business in Data Analytics.

    1. Instead of a test system (marking yourself on tests), focus on project based learning. Applying your knowledge in a project your interested in.

    2. Break down basic (year 1-2) courses into many small goal posts; perhaps chapter based goals/achievements?
    And the advanced (year 3-4) courses can be used as goal posts.

    1. Traditional University education in contemporary times is designed in a factory mindset. Private/Public industry desires certain skills in an employee, curriculums are built around this need, graduates hopefully have the desired skills after 4 years. Of course things can change over night.

    Now upon analyzing Scott’s methods of completing a 4 year curriculum in CS, I noticed that he followed this to the core as best he could. My problem with this as has been mentioned either in comments somewhere else on this site, or on another website is that the real world is not based on tests or your ability to memorize facts.

    Let’s be frank, that is all tests in reality do, test your ability to remember facts. Now some may argue that tests in different disciplines test your ability to reason and think. How much thinking or reasoning can you do in a 2-3 hour period? That is assuming an exam, tests are usually much shorter. There is a reason graduate studies take anywhere form 2 years to 5 years (and only increasing as most surveys have found). Real life projects are new problems, that require new solutions or at least evolved solutions from previous analogies, which requires countless hours and nights of thinking, testing, and modification. In essence, project based learning.

    So instead of tests, I would take from the MIT curriculum their meta-learning goals: basics, topics, maybe even notes or lecture videos along with practice problems. But I would develop mini projects on the side and a larger multi-year project. This way your applying what your learning directly.

    As an example:
    I have a degree in Biotechnology, and I want to supplement this with knowledge in electronics. My mini projects could be fun little gizmos, such as working with “Arduinos” or “Raspberry Pi’s” to make my room a “Smart Room”. Or maybe a simple robotics project, such as a cleaning bot that dusts my table every day. My larger project would be integrating my traditional interests with these new ones, developing wearable technology that measures health metrics. Heart rate or galvanometer in shirts or watches. Things that could allow me to learn how to implement between biology and electronics. (of course I’m not giving you my best ideas ;D ).

    2. I think there was a strict adherence to “finishing courses” in Scott’s MIT challenge.

    I have no problem with this, but when preparing an education challenge such as an MIT challenge, I think I would have different goals in mind.

    Courses are great when your at university for these reasons: universities can charge per course, curriculums are built much easier in a modular fashion, each course has a mandate to test what came before it and to improve your understanding a bit on that knowledge or base.

    A lot of basic courses in the beginning have a lot of base knowledge and facts that need to be understood, memorized, and practiced. Advanced senior (year 4) courses are not information dense as they are abstract. You may only have 5 major ideas in these courses, but the time to understand each will be like understanding a few of the basics in one.

    So for me, in the beginning I would take basic facts and use each (even if it takes you 5 minutes to understand and learn) as goal posts. In the advanced topics, you could focus on a more course based goal system.

    This I think is more of motivator for some people, having lots of mini-accomplishments in the beginning. But it may be different for you.

    I would like to know Scott’s comments on this though.

  • Ilham

    Hello Farhad,

    Sorry didn’t realize the earlier length of my comment. Here is a link to an article Scott wrote up on his blog. Entitled: “What I would change about the MIT Challenge.

  • VCB

    I wish to make two points with this post. The first is destructive, whereas the second is constructive.

    You say “Because I live in a Western society, where any lack of self-praise that doesn’t border on oblivious narcissism is somehow an illness that needs to be cured[…]”

    Contrariwise, such an attitude is by no means ubiquitous in the west. Unless you choose to disregard large parts of Europe, of course. But that would be quite a weird move to make, and it is liable to mislead people.


    I agree with much of what you say, but think we would benefit from distinguishing two senses of being satisfied with one’s past work. First, you could think that it is good enough simpliciter. Second, you could think that it is good enough given the position you were in when you worked on it.

    Someone who thinks that everything they do is good enough simpliciter is clearly just delusional, and ought not to be taken seriously. It is much less clear that someone who is satisfied with their performance in light of where they started out and the conditions under which they performed should come in for criticism. For clearly you can consider what you did a great achievement in the second sense, while thinking that it was utterly wrongheaded in the first.

    So you don’t have to choose between being happy with your past work and being critical of it. You just have to not confuse your sense of ideal performance with performing well given one’s circumstances.

    Keep up the good work.

  • Scott Young


    Depends on what you want to learn. Tests have biases, but they are often very good proxies for knowledge on less practical subjects. Programming, something you actually use directly, is better evaluated by what you can do. But a modal logic course? Algorithmic analysis? Number theory? Calculus? These do have practical implications, but most real-life projects that would require the full breadth of what you would study across a semester would be enormous and still probably wouldn’t test whether you truly understood the idea or merely learned how it applies in more narrow contexts.

    The MIT CS curriculum is mostly these types of theory courses. That could be a flaw itself (since MIT goes more theory heavy for classes, expecting students will pick up the practical skills on their own) but it really depends on what your objective is.


    Possibly, I’m making a fairly broad generalization on the East-vs-West philosophical differences, not a really deep philosophical survey, so it’s not entirely accurate.


  • Ilham

    Hey Scott,

    Thanks for the reply. I am just about to start my own MIT challenge, so I am sure that I will probably encounter those problems you mentioned. Thinking about it, in those cases test may be truly better.

    But I’ll get back to you once I complete a few of the courses.

    Thanks for the insight.


  • Mary Taylor

    I don’t understand why so many CEOs are struggling with self-criticism. Accepting others feedback and being able to accept mistakes is really the core of all leadership skills and coaching skills a manager should have.