Inspired by Dynomight’s viral post “Why doesn’t advice work?” I have been thinking about why does advice work so poorly. Or why is there such a stark disconnect between the application of said advice and the desired outcome? Why does advice work for a few and not most? I am going to ignore the edge case where the advice is ignored, in which the outcome is certain failure. So assuming the recipient makes a concerted effort at applying advice, why doesn’t it work?
Everyone is the product of unique circumstances, whereas advice tends to be one-size-fits-all. Advice tends to not take into account all the ways humans vary. This can be upbringings or innate attributes, like IQ, attractiveness, or height. It’s not unreasonable to assume dating advice will work better for people who are naturally more physically attractive or have more money, all else being equal. Or that studying advice (e.g. Cal Newport books) will be more effective for people who are overall smarter (Dr. Newport in 2004 earned a Ph.D. in computer science from MIT).
For example, to get ‘six-pack abs’, it’s commonly advised to cut calories and ‘do cardio’ until the sought leanness is attained. Simple enough. Yet in spite of the immense popularity of fitness content on social media, the vast majority of people who apply this advice fail. Maybe this is due to genetics or other factors outside of one’s control. Even the best advice will still be constrained by one’s innate limitations.
In my post “Individual differences of metabolism are real and matter” I give a real-life example of someone who despite only eating 1,800-2,000 calories/day, which he carefully tracks, and doing cardio, is still overweight at over 200lbs at 5’10”. How many more calories can he possibly cut when he’s already eating at well-below the predicted daily estimated calorie intake for his height, weight and activity level? Any fewer is likely infeasible or intolerable, and maybe even unhealthy.
There are many such cases like this, of people who are still visibly overweight and by BMI despite applying popular health and fitness advice. Likely someone with a naturally slower metabolism relative to height and weight will generally find it harder to get lean. I ran into this problem myself where basically my body refused to drop anymore weight, but at least I was nowhere near 200lbs. Maybe people who produce fitness content and espouse such advice are genetic outliers or are concealing the actual reason (e.g. drugs instead of only dieting).
This ties in with survivorship bias. People who are successful at applying advice leave positive testimonials, whereas those who fail either drop out and quietly go away or complain. Those testimonials are collected and shared as ‘evidence’, whereas the complaints, in the comments, are deleted if possible.
Also, what happens if multiple people apply the same advice in the same domain? For example, in Stephen R. Covey’s hugely popular book The 7 Habits of Highly Effective People, it’s advised to be a good listener by letting the other person talk and showing interest. That makes sense. But what happens if two people who read that book enter into a conversation; will they just stare at each other until someone finally breaks the ice? This is obviously a contrived example, but it applies to anything competitive, in which the efficacy of advice becomes diluted when many people are applying it.
If high schoolers are following the same advice or guides for scoring well on high-stakes math competitions to get into top schools, like by reading the popular The Art of Problem Solving books, scores will rise overall. But everyone is still in the same relative place yet working harder. Same for the recent trend of score inflation on the LSAT, likely due to more studying and apps. Now a higher score is required to get into a good law school compared to as recently as only five years ago. Or on the GRE, scores of 700-800 on the math section are increasingly common due to studying. Or in writing, what happens when everyone is following Strunk & White.
Other times advice so vague or non-specific as to be useless. Or it’s contradictory. For example, it’s commonly commonly advised to ‘write clearly’ or to ‘write like you talk’. What does this mean in practical terms? Human speech tends to be cluttered or full of filler (e.g. “ums” and “ahs”) when there isn’t a teleprompter or script to read from. This would make ‘writing like you talk’ very poor advice (unless you’re trying to meet a minimum wordcount for a school assignment I suppose). Regarding clarity of writing, this contrasts with other advice to vary the language. Spelling everything out explicitly or using as short of sentences as possible would leave minimal room for ambiguity or misinterpretation, but would also make for very boring reading. So the writer must make assumptions as to how much familiarity the reader has with the topic.
Or advice is highly situational or there are obvious exceptions. To be likable, it’s commonly advised to ‘be a good listener’ and to be unpretentious and humble, yet this does not explain the apparent success and likability or aura of people who embody the exact opposite of those traits (e.g. Donald Trump and Andrew Tate), who have tons of followers and evidently many people who do like them despite their arrogant demeanors. Elon Musk has the most popular account on Twitter/X, yet ‘humble’ or ‘good listener’ are not among the most common adjectives one reaches for when describing him. Or maybe people who are popular and successful later become arrogant, not that being arrogant leads to success.
In regard to career advice, the economy is also constantly changing. Now artificial intelligence threatens not just careers, but entire sectors of the economy. How does one learn coding when AI is doing so much of the coding work? The debate over whether using AI in college is cheating or not, shows there are no good answers. If the outcome is the same or people cannot tell the difference, does it matter if something was produced or assisted by an AI? Building ‘meta skills’ is one solution, but then you run into the above problems, in which innate talent matters or the advice is too generic or vague to be of any use.
But I don’t want to dismiss all advice as bad. Advice by an expert in the form of individualized instruction, such as math tutoring or music lessons can be useful, provided one has reasonable expectations.
A useful heuristic I have found is that bad advice does not agree with empirical reality. For example, in regard to dieting, it’s often recommended to take a ‘diet break’ or ‘refeed’ in the expectation of raising one’s metabolism before resuming in the diet, so that additional weight loss becomes easier. In theory this is how it’s supposed to work. I have argued that this is a waste of time and money (if you’re paying for coaching). The existence of 80% of the US being overweight or obese is clearly evidence this cannot work. Sure, you will rev up your metabolism as a consequence of gaining weight and eating more food, but this will be negated by fat gain. So you’re just undoing your progress. Rather then getting a metabolic tailwind after resuming your diet, your metabolism slows down as if you were dieting again, but only you’re heavier.
I think the problem overall can be summed up: ‘Almost everything is bullshit’. This is not to say that people are intentionally giving bad advice, but that advice is nothing more than old wives’ tales or anecdotes. When you accept this fact, then the high failure rate of advice is a logical conclusion of this. It’s ironic to end this essay with advice, but perhaps the best advice is to not have any expectation it will work.