3 ms·

Taking a principle and lauding it makes not an insight. There is no rationale given for the 1% rule e.g. why not 10% or 0.1%? There is not statistics behind th

Taking a principle and lauding it makes not an insight. There is no rationale given for the 1% rule e.g. why not 10% or 0.1%? There is not statistics behind the 1% rule. Yes there are "the winner takes all" situations but the article neglects diversity and other negative feedback loops e.g. in fashion the winner may succeed only to be swept away by a counter-revolution trend. About the author: > I'm an author, photographer, and weightlifter. Most people know me because of my writing about habits and human potential. I've created this weird “job” for myself where I try to be an advocate for the world's best ideas.

dev1n 9y ago
https://en.m.wikipedia.org/wiki/Zipf's_law https://en.m.wikipedia.org/wiki/Zipf's_law
- dsacco 9y ago
  To expand on this comment, since it's just a (albeit helpful) link: the author's point is that competitive advantage follows a power law. When phenomena appear to obey a power law, the specific percentage increase does not need to particularly rigorous to illustrate the growth pattern. You could plot the two axes without numbers and it still basically makes sense. You're going to see a disproportionately weighted side and a long tail next to it. That doesn't mean the author is necessarily correct in the overall observation here, or that you shouldn't add further rigor to the analysis; it just means that for the purposes of an article without peer review, "1%" is probably fine. Power laws are very often fuzzy when they're applied (or perhaps rather, "observed") in practice.
alttab 9y ago
In multiple places, especially towards the end, the author touches on negative feedback loops as much as positive ones. That said - in a zero sum game for resources, a positive feedback loop is in itself a negative feedback loop for others.
simonh 9y ago
It's a rule of thumb, not a law. As the article points out repeatedly the ratios differ from one example to another as you can see from the examples given. The 'rules' are ballpark averages or easy to remember round numbers across many different situations. What's important is to understand the effects that lead to these distortions in distribution.
- nathanvanfleet 9y ago
  Yes, I find that it's really remarkable, but every time I take a handful of 4-6 jelly beans, there is a dominant color. Sometimes it's blue, sometimes it's red, but there is always a dominant color. It can be 20%-40% sometimes. Remarkable. I think that I might write an article about this to justify current economic trends and hardships...
  - simonh 9y ago
    You really think that each person in society has a random share of wealth? That each season a random team wins the Super Bowl? If that's not what you mean, I'm afraid you're not making your position very clear to me.
anigbrowl 9y ago
It's an empirical rule, with a fair amount of evidence to back it up. You may be interested in this recent paper, which argues, in contrast to prevailing economic models, that Pareto-optimal outcomes may be rooted in production functions: https://www.econ.ucla.edu/fgeerolf/geerolf-pareto.pdf https://www.econ.ucla.edu/fgeerolf/geerolf-pareto.pdf As that is an economics paper the author is naturally at pains to relate such outcomes back to real-world causes, so as not to get lost in a maze of abstraction. My personal take on this, as an amateur number theorist and geometer, is that real-world Pareto distributions are degenerate cases of ideal power laws whose specificity and distinctiveness are illusory. I am a neoplatonist so you might want to think of this as a suggestion to take a step back rather than to peer closer.

dev1n 9y ago

https://en.m.wikipedia.org/wiki/Zipf's_law https://en.m.wikipedia.org/wiki/Zipf's_law

dsacco 9y ago
To expand on this comment, since it's just a (albeit helpful) link: the author's point is that competitive advantage follows a power law. When phenomena appear to obey a power law, the specific percentage increase does not need to particularly rigorous to illustrate the growth pattern. You could plot the two axes without numbers and it still basically makes sense. You're going to see a disproportionately weighted side and a long tail next to it. That doesn't mean the author is necessarily correct in the overall observation here, or that you shouldn't add further rigor to the analysis; it just means that for the purposes of an article without peer review, "1%" is probably fine. Power laws are very often fuzzy when they're applied (or perhaps rather, "observed") in practice.

alttab 9y ago

In multiple places, especially towards the end, the author touches on negative feedback loops as much as positive ones. That said - in a zero sum game for resources, a positive feedback loop is in itself a negative feedback loop for others.

simonh 9y ago

It's a rule of thumb, not a law. As the article points out repeatedly the ratios differ from one example to another as you can see from the examples given. The 'rules' are ballpark averages or easy to remember round numbers across many different situations. What's important is to understand the effects that lead to these distortions in distribution.

nathanvanfleet 9y ago
Yes, I find that it's really remarkable, but every time I take a handful of 4-6 jelly beans, there is a dominant color. Sometimes it's blue, sometimes it's red, but there is always a dominant color. It can be 20%-40% sometimes. Remarkable. I think that I might write an article about this to justify current economic trends and hardships...
- simonh 9y ago
  You really think that each person in society has a random share of wealth? That each season a random team wins the Super Bowl? If that's not what you mean, I'm afraid you're not making your position very clear to me.

anigbrowl 9y ago

It's an empirical rule, with a fair amount of evidence to back it up. You may be interested in this recent paper, which argues, in contrast to prevailing economic models, that Pareto-optimal outcomes may be rooted in production functions: https://www.econ.ucla.edu/fgeerolf/geerolf-pareto.pdf https://www.econ.ucla.edu/fgeerolf/geerolf-pareto.pdf As that is an economics paper the author is naturally at pains to relate such outcomes back to real-world causes, so as not to get lost in a maze of abstraction. My personal take on this, as an amateur number theorist and geometer, is that real-world Pareto distributions are degenerate cases of ideal power laws whose specificity and distinctiveness are illusory. I am a neoplatonist so you might want to think of this as a suggestion to take a step back rather than to peer closer.