Hacker News new | comments | show | ask | jobs | submit login
[dupe] Talent vs. Luck: the role of randomness in success and failure (arxiv.org)
38 points by manusachi 3 months ago | hide | past | web | favorite | 27 comments




N agents are placed in a square grid, each with T_k "talent", chosen from a Normal distribution with mean, m, and variance, v, (chosen to be in the neighborhood of [0,1], e.g. m = 0.6, v = 0.1) and C_k initial capital, with all C_k chosen to be the same initially. N/2 "events" are also placed on the grid, with p of them being "lucky" and (1-p) of them being "unlucky".

The simulation is run with the "events" wandering around randomly. If an event "hits" an agent, the agent doubles their capital (C_k) with probability T_k, trying to encapsulate the idea of "when preparation meets opportunity". In other words an agent's capital doubles proportional to their "skill" if a "lucky" event hits them.

An agent's capital is halved if an unlucky event hits them.

After running the simulation for a certain amount of time, a Pareto distribution is observed for the distribution of capital (C_k). That is, with an initial distribution of "skill" as Gaussian/Normal, the wealth distribution that results is power law.


Consider N agents, each with initial equal capital C. At every beat of the clock, each agent gives 1 unit of capital to a random other agent (or does nothing if their capital is zero). After running the simulation for a certain amount of time, the wealth distribution is exponential.


"The Drunkard's Walk: How Randomness Rules Our Lives" by Leonard Mlodinow seems topical based on the abstract. I would recommend it as a fairly quick read that presents a perspective you might not have thought much about.

I avoided it when I first came across it because I was worried (based off of the subtitle) that it'd just be fodder for learned helplessness. It turned out quite good though, and anything but (unless you're a movie studio executive).


You may also attribute the possession of various types of intelligence to luck and you are done!

Richard Wiseman argued quite convincingly that luck is mostly self-made.

http://richardwiseman.com/resources/The_Luck_Factor.pdf

In the famous picture-counting experiment, people were grouped into two categories (those who were considered successful/lucky and those unsuccessful/unlucky) and given the the task of counting the number of pictures in a newspaper. People in the "lucky/successful" group found the correct number in less than half the time it took people in the "unlucky/unsuccessful" group.

There was a small footnote in the front page of the newspaper: "This paper has 47 pictures." Guess which group was more likely to notice that.


If you specifically select a bunch of successful people and pitch them against a bunch of random people then of course the successful people group will perform better.

I'm sure though, that if someone finds a group of successful people (even billionaire level), I can easily find a group of poor/unsuccessful people who will beat them at any specific task.

It's no coincidence that people often complain about their boss being an idiot. It's probably true. People don't get into leadership roles because of their talent or intellect - They get lucky; right company, right time.


Perhaps. Some bosses are idiots for sure, I've had a few in the past - and even now I'm my own boss :)

We need more studies to back up this claim though.


Or perhaps it's the "Peter principle" at play.


Cue the people saying "yeah but we all make our own luck" or "luck favors the prepared"

Which misses the point. Of course you have to be good AND lucky to succeed. But not everyone can be lucky. So think of that the next time you're looking a few rungs down the ladder.


You miss the point of people saying that "we all make our own luck". It's not about whether you get lucky or not, but the mindset with which you approach the world and how that mindset helps you or not. I wrote about this here if you want a more extended argument https://github.com/SSYGEN/blog/issues/38


I have to read your essay in more detail. I see a lot i disagree with. But if the gist of what your saying is is "ignoring luck will help you persevere" I can see the logic.

However, I'm not advocating a cynical mindset. I'm advocating a compassionate one, and a realistic one. If you don't think luck matters, you can delude yourself into thinking the people below you didn't work hard enough, or that you must be especially talented and driven to have gotten where you are.

I've also seen people hurt themselves with that mindset. The guy who risks everything on some business venture that's doomed to fail, yet he thinks he can beat the odds because he's smarter or more hard-working.

How about having an absurdist approach towards luck. "I'm probably going to fail, and I'm going to do it anyway." It's honest, it's compassionate, and it doesn't discourage you from trying the impossible. That's how I think, and it's worked well for me.


I think the distinction between how you see yourself and how you see others is particularly important, and I find that it's often conflated when I have arguments about things like poverty.

Attributing things too much to luck in your own endeavours might lead to learned helplessness, among other things. Whereas assuming that the bum on the streets quite possibly got into his predicament because of bad luck seems like a safer and more compassionate assumption to make.

But factually speaking, I'd say much more than we'd like to admit is out of our control, which is one of the reasons why I am in favor of societies with strong safety nets and hefty base-line of compassion.


Side note: it's pretty clever you're using GitHub's issue tracker for blogging. Besides it being weird for navigation, I suppose it has all the features you'd need such as tagging, comments and search.


Github issue tracking doubling as a personal blog = effin genius


Central Limit Theorum tells us that luck will be normally distributed even if you’re sampling from people who work hard. Sure hard work may shift the mean to the right compared to a random sample of the general population, but you still get a normal distribution. Now if being lucky is being in the top 2% from a random population, how much does hard work shift you to the right? You need to shift this distribution pretty far to make the case that work alone can overcome luck.


The sampling distribution of the _sample means_ (averages) will be a normal distribution.

The population distribution need not be normal.


I don’t believe we all make our own luck nor do I discount the role of luck in my own success. However the model used in this study is very weak and does not supply much evidence, of you ask me.


FTA "almost never the most talented people reach the highest peaks of success, being overtaken by mediocre but sensibly luckier individuals."

It's also interesting that regardless of anything, in the end your genes and where/when you were born determine pretty much your whole life, and that's just luck. In the end success attribution to anything but luck is just ego.


> "almost never the most talented people reach the highest peaks of success, being overtaken by mediocre but sensibly luckier individuals."

One can be talented at many different things. Perhaps the people at the top were talented at a less obvious but more relevant skill than those who didn't "succeed".

Looking strictly at monetary success, people who succeed purely and obviously by luck (lottery winners) generally do not stay successful very long. They quickly lose their money.

AFAIK, the bankruptcy rate for people who get rich via working on something is far less than people who were randomly assigned riches. This would seem to imply that the former are doing something differently.


> in the end your genes and where/when you were born determine pretty much your whole life

Interestingly, your genes arguably determine (and at the very least heavily influence) your talent also, implying that having talent itself is a result of luck.


Sometimes I think the same. But then again I think that this is a very depressing view because it’s somehow deterministic: there’s absolutely nothing in this view that can be attributed to yourself, because either it’s in your genes or the place you were born.

That’s why I don’t like that thinking. I want to believe that you’ve were given a spectrum from a start and that you can move in that spectrum if you want to by exposing yourself to luck (you’ll probably say: depending on your genes...)


If everyone in the world is an equally hard worker, then yes, all success is luck. But that's just not the case.

Anecdotally, I knew two identical twins who were both obese. They both tried to diet, but one always relapsed, while the other got serious and lost weight. Ultimately only one of the twins was "successful" at getting down to a healthy weight, while the other "failed".

Is it just luck, then, that one of the twins was born with more mental fortitude? Is that your logic?


It's interesting that 10 years ago, people avoided this subject completely - The authors of such papers would have been accused of being jealous and lazy.

Now, the idea that we are not a meritocracy is basically common knowledge, it seems that more or less everyone (even among the rich) accepts that this is the reality. I don't think that society has ever been in such a state of economic self-awareness before. Unfortunately, all this doesn't seem to change people's attitude towards wealth; if anything, the rich are getting even richer and the poor are literally being wiped out in the opiate epidemic.

It seems that luck plays a bigger role than ever and yet the consequences of winning or losing are becoming more extreme at the same time.


TL;DR: Talent (inputs) are normally distributed but wealth (outcomes) have a power law distribution. So talent doesn't determine wealth. What does? Luck. (They do some simulations.)

I wonder if they're measuring the right kind of input though. IQ is distributed in a bell curve but that is one input. There is hard work, common sense, ambition, and energy. I wonder if you have a multi dimensional bell curve as inputs, you may end up with a power distributed output (all the elements have to line up for success).

Although, as someone pointed out, talent too is often a matter of luck.


> Talent (inputs) are normally distributed but wealth (outcomes) have a power law distribution. So talent doesn't determine wealth.

It seems to me that if the most talented are the most wealthy (etc. for the remainder of the people), then even if the distributions are not the same, talent can still determine wealth (?)


The (flawed) assumption that hard work inevitably leads to success is perhaps the biggest reason why capitalism is so successful.


If by successful you mean effective for the benefactors of luck to entrench their advantage at the expense of the less lucky, then I agree.




Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: