Hacker News new | past | comments | ask | show | jobs | submit login
Than Average (thanaverage.xyz)
326 points by dvt 20 days ago | hide | past | favorite | 208 comments



Humble suggestion to a great project: I'd put a different colour to a single 'person' of the crowd so to represent the user and to highlight in which group he is fallen. That changes for each answer obviously. Kudos for shipping this nice project.


I'd love for this to have rooms. Your answers could still count toward the global stats, but you would compare only with the group where it was shared.

I'd like to try these questions out in my friend circles (anonymously). Although obviously the results would have to be delivered afterward, since initially there's no baseline.


  74% of participants also think they dream less while sleeping than average.
  79% of participants also think they have sex less often than average.
  82% of participants also think they are a worse dancer than average.
  79% of participants also think they are more privileged than average.
  79% of participants also think they are smarter than average.
  78% of participants also think they are less extroverted than average.
Things we readily admit together because it's funny and makes us seem more human, or likable in some way.

  28% of participants also think they are less honest than average.
  34% of participants also think they are less kind than average.
  26% of participants also think they lie more than average.
  31% of participants also think they had a worse childhood than average.
  35% of participants also think they are more important than average.
  29% of participants also think they are less progressive than average.
  30% of participants also think they are less polite than average.
  33% of participants also think they are worse at keeing secrets than average.
Things most of you are afraid to admit because it makes you feel guilty of being a bad human.


Some of those questions should also have a “I don’t have a slightest clue” option.

On top of that I think there might be selection bias going on.


There is a pattern. It looks that cerebral introverts (or well: nerds) are more likely to answer these questions.


One of the main things I like about Quantic games (Detroit Become Human, Heavy Rain, Until Dawn) is the world stats that show what other people chose by % and how that matches the choices you made in the game

Its like reading Choose your own adventure Goosebumps books and seeing how many other people chose that adventure


The answers only show how the HN crowds consider themselves. The significant deviations of the answers from 50% are ridiculously stereotypical from men in STEM: wealthier, more progressive, have less sex, dance worse, etc


There's no "I'm about average" option, so the results will tend to deviate from average more than they ought to.


Some people have a strong desire to not be average, even if it means being far below average.


This website has been making the rounds on social media for a few days - it's not just a HN sample.

I agree it is still a heavily biased sample though.


The answers are definitely mostly from redditors and guys from here (many of whom are also redditors)


I mean your obviously right but I still think most of the massive deviations would still be there (like the lying one)


Forgive my ignorance - why wouldn't, "I'm average," not be an option? For multiple questions, I found myself feeling neither more nor less than average.


The problem with having such options in surveys is that everyone then picks it, even though it is statistically unlikely. So better to force them to make a decision instead.


I think four bins of below average, slightly below average, slightly above average, and above average would be reasonable. It would show how many people think they fit into the interquartile range versus the extremities.


But then it will just take longer. You'll still have to pick which side of the average you're on, plus how extreme (or maybe you'll just never pick the extreme).


I gave up at question 60 because it was taking too long.


I have the same issue with most of the popular personality tests. Many of my responses are either extremely neutral or could wildly swing depending on the week, month, year, etc.


Think of it as a bet. There's no harm in picking either above or below average when you're uncertain into which bin you'll fall, because those cases will even out in the end anyway. Sometimes you'll chose "below average" and actually be 0.1 standard deviations above average. Others will be the other way around, but for your return on investment those cases are not the ones that matter.

In fact, if there was a third "within +-0.3 standard deviations of average" you would probably be unwise to pick it, simply because "below average" has a greater expected value due to covering more of the range of possibilities, even when it's only true 50 % of the time.

Besides, being forced to really try to sense which way one leans is informative!


Also things like “better partner” are so conditional on who you partner with that a simple general better vs. worse decision seems impossible to answer in a meaningful way.


Maybe choosing a partner that will be happy with you is also part of being a good partner?


For a continuously measured variable like height or attractiveness or curiosity measured at arbitrary precision, your probability of being average is 0.


Average being -1 to 1 sigma, most people are just average.


> measured at arbitrary precision, your probability of being average is 0

You're talking about a site that currently measures with precision of 1 bit, in which case your probability of being average is basically 100% comparing to either extreme.


If you force someone, you truly measure them.


This. I closed out of the quiz when I saw that I couldn't answer "close to average, and without enough certainty or precision to guess above or below".


So people are insecure about dancing and sex. This all tracks.

I wonder if and how things cluster around age, income, country of residence, etc.

What I'd really be interested in is a set of tags people can click as far as "which of these do you strongly identify with?" And then just have some broad categories that are ostensibly independent such as "investing", "politics", "exercise & fitness", "video or board games", "charity/volunteering", "night clubs/partying", "theater & the arts", "sci-fi or fantasy", "religion", "travelling/siteseeing", "meditation" ...

I'd assume most to be total noise but there may be a few interesting strong correlations that could inspire further study.


Not necessarily. It's the average of people playing the game, not reporting on some survey.

I saw better than average at maths around 68%... and that might be true for the kind of people who play this game. Dancing and sex might be the same story (didn't see those questions).


Ah, this feels interesting. I would have said I’m a below average dancer, but above average person at dancing. The difference being whether I compare to people who I observe dancing, or consider everyone. Differently for sex, I assume that this is possibly bimodal (none versus some), so this could provide a measure of the actual skew in the underlying distribution (difference between mean and median) and not only the perception of bias.


> So people are insecure about dancing and sex. This all tracks.

I reckon this is due to only watching apex males perform those acts. As opposed to maths where you probably remember your classmates struggling and were actually tested for your ability.


I think there's also mixing up "less/worse than average" with "less/worse than my share", and an uneven distribution makes those pretty different. There are some very good dancers doing a ton of dancing. So if you dance, it's probably a below average dance even if you're a slightly above average dancer. It might also be why so many people think they lie less than average. There's a minority that just lies their pants off, so most of us are telling less than our "share" of lies.


There's a good if a silly factiod in a similar vein -

    More likely than not you have more hands than an average human.
That's because the average hand count is a bit under 2.


... thus highlighting that sometimes you really need the median instead of the mean to describe something


Expected value and 95% confidence interval is my favourite. Or just the confidence interval.


Or don't waffle about which single value to show and print a histogram/violin/box plot instead.


I think this experiment is about perception rather than finding the true statistical average. The sample space might not be balanced for studying poverty, intelligence, relationships, etc. but I believe that's not the point here. The test is trying to show you how is your perception of yourself compared to others who participated.

Take for example this question: "Are you smarter than average?" and (last time I checked) 72% think they are. Just imagine how the distribution looks like. By comparing this number to the IQ distribution (assuming correlation) which is a normal distribution, you could infer that either [0] a sizable portion of participants are wrong about how smart they are, [1] the sample size is too small, or [2] the people who've landed on the page are exceptionally smart!

[0] I have a "feeling" that this is the case!

[1] Maybe this can be taken into account with better statistical analysis.

[2] The assumption of participants coming mostly from HN, does not prove that this is the case.


There's probably a bias towards sightly higher than average IQ, because extremely low IQ usually means intellectual disability, and because IQ correlates with the ability to apply effort (at least at lower IQs). I also suspect but have no evidence that low IQ users tend to stay within walled gardens like Facebook more.

I think a lot of HN would be surprised at what an IQ of 100 looks like.


A remarkable number of them were around 50%. I guess we know ourselves pretty well! Although I wonder how many of those people are wrong, just counterbalanced by the other side also being wrong....


And the rest show a humorous bias, like "75% say they are less conservative than average".


Probably not a humorous bias because the audience of a site like this is not average on some criterias.


Well that's humorous to me. /shrug/


Assuming majority of inputs are from HN, that makes sense. HN demographic is higher education level and younger than a given average population, two demographics that correlate to “not conservative”


> HN demographic is higher education level

"Do you believe the sites you frequent are visited by higher educated, more intelligent people than average"

Nervous shuffling of little green ghosts to the left.


I believe this site’s main referer is HN

I believe HN is mostly working age English speaking computer professionals - especially American software engineers

I believe that American software engineers are broadly educated to an above average level


This could be truthful if there is a biased data source.


Indeed, I willingly believe that the answers are truthful and that the source is biased. It's just a humorous window into the participation.


Could also be true in an unbiased datasource in a skewed distribution.


"60% say they are more average than average"


Or just random. I noticed that in the handful of questions I checked, I often thought: I'm about average (e.g. in hours of sleep), but I had to pick. That might bias towards 50%.


Some of these questions produce obviously weird responses. E.g.:

> Do you have better music taste than average?

Of course I do. Logically everyone should evaluate their own taste in music to be better than average, since your own music taste determines what you think constitutes good music taste. Somehow, only about half of respondents selected that option, though. That's interesting.


I could imagine somebody who is pretty self aware could judge their taste in music to be worse than average, if they really only ever buy music from artists that they are later embarrassed by.

But in general, I think this would skew towards people having the preference for their preference.


> your own music taste determines what you think constitutes good music taste

No, your own music taste determines what you think is good music. Taste is not about holding your own opinion higher than others', it's just about what you like or not.

I still find it weird that it's a 50/50 split though. Maybe a lot of people aren't thinking so much about whose taste is better, but whose is more refined? Like how people who aren't into wine might not take their own lived experience as seriously as the words of an experienced sommelier?


If that's how you interpret the question, then the whole concept of having "good" or "bad" taste is incoherent to begin with. If it's just your opinions on what you enjoy while not harming yourself or anyone else, then obviously nobody's taste can be better than that of anyone else.

Granted, starting from this interpretation could easily lead people to interpret "good" as "refined"; I suppose that makes sense.


Many of us click randomly or invert our true response to throw off data collectors. This is pretty low stakes after all.


I wonder how much of the results reflects the biased group that is answering the questions.

For example, 75% said they were more privileged when I looked, but if they are all coming from sites like this and have good jobs, that might be quite reasonable.


Lots of bias in the sample set. I'm pretty sure I'm less religious, more privileged, better with technology, more liberal etc. than the average American (or even my friends, family, neighbors), but probably not so compared to the average HN reader who took the survey.


Only 19% think they are more religious than average. I find that very interesting. Not sure if it's an artifact of the kind of people who have taken the survey so far, a statement about how people want to be perceived in our culture, or a bit of both.


Theoretically it could also reflect the society. If 80% of people aren’t religious at all, then at most 20% are more religious than average.


It's been a while since this was updated (2014), but it indicates that over 50% of people consider religion "very important" in their life. Only 11% said religion was "not at all important."

So while you could imagine a society where 80% of people believe they aren't religious at all, the United States at least does not appear anywhere near that number.

https://www.pewforum.org/religious-landscape-study/importanc...


Is this website primarily polling Americans?

(You’re probably right anyways - there’s a reason I put ‘theoretically’ at the beginning of my first post)


> Not sure if it's an artifact of the kind of people who have taken the survey so far, a statement about how people want to be perceived in our culture, or a bit of both.

It could also be that the (loose, intuitively approximated) mode on some level of granularity (the individual category perceived as most commonly seen) rather than the median is going to frequently be the intuitive yardstick for “average”, rather than median (mean of a ordinal categorical trait with no unique obvious reduction to an interval-level measure has no coherent definition, so its not really a candidate average.)


Yeah, this one is pretty easily explained as a statistical artifact in this way. The more religious someone is, the more time they are likely to spend around other religious people (at church, with friends met via church, various knock on effects of these things). As you become more religious, this effect will tend to increase your estimate of the population's average religiosity.

This also explains a number of political effects, especially once you start looking for it in other high time commitment social activities.


Just a reminder for the readers:

1. Averages and medians have no reasons to be similar or even close.

2. "Averages" do NOT behave the same way at all in high dimension than intuition suggests. If you answer 100 questions it is expected that you will be an extreme outlier on specific dimensions even if you are average "globally".

Not that the website argues otherwise mind you.


It should be a game. Try to hit the center. 2 points for 45%-55%, 1 point for 35%-65%. And the last question should be: Do you think your score will be better than the average?


I'm impressed with how honest people are. Most answers were in the 40 to 60 percent range, which is about where it should be mathematically of course.

Two big outliers I saw was that apparently everyone thinks they have more common sense than average, and also more polite than average.


> Most answers were in the 40 to 60 percent range, which is about where it should be mathematically of course.

Not necessarily. For instance there's:

> 77% of participants think they give less to homeless people than average.

Say most people give $0 to homeless people, but 30% of people give some. In that case 77% giving less than average may be correct.


"Almost all of us have an above average number of legs."


Maybe. Body brokers are companies that own cadavers and sell them for medical research. The people who own these and similar companies have large numbers of legs and probably raise the average number of legs more than the few people who have one or fewer legs.

If we were betting on average leg ownership in the US I'd bet on slightly more than 2.


Your argument doesn't make sense.

Those legs came from people which had at most two of them. They don't manufacture legs or whatever.


But those people are dead. So if you count up all the live legs, which will be slightly less than 2*live people, and then add in the dead legs, this may be somewhat more than 2*live people. And since the humor in GP comment is attributing the "having" of the legs to the owners of the cadavers, the denominator remains 'live people'. Thus the average number of legs to "have" is greater than 2.


Why do you count dead legs, but choose not to count the bodies they're attached to?

Or if you chop them ... does the rest of the body cease to exist?


Because the dead person they were attached to doesn't answer polls, nor own much of anything.

But I suppose a live person can also own a bunch of torsos. The statement however only pertained to legs.


And if some of this chopping happens. Is it more likely that the legs or body is destroyed? I would guess that more of the bodies are stored than the legs... But now if we extend to organs, heads and brains... How do we calculate those? Like do we need complete sets, or something less? Or is sufficient mass enough to count as one unit.


Yes, the legs come from people who had at most two but the people who own the body brokers now have those legs. For simplicity let's say person X is the sole owner of a body broker that sells cadavers to medical schools. If a school calls up X they may ask how many legs he has, and may give an answer like "I have 100 legs" (plus the approximately 2 legs X is likely not willing to sell).


That raises an interesting philosophical question, do you have to be alive to retain ownership?


I would say no, but in this case I think only "attached legs" should count.


Depends on the average. In your example the median gives $0; the mean gives 'less than half of some'; the mode gives $0.


I know people learn this in middle school, but I've never seen an actual scholarly work call the median or mode "the average"--and I think I'd find it slightly misleading if they did.

"Typical", "likely", (etc) are all fine, but to me, "average" strongly implies "mean."


Scholarly works will pretty much never call the mean "the average" without clarifying what they mean either.

As for the implication, a statement along the lines of "the average person" will almost always be interpreted as median, not mean. For example, Jeff Bezos lives in Medina, WA, a remarkably anagrammatic suburb for this demonstration. You can probably understand why saying "the average person in Medina has a net worth of $65,000,000" is both somewhat misleading and remarkably unhelpful for pretty much any purpose you'd want that information.


People use "mean" and "average" in way that suggest they're the same, but one certainly doesn't have to say "average (herein defined as the mean)" because that's the overwhelmingly dominant meaning anyway.

Two random examples from other tabs I had open: https://direct.mit.edu/neco/article/16/3/477/6879/Mean-Insta... or https://www.cell.com/neuron/fulltext/S0896-6273(13)00798-8 Mean and average are clearly the same thing, but only via context.

In fact, I'd say that "averagING" (as a verb) almost has to be the mean; I'd find it deceptive if it were mode-finding.

You're right that people often conflate the mean and median, but I think that's more to do with underestimating the very, very long tails of some distributions.


You're replying to a perfectly clear demonstration that people often mean the median when they say average, and completely ignoring it.

What was the point of that?


That it's rarely deliberate.

I certainly agree that people mix them up. If you ask someone to quickly estimate the average/mean of something, you are likely to get some other location measure, like the median or mode. Some of this might be a heuristic: it's easier to mentally estimate the middle/most common item than it is to keep a running sum, and for many distributions, you get a similar answer. Some of it might be due to fogginess over the definitions. And, of course, these other summary statistics are often closer to what people regard as a `typical' value, especially if there's a long tail, so it

On the other hand, suppose we had some data like this

    X=[2, 2, 2, 4, 4, 8, 16]
If I asked you what the average of X was, the answer I'm expecting is "5.4", or perhaps "the median is 4", "the mode is 2", etc. I would be surprised if you just answered "2" or "4", and once I figured out what you did, I would find it annoyingly pedantic.

I've worked with a lot of different people from different backgrounds, and I don't think anyone has ever used (unadorned) average to mean something other than the mean. Perhaps your experience is different.


Ah.

Alright, this is rooted in a confusion between the normal use in statistics and the normal use in human speech.

Absolutely, average-without-qualifications means the mean, if we're going to be handed a list of actual numbers and expected to produce another number from them.

But especially for highly right-tailed distributions, when people talk about the average, they mean something more like the median, if there's a significant difference.

I would find it annoyingly pedantic if someone said the average net worth in Jeff Bezos' municipality was 60mm, if the median was closer to 0.75mm. If I was handing someone an Excel spreadsheet of tax returns, I would of course tell them that we needed the median. Or the mean but in this circumstance I wouldn't use the word 'average' because of the wide skew. If I did anyway, yeah I would probably get the mean back.

That's how average works in statistics, but not in common speech.


Whatever 'scholarly work' does or doesn't do, journalists certainly interchange them; sometimes leaving a hint like the implied distinction between 'the average <trait>' (mean) and 'the average person's <trait>' (median; sometimes mode for physical discrete things). Also, though, that can just be sloppy and have no intentional implication, misleading depending on the reader.


That precision of language is not the way most people would use that word, eg "a single value (such as a mean, mode, or median) that summarizes or represents the general significance of a set of unequal values" - https://www.merriam-webster.com/dictionary/average

"be conservative in what you do, be liberal in what you accept from others" - Robustness principle

It's good to be precise when communicating, but in most contexts you shouldn't expect it from others.


I understand that some sources say that one can use "average" to mean something other than the mean.

My claim, as an actual working[0] scientist who encounters those words a lot, is that nobody really does, outside of "Well, actually the average can be the median or mode too...."-type statements.

I'd put describing the mode as "the average" on par with giving someone a tomato and zucchini salad while describing it as "mixed berries." Not technically wrong but....

[0] Well, slacking scientist at the moment, but...


For the record I was assuming that "average" meant mean in my original comment.


On subjective qualitative questions, the average should probably be above 50% because everybody has different scales and are probably following their own scale better than average. eg. If you believe good driving means following traffic laws, you probably follow traffic laws better than average. If you believe good driving means low lap time on a race track, you might have a better lap time than average.


It says "average" not "median" so it's not necessary for the result to be 50/50 even if everyone was perfectly honest.


When it comes to personality traits/skills, average has to refer to the median. I can't think of a qualitative way to do it otherwise.


You can sort the results by agreement percent to see which questions are most and least skewed. Interestingly one of the most precisely balanced is "50% of participants think they are better looking than average."


77% think they are smarter than average.


It could well be true that 77% of people are smarter than average if intelligence is a heavily skewed distribution.


That's part of the Lake Woebegone effect.

For me it's different, I think I'm smarter than average 77% of the time.


I believe in this. Smarts are more like a distribution curve, even in a single individual.

There are areas I am in the 1 percentile, I think I can say without exaggeration. Those areas I seem to have insights in without effort, coupled with thousands of hours of focus.

But I experience a high cost of many things I do very very badly compared to almost EVERY SINGLE PERSON I know.

I am not alone in this assessment of myself. Friends & family all agree.

My many areas of remarkable incompetence, besides occasionally causing me great shame, make it very clear to me that everyone I know or meet has something to offer, at least to me!


There's so many facets to intelligence. There's a lot of areas that I'm more competent than most in and I learn very quickly. Then there's things like math where I have plenty of education, but the simplest real-world applications throw me for a loop. Add in frequent blonde moments and I can go from looking like the smartest person in the room to the stupidest in no time.


Smarter at what? Smarter at the things you are good at of course and there are lots of people who are not good at the things you are good at, so you consider yourself smarter than average.

If you asked people if they were smarter at everything than average what would they answer?


My definition of dumb is people who don't agree with me. So it's safe to say I'm basically Einstein ;-)


Most people think they are better drivers than average.


Right, but some of us know we are. :)


This would be more fun if this could be done with the median. Then the "true" answer to any question would be 50% on both sides. (For example, more than 50% of people have a lower income than the mean income, due to extreme outliers.) I suppose in lay language the distinction gets muddled.


These questions and some of the percentages really makes me wish I could see what other people think average is.

It would be interesting to see how often those who are actually less than average said they were more and vice versa.


This is cool. I adore the animation. I think you can tell how the data's a bit skewed by the HN audience so far as a result of 79% thinking they're more progressive than average. That question doesn't have the same type of subjectivity as how you smell, for instance, because anyone who's progressive at all in some way can correctly answer yes to it.


The one question I saw 50% on was asking people if they were better or worse looking than average. The most skewed questions/answers were people saying they were more perceptive, lie less, are more liberal, and give less to the homeless than the averages.

EDIT: you can see the full list of questions and answers as well, and it’s more interesting than just what I wrote above.


"Is your salary higher than the average salary" and "Is your salary higher than the average person's salary" are two different questions when you think about it. I interpreted all of these questions more along the lines of the second example, which meant I was basically using medians and not means.


> 23% of people say they're less smart than average [Paraphrasing].

That's interesting. Most of the other ones sit somewhere in the 40%-60%, but this is the only one I've seen where most people pick the "good" trait over the bad to such a high degree. I guess this says something about Hacker News.


New idea: this same questionnaire, but all questions ending in "average HN user".


on Firefox 88/Windows there seems to be a bug where the 'crowd' isn't re-flowed after window changes size, but the fonts/ui ARE being reflowed. This leads to a issue where the 'crowd' may sit over the buttons and make the UI unusable.


On Safari/iOS I can’t read the questions because the crowd is over a few words on each one


It's missing the "about average" option. Many of the questions I'd like to say that I'm about average, I think. Like, for instance, I think I'm an about average driver and I think I'm about average when it comes to optimism.


The 'confidence' question was hard to answer. In some areas I am confident, in others not so much. I think most people are like that. There were a few other questions that I found hard to answer as yes or no. Or maybe I am misunderstanding confidence or don't want to see myself as lacking confidence.


Reminded me Tom Scott's video on his survey with similar question and interesting interpretations: https://youtu.be/dcuNq3Bw9Xs


"75% of participants think they have fewer friends than average."

https://en.wikipedia.org/wiki/Friendship_paradox


I think it can apply to skills.

For example, if you go to a gym, you are probably among the weakest in the room. Simply because the strongest are more likely to be those who go to the gym the most and therefore, the ones you are most likely to encounter.

If you hit the dance floor, the same happens, and as expected, most people think they are below average dancers. In fact, the worst dancers may not dance at all, so you can't compare yourself to them.


The Friendship paradox has nothing to do with what people think: it's an observed fact and related to the way that node degree in social graphs work (or random graphs in general).


80% think the selection bias of this website is worse than average.


Is the population that visits your website more biased than average?


There may be a reverse-Woebegone effect operating here that tends to make people answer something other than what they might initially think.

Maybe that's the point.


Went through these, and I wish I had an "about average" response for many of them. Pleasant interface for this sort of thing!


Remember from somewhere : "You are never as good as they think you are OR You are never as bad as they think you are"

Good day !


> 81% of participants think they are a worse dancer than average.

Made me laugh. Obviously skewed demographics, but entertaining nonetheless.


I’m betting the results will be skewed by being posted here and being on the front page. HN can produce a lot of traffic, but doesn’t represent a good sample of the population.


Should be pretty easy for the site owner to analyze that.


> 50% of participants also think they ponder their own mortality more than average.


What a great way to get personal ad information.


This exact thing is in my idea file :D


fantastic design


This is amazing. We’ll done.


The first question I got already gives me pause:

> Do you have a better music taste than average?

What does this even mean? How can an opinion be averaged?


I had this exact problem.

> Are your feet better looking than average?

Ok, I understand what they mean but I have no opinion on that, just generally not finding feet attractive and I can honestly say I have never considered the question before. I consider it for a few moments and realize I have no values that would allow me to formulate an opinion on it. So then I fall back to, would an average foot lover find my feet more or less attractive than average? I still have no idea. So then I fall back to a meta strategy, ok if I am going to answer questions stating my opinion when I have no opinion, what should I do? Alternate answers? Always answer that I am better? Always answer that I am worse? I understand that they are interested in what the distribution is going to be after people answer the questions so someone hedging is not giving helpful information.

I decided that my strategy was to leave the website. I was curious if anyone else found being presented with a binary choice for an estimate at a continuous random variable off putting. I would have been much more willing to give a confidence interval around some abstract mean.


The entire exercise is an inversion. It's really asking you about confidence and insecurities


You don't need to average opinions. Just imagine you pick a person at random from your peer group and ask yourself how likely it is that you have better taste in music than that person.


Right, but what does that even mean? My taste in music is the best, if there were music that I liked better, that music would be my taste in music. It's just an odd concept, comparing a taste like that.


The concept that there is such a thing as objectively better or worse music is not, I think, a totally new idea.


you're not averaging actual numbers, its just "X than the average person"


Right, but comparing opinions in the first place is odd. Is this concept more common than I'm giving credit for? I would assume that everyone thinks they have the best taste in music - if they like something better that would be their taste in music.


Yeah i think you're just missing something here. Its pretty common for people to compare themselves to others like this.

Not all of the questions are equally qualitative. Some are "i give more money to the homeless than average". You could theoretically measure this.

The purpose is probably just to understand peoples perceptions of themselves. (For example), i think i have better than average taste in music... after all, i meet artists and spend my free time going though new albums to find the next needle in a haystack.


No doubt mine is best ;)


So sad, pretty much all the comments in this thread are "it would be better if ...", "I think that ...", "I don't think that ..."

Well, this is your chance then, buy a domain name and set up your website, shows us how much better you can do it, or, you know, just shut up.

I personally found the site amusing and interesting, I'd love to come back and check the statistics once a lot of people have gone through it and the sample size becomes significant. Props to the authors for developing their idea and delivering it to the world in such a simple way that I only had to click a few times to play with it, for free :D.


That is a very common but misguided attitude towards criticism. It is fairly easy to determine you dislike something without necessarily knowing how it could be improved. Knowing how much people dislike something is useful information in regards to that thing. Maybe the author doesn't care, kudos to them in that case. There is nothing wrong with creating something which strongly divides people into love/hate categories. In any case, it is presented here for an audience and the audience can choose to respond as they see fit, they do not need to acquiesce to something distasteful to them.

From a cusiosity perspective, the aversion some people (including me) seem to have towards it makes me more curious if it suffers a strong selection bias. I tried a couple of questions but couldn't actually answer a single one. I was simply incapable of formulating an opinion on the question posed. I was even incapable of choosing one of the options simple to progress. That was what I found most amusing in it: my inability to answer a seemingly easy question.



Whoa, I wasn't aware of that.

Thanks for pointing it out.


You're welcome! thanks for taking a look :)

(Your GP comment was the top subthread when I posted that, which is what made me think of it.)


While I do disagree with the chosen wording (using 'a cool new feature would be to...' sounds a bit more friendly, if you will) I don't think that anything is said in bad faith. This is a passion project for the creator and he most likely wants to improve it. He should still follow his vision of what he wants this website to be, but that doesn't take away anything from the suggestive comments posted here.


"I don't think that" just because someone has an opinion that varies with that of the author, that they necessarily should try to outdo them.


I think what parent is getting at is that criticism and nitpicking take no effort. Obviously that's not to say nothing should be criticized.

There's a similar, ruder, expression in fashion communities: talk shit, post fit.


> criticism and nitpicking take no effort

Sort of unrelated aside. I recently started working with a new company and team, in a new ecosystem. I’ve found it difficult to ramp up and get productive. But the thing I’ve found most difficult is getting familiar enough to provide meaningful review.

It can take me hours of exploration to fully understand what might intersect with single line code changes. Often I’ve come back with a minor change request, even simply for clarity.

Where it relates is obviously these things are critique, but even picking nits has been hard. Not that I’m even predisposed to do so, but I’m trying to get familiar and provide value.

It turns out that effort has indeed provided value, I’ve caught mistakes just by being thorough and leaving as few stones unturned as possible.

And it’s reinforced to me that criticism and nitpicking take a lot more effort than might be obvious. Sometimes it’s quickly available based on experience, but even then that’s based on prior effort. I’m not saying it’s always welcome or appropriate. But it’s surprising to me to hear people seemingly reflexively say it’s easy.

A whole discipline (QA) is dedicated to it, and integrated disciplines (TDD, property testing) are still suspect for many.

Again, not saying it’s welcome or appropriate here to criticize or nitpick. I’m holding my tongue on several bits of feedback I wanted to say upon using the thing. But I think it’s a disservice to both the effort and value of critique to frame it as effortless and implicitly valueless.


43% of the people here thought the same, that negative criticism should not be shared here.

There is always room for feedback, but please make it constructive.


100% of me disagrees with you, it was constructive he just wasn’t nice about it, there’s a difference.


> 70% of participants think they procrastinate more than average.

This is what I call selection bias.


Absolutely, across the board, but I kind of enjoyed the meta-game of "guess how people selection-biased to be like you will answer".


Was thinking the whole thing stinks of selection bias but this one is the epitome hahaha


I have an above average number of hands: 2 (like most people)

Or in other words, it’s not contradicting if you’re in a 70% group of people who said some answer.


The flaw in this survey is that they do not specify what type of average they are referring to.


Average = mean.


> Average = mean.

No, Average is any measure of central tendency, which include all of the median, mode, and the various means (harmonic, geometric, arithmetic,...), as well as some others.

The mode is the most broadly applicable, since (at least if you accept multiple values) it is well-defined on any, even merely categorical, data. The median requires ordinal data, and the various means tend to require at least interval-level measures (some of them require ratio-level measures.)

The arithmetic mean seems to be the most common grade school mathematics “average”, possibly because its the one that does the most to exercise basic arithmetic skills. But its rarely, when distinct from the median and mode, the most useful, and often doesn’t match the intuitive understanding of “average”.


No, it can mean:

Mean, mode or median. These are mathematically distinct figures and they can be extremely different numbers.


The mean, median, and mode are distinct, but mean is the average. See: https://en.m.wikipedia.org/wiki/Mean which says:

> In descriptive statistics, the mean may be confused with the median, mode or mid-range, as any of these may be called an "average" (more formally, a measure of central tendency). The mean of a set of observations is the arithmetic average of the values...


“the mean is the arithmetic average” is nonsense verbiage. “The mean” isn’t even well defined, there are a number of different means.


Aritmentic mean is defined at the top of the linked Wikipedia page to be the average as we know it (sum/count)


“arithmetic mean” is well defined (“the average as we know it” is not a particular sensible phrase to use in a subthread where the point in debate is specifically whether the average refers to one specific thing.)

That doesn’t make the word salad sentence used later in the article— “the mean is the arithmetic average”(emphasis added)—make any sense. Were it “the average is the arithmetic mean” it would merely be wrong, but at least make sense (and it would be making the same incorrect point it was being cited to support), but flipping “mean” and “average” so it claims that the former is the “arithmetic” version of the latter goes beyond wrong into being a pile of words that are individually sensible but don’t mean anything, right or wrong, as assembled into a sentence.


The mean is what is often meant by the term but not always. "How to lie with statistics" is a great book and I highly recommend it to anyone still confused by such terms -- or anyone who ever sees, say, ads.

"There are three kinds of lies: Lies, damn lies and statistics" -- Mark Twain


The average person has less than 2 hands :)


And just under 1 testicle.


Brilliant :D


It would be helpful if the site could define what they mean by the average. The average visitor to the website thanaverage.xyz? The average American? the average person around the world? It seems like it would be more interesting to make a guess about the average visitor to the website, but it's not clear if that's the case or not. Compared to the average American or the world, the numbers for this small segment of people who would visit this website would not reflect the average.


But it does define.

"Imagine you are in a room with 100 strangers, Imagine they're similar to your peers and neighbours."


But then aggregating the results together doesn't really mean anything. 100% of responses could be "I am more XYZ than average" and they could all be correct.


In the "about" section of the webpage: "ThanAverage is a small unscientific investigation into how we value and compare ourselves to each other."

It's just a fun website... don't think too much about it.


You realize you’re on HN right?


Do you overanalyze and intellectualize more than average?


Depends on who we consider average I guess :)


Still not clear IMHO. What is "peers and neighbours"? Same country as me? Same age as me? Same gender as me? Same socio-economic group as me? If I take all that to one extreme, then I have zero peers and neighbours, because nobody is in all the exact same categories as me. If I take it to the other extreme, then all other living (or dead?!) humans (or lifeforms?!) on Earth (or elsewhere?!) are my peers and neighbours.


So "more than median" it is.


Important to realize that the average taker of this survey is likely nowhere close to the average person on quite a few of these qualities, or even the average internet user. It's going to skew more tech savvy / progressive / intelligent / etc -- the Hacker News bump alone renders the "insights" you might glean from the results almost meaningless.

For example, that 76% of takers think they are more privileged than average is probably an underestimate, not an overestimate as you would first believe...


> 61% of participants also think they are better at maths than average.

Yeah, given my discussions about math with people in non-technical fields, I think this answer is very skewed.


I'd assume the majority of people think they're bad at math based on passing observation.


There must be a term for this, but I imagine no one is average across all attributes. Curse of dimensionality maybe?


“Average pilot fallacy”? https://www.thestar.com/news/insight/2016/01/16/when-us-air-...

When designing cockpits using the average dimensions of pilots the US Air Force found exactly 0 pilots who were average in all 10 dimensions. Even just using 3 dimensions only 3.5% were considered “average”. Take the average dimensions of a human hand and draw that hand. Almost nobody has the dimensions of that hand.


That’s so cool: just occurred to me it’s like the inverse of crowdsourcing. No one guesses the right weight of the cow, but the average of all guesses is almost exactly the right weight.


Well, here you never know the weight of the cow, per se. If the weight-of-cow question were asked, it would "If you were asked to guess the weight of some cow, do you believe you would overestimate it?" And then you find out, that, 63% other people share your self-evaluation, or else do not. There is no info about any actual cow.


On HN, there was recently an article [1] on this topic. You may also be interested in the discussion [2,3].

[1] https://www.thestar.com/news/insight/2016/01/16/when-us-air-...

[2] https://news.ycombinator.com/item?id=22956832

[3] https://news.ycombinator.com/item?id=11230287


I don’t know a term for it, but here's a relevant blog post:

https://www.johndcook.com/blog/2011/09/01/multivariate-norma...

> For a multivariate normal distribution in high dimensions, nearly all the probability mass is concentrated in a thin shell some distance away from the origin.


It asks you to compare yourself to your peers, not to the global population, so the results should still be close to 50%, even if it does say "average" and not "median".


I don't personally consider "HN readers with tech jobs" to be my peers. I'd instead pick my friends, family, coworkers and extended social circle.


Most of the people who I imagine to be my "peers and neighbors" probably wouldn't visit a site like this or spend as much time on Hacker News as me.


You aren't my peers, guy!


That's certainly an interesting point. The questions have ~9K responses at this time; I wonder how many of those can be attributed to an hour of being posted on HN.


Also 81% have less sex than average...as well as upper 70% have less friends than average.


Fewer friends than average actually kind of makes sense:

> The friendship paradox is the phenomenon first observed by the sociologist Scott L. Feld in 1991 that most people have fewer friends than their friends have, on average. It can be explained as a form of sampling bias in which people with more friends are more likely to be in one's own friend group. Or, said another way, one is less likely to be friends with someone who has very few friends.

> [...]

> In spite of its apparently paradoxical nature, the phenomenon is real, and can be explained as a consequence of the general mathematical properties of social networks. The mathematics behind this are directly related to the arithmetic-geometric mean inequality and the Cauchy–Schwarz inequality.

Strangely, however:

> In contradiction to this, most people believe that they have more friends than their friends have.

https://en.wikipedia.org/wiki/Friendship_paradox


I wouldn't say "progressive" is a word I'd associate with the HN demographic, based on some of what I've seen here. There's a definite libertarian skew, but that seems to fall along both sides of the left/right political spectrum.

As to whether the typical HN user would answer "yes" if asked if they are more progressive than average, however, I do also suspect they would answer "yes."


Yup. And given that an almost exactly inverse group said they're more conservative than average, that's a hot take on how tech libertarians view themselves if given a binary choice.


I wonder what HN's result would be for "I am more/less likely to rush to HN to post banalities about the survey sample"


"Please don't sneer, including at the rest of the community."

https://news.ycombinator.com/newsguidelines.html

That guideline is not there because we don't know how annoying HN comments can be, but because we do know. Reacting by adding snark or denunciation to the threads just makes things even worse.

The only response that really helps, in reaction to bad or unsubstantive or obvious comments, is to post good, substantive, non-obvious ones.


I wish I had seen the deleted comment. My take was obvious, but it was intended as a compliment to the creator of the piece, not to incite argument. Still, I'm curious to know why I got trolled for it.


> Are you better at washing up dishes than average?

What does that... even mean?

That I do it more often? More often than my partner? More quickly? More sanitarily?

I never in my life thought washing dishes was something you could be better or worse at, any more than tying your shoes...

I'm guessing other people are confused as well and just answering randomly because it's the only question I've seen with a perfect 50% split.


Summers during college, I worked as staff at a summer camp, where part of the "camp experience" was that the kids got to (had to) help wash the dishes are meals. This was usually done by a rota, so a different subset (4 to 5) of the campers (mostly 6-12th grade, but there were adult camps too) helped after each meal, and the staff took turns supervising this.

As such, over the course of a few summers, I have watched 100s (18 meals/week, 10 weeks/ yr, 3 years) of discrete groups of people of varying ages "wash" dishes and can confirm that there is a wide range of dishwashing skills across the following criteria:

- cleanliness (what % of dishes had to be sent back for rewashing because they were too dirty for the sterilizer)

- speed (fast groups would be done in 20-25 minutes. Slow ones could take an hour+, mostly because they would spend time complaining or arguing rather than actually washing anything)

- how much of a mess was made (sink water splashed on the floor, mostly) during the cleaning progress

And our dishes were unbreakable, but if they hadn't been, I'm sure "breakage" would have been a metric too.


Are there any anecdata you noticed in the trends there that you'd care to share?


Sometimes when I go to a friend's house, they have dishes in their sink that have obviously been there for a number of days. Conversely, I have had roommates who will cook a complex meal, dirtying many dishes in the process, but still have very little cleanup after the meal because they clean efficiently as they go (a practice I try to emulate to varying success). I would say the former group is not very good at washing up their dishes, while the latter is (from my own observations, at least) better than average.

edit: this is also one of the easier questions to actually have a good sense of where you stand. I don't know how often most of my peers cry, for instance, but I do know how often I visit someone's house and they have dirty dishes lying around, or they leave dishes from a dinner party or get-together in the sink for the next morning, etc.


Have you ever lived with other people? Some of my housemates of uni days would leave large chunks of semi-masticated food on all the items of cutlery. Others would wash greasy frying pans before wine glasses.

I think the problem with this question is, you either think you're better at washing dishes, like me, or you don't care, like my housemates (and possibly you). Who thinks they are worse at it?


I've seen way too many plates that have been "washed" yet have grime solidified to them, or worse, still damp on them.

Some people really can be worse at washing dishes. But I'm not sure how I'd define "good" or "average".


> Based solely on your own instinct, perceptions, and self-reflection, answer the following questions

My gut immediate instinct says yes, for me.

It's hard to articulate this stuff. As humans, we're pattern matching machines. Try not to think too hard about it.


Why can't I answer that I'm the same as the average


Why, is it so fucking hard to place 101 points to be able to pinpoint where exactly you stand (and then show a graph of responses), instead of this beaten-to-death-horse more/less than average trickery (probably done solely for some BS psychology paper)?


Wow you got angry. This is not a tone people expect on this forum.

https://news.ycombinator.com/newsguidelines.html


No, I wasn't angry, even though I can see how that could come off as such. But I do feel like it's yet another highly-upvoted time waster with no substance to it on the front page.


While not exactly what the article is suggesting, a fun thing to know is that in a room of 100 people, with normally distributed IQs. If you want to have a 50% chance of being correct that you have a higher IQ than everyone in the room, you need an IQ of about 136.9.


While I realize the purpose is to force a choice, I'd be interested to see the results if there was an "I don't know" option.

There were some questions where I'm very confident that I'm above/below average, and others where I basically have no idea, but I had to pick something.




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

Search: