Oh god, this reminds me of the time I told Bram Cohen that Bit Torrent would never work. Of course, I just didn't understand what he was trying to do. Face, meet egg.
I was once superficially polite to the founder of twitter, while laughing patronizingly on the inside. "So, it's like chat, but with more restrictions and fewer features? Completely unmonetized, and the name literally means ignorant babbling? Huh, cool. Hope it works out for you!"
Comments on HN can be enlightening as to how people think. Never look here for validation of innovative ideas though. It's just not meant for that kind of discussion.
Oh my God, but without HN, how will I validate my existence?
... said me, while sipping expensive wine out of a crystal chalice that came from the 9th century while sitting on a throne made of the finest organic pasture raised bovine hide, while smoking a Stradavarius (the violin, not the cigar).
I find that's how HN is in general. There's a Rust contingent. There's a general FP contingent with a Haskell flank. A few years ago there was a Clojure Independency Party that provided things. I'm a member of the ArangoDB Activists Movement. The noisiest group right now is focuses on Machine Learning with NN. Fun times. Good way to get exposed to different ideas.
There's a certain amount of insane luck involved in any startup succeeding; you have factors like financing, broader industry investment trends, initial financing that involve just blind luck that is out of your control. To say nothing of all the ways you could control but ultimately screw up, in some way that ultimately kills the company.
Add to that some idea that is not normal and involves actually changing the way society thinks and acts. Or some idea that runs headlong into a highly regulated industry and involves a gamble on winning (zenefits, Uber, lyft). Calling these ideas crazy is basically stating the obvious. The unnormal result is that they actually succeed. And good for them for doing it, and I'm not discounting the skill and ideation or how well these people "earned" the success. I'm just saying let's not discount that they were still lucky and that the ideas were crazy -- this is statistical fact.
HN/Slashdot commenters can be valuable for pointing out what's technically wrong with the product. They also tend to put too much weight on product's shortcomings. However, they are blind to what's right about the product or discount the product's strengths for the non-geek consumer.
However, that doesn't mean that average consumers are ultimately right either. On the one hand, they don't have the technical understanding which often skews the prediction towards market failure. Their naivete just lets them try the product and vote with their wallets. On the other hand, that lack of knowledge leads to unrealistic hypothetical products... e.g. "if they can put a man on the moon, why can't my smartphone battery last 3 months and project a 4k video with 5000 lumens on my living room wall?" (relevant XKCD: https://xkcd.com/1425/)
There are so many comments on hacker news of the flavor: "Why would anyone use this when you can build it yourself in five minutes using x, y and z open source tools" completely ignoring how much knowledge you have to have retained in order to build it in five minutes.
Any time I read someone saying something like that I internally replace the subject with food. "Why would anyone buy food professionally made at a restaurant when you could make something similar at home?" Because I want something nice and I'm willing to spend some cash to get it. C'mon..
Counterexample by way of metaphor: I used to feel the same way about restaurant food, but then my cooking skills increased to the point where I can make most Italian or French food just as well as if not better than most restaurants I've been to.
Now, I rarely if ever go out for French or Italian any more.
Aside from pommes-frites and crepes, finding French food at the bottom of the price scale is pretty difficult.
And I wonder why that's true. I mean, in Los Angeles, I can find food from other great culinary traditions at both the high end and the low end of the price scale: Korean, Mexican, Chinese, Thai, Indian, Peruvian, Vietnamese, Middle Eastern -- even Japanese, which, considering general price levels in Japan, is surprising.
But French food? It's almost always some high priced, sit-down place with a valet out front. There don't seem to be a lotta reasonably/low priced French food joints out there.
(Even stranger is the price of "British" food, usually found in pubs: always seems overpriced. And British is not exactly known as a great food tradition.)
i love that comparison. also reminds me of the idea from economics that even a country that produces "widget x" domestically can in certain situations still come out ahead by importing some units of "widget x" from another country.
sometimes i'm better off buying takeout food from a restaurant so i can spend that time wrapping up some work task so i can just forget about it. i can bill a little more time that day. and i'll sleep better that night. it's worth it.
What I would find interesting is not just the positive\negative ratio of comments for successful projects, but also a comparison with less successful\failed ones.
Do successful projects receive more positive comments?
There doesn't appear to be any justification given for the eight posts selected for analysis. Given that nearly everything fails, choosing seven well recognized successes and one started that scaled quickly but failed seems an interesting methodological choice. Neither to they present any evidence that anyone is, in fact, obsessing over HN comments. The whole thing is "don't let the haters get you down" tweet expanded endlessly. With a infographic.
One can admire the article as a piece of analysis shaped SEO-bait, even as one despairs over the value it subtracts from the world by existing.
It would be nice if the Rust evangelists could show a little restraint. When a Rust article is posted (multiple times a day) and people disagree with its greatness there is no need to post even more articles about Rust closely related to the previous stories.
I'm not sure how valid this premise of this article is.
I would imagine that most startup founders probably spend maybe 10-15% of their day worrying about HN/Product Hunt comments. I forget who said it (maybe pg?), but HN isn't meant to be used as support board, especially for YC companies. Of course there are exceptions to the rule, especially if there is malfeasance or fraud involved.
This reminds me of something Joel Spolsky said on a podcast several years back. The gist was many newer ideas for startups that turned out to be great, were first met with criticism or labeled bad ideas. Because if the new idea was too obvious, it was probably already done by someone else. (At least that is what I think he meant. Wish I could find the show notes.)
Found it: https://stackoverflow.fogbugz.com/default.asp?W25795 Podcast 025. It is a good one, with Jeff, Joel and Steve Yegge. Joel says: "For a startup to work, it has to be an idea that is not very convincing. It has to be a completely terrible idea."
This really just puts more strength in the idea that success is largely luck-based.
No one can accurately predict success. Not specialized forums, not specialized workers in the field, and not even VCs whose sole incentive is to make money by investing in great ideas.
Even YC says as much that they have no clear idea as to what will be a success, and their strongest signal is the strength of the founders, and that there should be co-founders. I'm not sure how their experiment went with inducting people into YC just based solely on who they were, without an idea, but that's just further evidence that even people obsessed with finding success basically have no clear idea or signal as to what makes a success.
I agree with some of the sentiment but I feel like saying "largely luck-based" frames it as being mostly out of our control.
Yes, there are absolutely economic/industry issues that are completely out of our control and hard to predict but I believe a major part is the strength of the team to focus, execute, get along and continue when sh*t happens, and adjust as their Vision hits reality.
I'm not sure what the division is between luck/timing vs team/product vs all the other things but it's hard to accept that it's "largely" out of my hands. Maybe you're right but I hope not.
Assuming that the minimum threshold for any successful company are things like "strength of the team to focus, execute, get along and continue when sh*t happens, and adjust as their Vision hits reality", then I do believe that everything else is out of your control.
If it weren't out of your control, then wouldn't someone have already come up with a more surefire way of creating a successful company than what we currently have? The smartest people are here in Silicon Valley, all trying to become billionaires and yet only a very small percentage actually succeed, and there's no way to accurately and consistently predict before it happens.
Luck vs. skill is a false dichotomy. When people (or products, startups, organizations, governments, etc.) are successful, it's usually because they had developed very good skills at a.) putting themselves in situations where good luck is going to happen to them, b.) being able to continue doing so until good luck does occur, and c.) positioning themselves to take advantage of that luck when it happens.
I agree with the gist of the story, but I find the example poorly chosen.
The launch post regarding AirBNB set a tone with it's title that set people up to be critical..."Sleep under my kitchen table.." Even for those that understood the tongue-in-cheek there, the title is self deprecating.
I suspect the thread would have had a more positive tone with a different title.
Ugh. There really should be a logical fallacy around the fact that Hacker News was incorrect about Dropbox and other companies, and how it's not a license to ignore comments altogether. I really wish people would stop referring to the Dropbox story on HN as an example of "ignore the haters because they are wrong." Dropbox, and the other startups mentioned in the post, are outliers, not typical examples, and also cases that people have learned from since 2008.
Agreed. And technically, if the author wanted to make a useful point, they should have looked at a random sample of startups that were posted on HN. See if HN was generally correct or not. This is just looking for HN to be wrong.
Agreed. There's major survivorship bias in the author's selection.
I'd love to see an analysis of all the "Show HN" posts and which projects/companies are alive and thriving at +1 year, +3 years, and +5 years. It still wouldn't be a complete picture but better than cherry picking winners.
When I submit my project to Show HN it'll be to have the stupid obvious things, or gaping security FU, I've missed ripped to shreds. That's the point to my mind. Same if I were showing it to /r/programming. They're my sanity check before pushing for users.
Hopefully I've already got a good idea of the market before I do that. If I want positive outlook, perhaps equally unrealistic or wrong, I'd submit to whatever non-technical reddit suits the service.
Does anyone submit MVP to HN in the expectation of glowing praise or predictions of success?
e.g. I still don't like Quora, and rarely even visit, especially after the sign-in to view more change.
Just in case you're reading the comments here -
In the infographic section "Comments were generally negative and top comment tended to reflect this sentiment."
the POSITIVE / NEGATIVE categories are swapped.
Who'd have thought a community of grognards seemingly interested only in the latest JavaScript framework and a priori inserting Latin idioms into places they don't belong makes a poor predictor of startup success?
AirBnB poses a safety risk to homeowners. The fact that it's also a billion dollar business now doesn't change that. Additional problems have been discovered with AirBnB's business model, which are arguably even worse.
Sure, if all you care about is money, those criticisms don't matter. But personally, I care about if I might have a negative effect on the world.
Interesting article, but the example at the beginning doesn't strike me as a very good one.
The commenters were correct in being concerned about safety, and the suggestion to link Facebook was implemented.
I remember when I signed up for AirBnB my biggest concern was not being able to trust the person renting the place. Their account being linked to a Facebook profile (mostly) solved this problem for me.
57 comments
[ 3.4 ms ] story [ 61.5 ms ] threadhttps://hn.algolia.com/?sort=byDate&query=comments%20obsessi...
... said me, while sipping expensive wine out of a crystal chalice that came from the 9th century while sitting on a throne made of the finest organic pasture raised bovine hide, while smoking a Stradavarius (the violin, not the cigar).
Not passing judgement, mind you, but at least it isn't a Goddamned echo chamber like most other websites. Well, at least, not nearly as bad.
That's not a bad description, at least of one segment of the spectrum.
Add to that some idea that is not normal and involves actually changing the way society thinks and acts. Or some idea that runs headlong into a highly regulated industry and involves a gamble on winning (zenefits, Uber, lyft). Calling these ideas crazy is basically stating the obvious. The unnormal result is that they actually succeed. And good for them for doing it, and I'm not discounting the skill and ideation or how well these people "earned" the success. I'm just saying let's not discount that they were still lucky and that the ideas were crazy -- this is statistical fact.
My take on it is:
HN/Slashdot commenters can be valuable for pointing out what's technically wrong with the product. They also tend to put too much weight on product's shortcomings. However, they are blind to what's right about the product or discount the product's strengths for the non-geek consumer.
However, that doesn't mean that average consumers are ultimately right either. On the one hand, they don't have the technical understanding which often skews the prediction towards market failure. Their naivete just lets them try the product and vote with their wallets. On the other hand, that lack of knowledge leads to unrealistic hypothetical products... e.g. "if they can put a man on the moon, why can't my smartphone battery last 3 months and project a 4k video with 5000 lumens on my living room wall?" (relevant XKCD: https://xkcd.com/1425/)
Counterexample by way of metaphor: I used to feel the same way about restaurant food, but then my cooking skills increased to the point where I can make most Italian or French food just as well as if not better than most restaurants I've been to.
Now, I rarely if ever go out for French or Italian any more.
Aside from pommes-frites and crepes, finding French food at the bottom of the price scale is pretty difficult.
And I wonder why that's true. I mean, in Los Angeles, I can find food from other great culinary traditions at both the high end and the low end of the price scale: Korean, Mexican, Chinese, Thai, Indian, Peruvian, Vietnamese, Middle Eastern -- even Japanese, which, considering general price levels in Japan, is surprising.
But French food? It's almost always some high priced, sit-down place with a valet out front. There don't seem to be a lotta reasonably/low priced French food joints out there.
(Even stranger is the price of "British" food, usually found in pubs: always seems overpriced. And British is not exactly known as a great food tradition.)
sometimes i'm better off buying takeout food from a restaurant so i can spend that time wrapping up some work task so i can just forget about it. i can bill a little more time that day. and i'll sleep better that night. it's worth it.
Do successful projects receive more positive comments?
One can admire the article as a piece of analysis shaped SEO-bait, even as one despairs over the value it subtracts from the world by existing.
I would imagine that most startup founders probably spend maybe 10-15% of their day worrying about HN/Product Hunt comments. I forget who said it (maybe pg?), but HN isn't meant to be used as support board, especially for YC companies. Of course there are exceptions to the rule, especially if there is malfeasance or fraud involved.
No one can accurately predict success. Not specialized forums, not specialized workers in the field, and not even VCs whose sole incentive is to make money by investing in great ideas.
Even YC says as much that they have no clear idea as to what will be a success, and their strongest signal is the strength of the founders, and that there should be co-founders. I'm not sure how their experiment went with inducting people into YC just based solely on who they were, without an idea, but that's just further evidence that even people obsessed with finding success basically have no clear idea or signal as to what makes a success.
Yes, there are absolutely economic/industry issues that are completely out of our control and hard to predict but I believe a major part is the strength of the team to focus, execute, get along and continue when sh*t happens, and adjust as their Vision hits reality.
I'm not sure what the division is between luck/timing vs team/product vs all the other things but it's hard to accept that it's "largely" out of my hands. Maybe you're right but I hope not.
If it weren't out of your control, then wouldn't someone have already come up with a more surefire way of creating a successful company than what we currently have? The smartest people are here in Silicon Valley, all trying to become billionaires and yet only a very small percentage actually succeed, and there's no way to accurately and consistently predict before it happens.
The launch post regarding AirBNB set a tone with it's title that set people up to be critical..."Sleep under my kitchen table.." Even for those that understood the tongue-in-cheek there, the title is self deprecating.
I suspect the thread would have had a more positive tone with a different title.
[1] https://intelligence.org/rationality-ai-zombies/
I'd love to see an analysis of all the "Show HN" posts and which projects/companies are alive and thriving at +1 year, +3 years, and +5 years. It still wouldn't be a complete picture but better than cherry picking winners.
When I submit my project to Show HN it'll be to have the stupid obvious things, or gaping security FU, I've missed ripped to shreds. That's the point to my mind. Same if I were showing it to /r/programming. They're my sanity check before pushing for users.
Hopefully I've already got a good idea of the market before I do that. If I want positive outlook, perhaps equally unrealistic or wrong, I'd submit to whatever non-technical reddit suits the service.
Does anyone submit MVP to HN in the expectation of glowing praise or predictions of success?
e.g. I still don't like Quora, and rarely even visit, especially after the sign-in to view more change.
And, consequently, why would you then feel smug posting ad-homs about how you aren't surprised that it isn't?
http://comments.network/
Sure, if all you care about is money, those criticisms don't matter. But personally, I care about if I might have a negative effect on the world.
The commenters were correct in being concerned about safety, and the suggestion to link Facebook was implemented.
I remember when I signed up for AirBnB my biggest concern was not being able to trust the person renting the place. Their account being linked to a Facebook profile (mostly) solved this problem for me.