Hacker News new | comments | ask | show | jobs | submit login

I think the story of this game exemplifies the fact that content needs a minimum amount of attention to reveal it's true value.

The journey of Flappy Bird appears to me like this:

Stage 1: Almost no attention, almost no growth

Stage 2: Some kind of grey market paid downloads / ratings service

Stage 3: Attention of minimum critical mass of early consumers (increase in audience of 1 or perhaps 2 orders of magnitude)

Stage 4: Attention of influencers on other platforms, especially pewdiepie on youtube (increase in audience by a couple of orders of magnitude)

Stage 5: Attention by large community of enthusiasts (increase in audience by a couple of orders of magnitude)

Stage 6: Mass market attention

where each stage is dependent on the one before it, with the exception of stage 2, which is performed externally.

What's interesting to me is that the qualities the game possesses, which were sufficient to carry it from stage 3 to stage 6 without promotion on the part of the dev, were not sufficient to propel it from stage 1 to stage 3. We could say that the true quality of the game was unknown in stage 1.

My personal opinion is that the vast majority of content created languishes in the equivalent of stage 1 for its particular ecosystem. Of this content, the vast majority will be garbage, a small minority will be reasonable and a tiny minority will have the potential to be a widespread hit. But of this last group, most or all of it will never emerge from stage 1 because stage 1 does not provide it with enough attention to separate it from the rest of the (bad) stage 1 content.

Increasingly I think that the journey to stage 1 to stage 3 is the most important, most difficult and most overlooked part of the progression.

For example, when something is submitted to HN, from my anecdotal observation it will typically get something like 5-10 simultaneous visitors from the new page. If it gets a minimum critical mass of votes to hit the front page, this will increase immediately to something like 50-100 simultaneous viewers and increase from there. But often, it only takes 3-5 upvotes for it to hit the front page. So the relatively trivial actions of the small critical mass of early consumers has an extraordinarily large effect on the dissemination of the content. Indeed, as a content creator, it's often struck me that the actions of those first 3-5 people have an equivalent significance, in terms of the world's experience of my content, to me as the creator.

And if you create good content, then just getting to the front page is by far the most difficult part of the process, because once there you will naturally attract upvotes from the vastly increased exposure. But with only a few random bits of cosmic entropy set differently, the creator could create exactly the same content, fail to get those first 3-5 votes, and the number of views on the article/app/etc. could be 10 rather than 10,000 or 100,000.

In my experience, lots of platforms follow the same model. Reddit is a very obvious one, but the same holds to for trying to get press interest: so much depends upon the decisions of a few key journalists, and that decision may depend on how many other emails hit their inbox that hour, or whether or not they've had their coffee yet, or some other particle of background entropy. This isn't a criticism of the press, it's just a consequence of the current system. The same is true for people who run influential blogs or social media accounts with large followings, or newsletters.

The problem is that there is less attention available per item of content than the minimum quantity of attention needed to rate the quality of that content well. When I create something new, my worry is never "Boy, I hope that people don't simply dislike this," it's always "Boy, I hope that enough people see this to give it a shot of achieving its potential, whatever that turns out to be." If it turns out people don't like it, which is of course always a possibility, that's fairly easy to handle. If it disappears into obscurity, that's much more difficult to accept.

When I create content now, my promotion strategy is 99% focused on that first stage. If it reaches that critical mass of attention among a small number of early, influential people, I feel like my job is done and the chips should fall as they may. If it reaches that initial critical mass, people will want to write press stories about it and post it on social media and ask for interviews and tell their friends about it, assuming its any good. This may be an obvious conclusion, but the interesting part, for me, is that the initial critical mass is much smaller, much more random and much more difficult to achieve than most people realize.

You should add the simpler, deeper issue of upvoting mechanically filtering out good long form content.

Say I found a link here and it's a really good long form story, I really got into it, even had a coffee break while reading it. Then I want to vote for it, but the link has fallen down and is hard to find back.

It is even worse on Google, Facebook or Twitter, because of the infinite scrolling. I always open seemingly interesting links in the background and continue scrolling, but then hardly ever even try to like, plus one or favorite good content.

It's like a perverse filter, selecting cat pics and crash gifs over really nice content that require more than a few seconds to digest.

The one who will fix this will actually revolutionize content selection and probably kill curation.

One simple change would be to replace time-based "gravity" dragging items down as they age. Instead, gravity could be based on the number of users who have "finished" consuming the content. Perhaps by just clicking "Back" to the aggregator, or by explicitly acknowledging the completion. There could be three buttons: Like, Dislike, and Saw But Don't Care. In the algorithm to drive items downward, time would be replaced by the count of all three actions. Long-form content would then not be penalized so much (but would still be disadvantaged by users who dislike or don't care about long-form content, of course).

A New page might also need to be sorted differently, e.g. by putting items with zero count at the top, and ignoring time as with the front page.

This is an extremely well articulated point. Nice job.

The music market has worked similarly for a long time.

Almost all markets work like this it's basically the innovation diffusion model from Crossing the Chasm.

I think this is the first and last time that i will read the name Pewdiepie as HN comment :9.

Applications are open for YC Summer 2019

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