Hacker News Reader

Comment by firasd | Hacker News Reader

I'm not sure this whole drive towards 'original content' makes sense. It encourages people to make long rambling pages that don't get to the point

In fact it seems like Google worked better when wikipedia was almost always the first result on various topics

What I mean by rambling is if you type something like "types of oranges" and suddenly land on a page where some dude is clearly just filling up paragraphs for google like "so you want to learn about oranges? an orange is a great type of fruit. here's a bunch of text about oranges being described in historic literature"

View on HN

The dissonance is between Google's native perception and external reality:

reality 1: The www exists. Google indexes it, analyzes it and delivers it to users. Users like certain things, like original content.

reality 2: Google's ranking policies/algorithms influence the web. The "original content" that exists in a world without Google is different to the content that exists in a world where Google ranks such web pages more highly.

Google refuse to see or present themselves in the role that they actually occupy. They avoid thinking of the search algorithm as encouraging or discouraging anything. It's just analyzing.

On youtube, This mindset is even more loopy, because on youtube they actually own the platform. The recommendation engine or whatnot implements what is clearly a new policy, Youtube pretends that there was no policy to change in the first place.

>Google refuse to see or present themselves in the role that they actually occupy. They avoid thinking of the search algorithm as encouraging or discouraging anything. It's just analyzing.

Can you clarify what you mean by this?

As an outside observer, it seems that Google recognizes that their search engine does have a "Heisenberg" effect akin to "measurements of certain systems cannot be made without affecting the system"

E.g. the Google search ranking causes the rise of content farms. Google then fights back with a new revision to the algorithm to downrank them. That constant arms race between various blackhat SEO and Google's new algorithms was commented on many times by Matt Cutts (Google former head of search quality): https://hn.algolia.com/?q=matt+cutts

Another example is the RapGenius punishment by Google to discourage content that tries to game the algorithm: https://www.google.com/search?q=rapgenius+penalty+google+ran...

Google often makes manual human intervention e.g. M Cutts team telling HN they're looking into RapGenius SEO hack: https://news.ycombinator.com/item?id=6956658

Are those are not examples of Google understanding its effect on internet content that tries to game their algorithm?

I think the issue is that Google's algorithm isn't perfect -- and therefore it appears like they don't discourage bad content.

OK..

In a sense, I am overstating. Having an anti-spam team is obviously a recognition that SEO/Spam exists and that Google is trying to discourage or mitigate.

But Google is well beyond just attracting spam. Success in Google search rankings is, for many sites, the better part of online success. If Google ranks needlessly wordy recipes more highly than concise & useful recipes... then wordy recipes get written. It's not about ranking recipes anymore... it's about the effects google has on recipes. The recipes get written in a wordy fashion, because of the ranking system... not just ranked because of the writing style. Google is dictating the nature of online recipes with their rankings.

A few years back, Google made changes to youtube's recommendation engine in a way that really "encourages" frequent, regular videos. Viral hits became much less common. The result has been to send many professional youtubers into a frantic grind. There's no point in taking time with a video, or trying weird ideas. It won't go viral anyway. A week off or a couple of flops is harshly punished. There are tens of thousands of these youtubers, basically small businesses.

There's never any reckoning with 2nd order effects. It's always communicated as it is here. I believe this is how Google execs actually think about the issues. As a quasi-spam problem to be mitigated with the next “helpful content update.”

>There's never any reckoning with 2nd order effects.

Can you articulate what concrete actions Google could take that would address these 2nd order effects?

>, Google made changes to youtube's recommendation engine in a way that really "encourages" frequent, regular videos. [...] A week off or a couple of flops is harshly punished.

I've seen this repeated many times (especially from Youtubers making videos about "burnout") but the analysis about cause & effect seems incomplete. As one counterpoint, Ben Krasnow "Applied Science" channel slowed down from weekly uploads to a video every few months and yet his views and subscribers went up not down: https://www.youtube.com/c/AppliedScience/videos

Another yt channel that had a year between uploads and the views went up : https://www.youtube.com/channel/UCX7katl3DVmch4D7LSvqbVQ/vid...

My pet theory on the contradictory anecdotes: the Youtubers making videos on a topic with lots of competition from other Youtubers uploading every week -- such as fast fashion clothes shopping -- are the ones that seem to suffer if they slow down. E.g.: https://www.youtube.com/results?search_query=zara+haul

However, if you're making videos in niche topics with originality (maybe "weird ideas" as you put it), you won't be penalized by infrequent uploads. Many examples including Technology Connections, Applied Science, etc

So a different conclusion can be reached... if one makes a high quality videos, one can even upload just once a year and the Youtube algorithm won't penalize you.

I suspect some of the contradiction comes from whether you're measuring views per video or ad revenue per month.

If one youtuber produced a 32-minute video per month and got 500,000 views while another produced eight 4-minute videos per month and averaged 200,000 views per video, who do you suppose gets the most ad money? I'd wager the latter.

Goodharts law

I think it’s more that nobody knows the policy. It’s all ML changes few people understand or can even communicate. I don’t expect google would explain if they knew how it worked, but they also cannot explain how their technology works.

IDK. I suspect this is less true than most acknowledge.

Black box or not, google are creating these systems, analyzing, optimising and implementing them. It's indeed hard or impossible to trace back individual examples and their whys. But the macro effects are visible and choices are made.

It's just easier to "blame the computer," internally or externally. It's just like bank employees merge always blame "regulations." In fact, whatever they are blaming is a policy created to implement that bank's compliance framework which exists to satisfy the regulator's goals and the bank's goals." Most of the time, the underlying regulation is distantly removed from whatever piece of bureaucracy is annoying the customer or employee. But, if it can be blamed on the regulator, it will be.

It's just easier, if at all possible, to blame an immovable and mysterious force rather than the more likely culprit: humans doing a bad job.

This resonates with me so freaking much. I am disgusted by how every bit of information on webpages seem to be buried between paragraphs of empty EMPTY talk and even when they seem to get to the point, most of the time there is no real information there. Sorry, this had to come out. We really need better search engines.

I think it's going to require a social rating system. Which is why people append "reddit" to searches these days.

It’s safe to assume that a social ranking system covering the whole web will be gamed.

Yes, but if it's truly social it won't be easy. There need to be ways to see who did what and filter them out and various groups / spheres of influence. Just like in real life.

If it is a hidden algorithmically social function then of course it will be gamed.

But IRL there are certain people whose advice and recommendations you value and those whose you ignore. And in other cases you can easily ask the source of other information to find out if it is high value or not. MLM is the gamification of the IRL social structure and it's fairly easy to opt-out.

That's what search needs, a way to see the path information took to be presented to you and a way to filter it.

Unfortunately right now, so much of the best information is in Facebook groups, post and comments. The interface there is absolutely horrible though and not designed to provide you information, but to maximize the amount of ads that come across your screen.

The same is true for video information. It's not easily searchable or digestible. The web peaked when information was predominantly text form and not fragmented into walled gardens.

> Yes, but if it's truly social it won't be easy. There need to be ways to see who did what and filter them out and various groups / spheres of influence. Just like in real life.

Appending site:news.ycombinator.com instead of Reddit?

There is an extent they can go to with account verification where gaming is not very feasible.

Can you create a fake account with a verified credit card number, verified phone number, passport, drivers license, account history consistent with human usage, Google One subscription, etc.? You probably can, but doing it at-scale is going to be quite costly.

I don't think we need this that bad so as to jeopardize our privacy and freedom over it. So I'll pass on this idea.

Can you get a lot of people to install a piece of software and then use their account instead? Could you even pay those people per hour of usage of their account?

Quite cheaply

My understanding from reading a number of articles about this "helpful content" update from Google is that they want to get rid of pages like what you describe and instead prioritize pages that get right to the point. So perhaps you'll get what you are asking for.

This is true of most non-fiction literature as well (self-help and business books especially), so I'm not sure if Google can necessarily discourage people from adding unnecessary padding.

Google is already scraping bits of website content and showing it to the user as a "featured snippet". Nobody is going to write short pages if Google can already rob you of a click so easily.

Encourages people to make long rambling pages which don’t get to the point? Ever seen YouTube, it’s exactly the same. 10-11+ min videos saying almost nothing because it’s the most lucrative length. The internet of today is just everyone gaming everything for a dollar.

At least Youtube has a timeline (with sometimes even chapters) that let you jump ahead. The thumbnails are also a good indication on what type of content you can expect from a video.

Web pages have a timeline in the form of a scroll bar on the right

Actually that's a cool idea, put a graph on the scroll bar which shows where most people spent their time, same as the Youtube widget.

Medium kinda does something like this by highlighting the most shared part. Alternatively, you can also use AI to summarize the text on web pages.

Most YouTube junk does not even have a proper meaningful description, just some buzzwords for SEO.

In the case of Youtube, the video thumbnail itself is usually a strong enough signal as to whether it's worth watching.

Enormous swathes of the platform are filled with gossip channels (politics, crypto, stonks, celebrities, music) hosted by clout chasers desperate to be seen as authorities on that topic.

Of course there can be good content inside of these categories, but you can generally tell which channels are "optimized for engagement", i.e. run like a business, and those are more amateurish and perhaps more authentic.

I sometimes search for specific keywords in the transcript of the video to directly jump to the point in very long videos. Otherwise i just click at the middle & judge.

Those types of rambling posts are clearly one of the targets of this update. From the Search Central guidance, here’s a question for site operators to ask themselves to “Avoid creating content for search engines first”:

> Are you writing to a particular word count because you've heard or read that Google has a preferred word count? (No, we don't).

https://developers.google.com/search/blog/2022/08/helpful-co...

It's funny because one of the points is

a) dont summarize other sites

b) dont write overly long articles just to make google happy

It's a shame a) is never helping with b)

I think there's a different driving factor that. I've heard of a metric that is how long you spend on a given URL. Not sure how they gather that data, probably only gathered from Chrome users.

Anyway, from what I've heard, that's the reason recipe sites have started posting 4 pages of drivel about the dish before getting to the actual recipe.

Or so I've heard, I don't really have any good sources for this, could just be hearsay

> probably only gathered from Chrome users.

Or, Google Analytics. Or Google ads. Or, if you return to Google and try other links for the same query. Or if you return to Google and refine the query (e.g., a new search where the term is within a predefined threshold of word vector similarity)

There are plenty of ways to approximate time spend on link.

A couple of months ago I read on various SEO blogs (they all seem to mostly copy and paste each other with little original research) that these days it is less about keywords and technical things like semantic html, but if and how long people engage with the search result.

E.g. if a link to your page is shown to the user in search and they don't click or they return to search page too quickly, Google sees this as a signal that the result was not helpful.

If it's based on length of engagement this does explain the trend towards verbosity I've noticed. I assumed the long leading paragraphs and introduction and background sections were just excuses to keep dropping the keywords in, but I guess it's to make me have to spend a lot of time reading and scrolling to get to the meat of the answer.

I guess that means we can use adds and tracking against itself. Any time you go to a webpage that is ad heavy quickly leave. Over time they get ranked down until they shape up or die.

You have just described the typical content-farm page created to game Google. I hope this update is going to lower the ranking of those pages (pages that usually have an ad every paragraph or more).

Every site needs a commercialization score and a way to filter them down. Number of ads should be a dead give away, but that's antithetical to Google's business.

Yeah, I was going to suggest the same: farm-factory sites are easily identifiable by the number of Ads, especially on the doubleclick.net domain. But maybe for Google in the long-run would be a net positive to ban those sites anyway, due to better "organic" results and how people will see Google results.

The first headline is "Better ranking of original, quality content"

The update literally says they are working to get rid of the type of useless content you are talking about.

Feels like the early days of the internet when the computer savvy knew how to find what they were looking for (possibly even using early google) and everyone else was condemned to sifting through trash to get something useful. These days I find myself reaching more often for specialised search like going straight to wikipedia or amazon if I want to get a direct answer.

I almost spit coffee all over my keyboard. "straight to ... amazon if I want to get a direct answer.". Hilarious.

This is especially annoying with recipes. There's usually a massive backstory that takes up 90% of the page. Some recipe websites even post a button to jump directly to the recipe, at least they're self-aware.

I used to search for recipes on the Internet. The spam became unbearable. Then I spent $30 on a cookbook, one recipe per page. Very happy with the outcome. Simple, fast, effective: it just works.

Happiness meta-rule: Throw electronics away, use analog equivalents instead.

My impression is that this update is specifically about deranking those types of pages

I agree. That's why I habitually add "wiki" to most of my search phrases in the first place, in order to circumvent the whole SEO trash and get to the point.

Google prolly cannot earn money simply by featuring high quality stuff like Wikipedia or nerd websites. Usually SEO farms spend money on Google to earn money. So that's why search really cannot improve in my opinion due to this conflict of interest.

I hadn't thought that the reason they do this is to appear to have original content. It seems like distinguishing between the original source of information and a site that has a bunch of filler "original content" but just exists to republish/plagiarize some other site's content must be a difficult problem to solve.

An AI trained on the body of verified human knowledge that understands natural language could replace a good deal of the googling done by people.

In an ideal world, such an endeavor would be supported by governments all around the world (or a single "Earth government"), with yearly updates to keep up with the state of the art.

Clearly the end game for Google is to, also, become TikTok. 100% algorithmically sorted original content.

I have no problem with that if it can also read my mind and serve up what I need. Until then, I want the option to correct their generous recommendations.

Wikipedia doesn't bring in adsense dollars, I'm surprised it is usually still on the first page (frequently it isn't for disease searches, which have some of the highest ad spends).

But it's very easy to exclude most popular words and use TF-IDF and Google probably uses that to devalue rambling.

Yeah. The more words, the more space to insert ads. Then people tap on those ads by mistake when scrolling.

yes, yes, and yes!

I've started just closing the tab. I don't give a shit that a certain country planted oranges and it gave them naval superiority, I want to know what types of oranges there are!

literally every dang recipe website in the universe... Ugh.

Reply on news.ycombinator.com