Hacker News Reader

Comment by pupppet | Hacker News Reader

What really ticks me off are all these sites clearly scraping content from Stack Overflow that somehow end up near the top of the search results. How is there not an automated fix for this. Lemme help you out with this Googlers: If site content A == site content B, pagerank = -1.

View on HN

They have a fix already and applied it to the Wikipedia clones. They intentionally ignore other content scrapers because they drive ad impressions.

> because they drive ad impressions.

Could you explain why?

If Google replaces SEO trash with decent results, the ads they will show on the Search page will remain unchanged. How would the impressions suffer? Or rather, since Google charges per click not per impression for search ads, how would ad clicks suffer?

In fact, I thought low-quality organic search results are purely a negative for Google profits, since users (over a very long run) move their search elsewhere (Tiktok, Reddit, etc.).

It's about the ads on the SEO trash pages. Those pages are typically loaded with ads which increases the likelihood you click on them.

I wish google had a feature to permanently ban domains from search results.

They had a personal block list at one point... but removed it.

There are browser add-ons that can do this (for example, uBlacklist).

Kagi has that feature.

Why not just create your own search box and put -site:domain.xyz in the query? Maybe there is a length limit but at least you can ban some

There is a chrome addon called uBlacklist that auto-hides domains (configurable) in the google search results.

this list is really helpful

https://github.com/quenhus/uBlock-Origin-dev-filter

Cue a HN post “wtf google, my archive website is getting marked as spam.”

I don't want your archive website near the top of my search results when the original is readily accessible.

Won't stop the outrage post.

The actual outrage should be when a small site posts something and then that content gets stolen and posted verbatim as an answer on stackoverflow.

There's even the possibility that the Google crawler will have seen the SO page before discovering the original small site.

The suggestion is not to penalize all websites that share some minor amount of content with SO but to remove websites that copy tens of thousands of comments from SO and do not provide any other value. It’s a pretty easy to identify and implement this distinction.

You have any examples of that? Sounds far-fetched to me, since SO posts:

(a) are in response to very specific questions

(b) can be easily reported to SO for plagiarism

You misunderstand the parent comment.

They’re saying there are full sites which just scrape SO content. These are not hosted in StackExchange and therefore are not subject to their reporting system, and generally are filled with ads which provide the money and therefore incentive.

> You misunderstand the parent comment.

I don’t think so. The parent seems to talk about StackOverflow answers plagiarizing some small blogs, not StackOverflow itself being scraped.

Not an issue I’ve witnessed though.

I see the parent comment clearly stating small blogs lifting content from SO and SEO-optimising it.

This happens a lot. For instance, a Google search for “maps vs list comprehensions” shows an ad-based website result featured at the top, despite having re-hashed content from the SO post on the same topic.

Clearly, Google’s SEO favors recently-published articles over quality.

I ran a search “react node.js”, and a digital marketing blog post is highlighted in the “Featured by Google” despite being low-quality. So, it’s just spam at the moment to milk views by content-marketers.

Also, it’s hard to plagarise from small blogs when SO is the more active forum for day-to-day solutions. I don’t see that happening a lot.

> I see the parent comment clearly stating small blogs lifting content from SO and SEO-optimising it.

Are we talking about the same comment?

> when a small site posts something and then that content gets stolen and posted verbatim as an answer on stackoverflow.

This is about SO answers stolen from small blogs, not SO answers stolen and copied to SEO optimized pages. Unless we’re not reading the sentence the same way somehow?

Sorry. I took the top-most comment as parent.

Having said that, SO answers getting stolen from small blogs is a rarity as I wrote in my previous reply.

It wouldn't have to be malicious or uncredited, SO answers often quote manuals, books, etc.

That's a good one too.

These can be hidden using uBlock origin. I’ve got a list you can import (just use the raw link). The selector isn’t perfect in Google (sometimes leaves some unattached content) but the link is always removed. Feel free to submit a PR if you find any other SO clones - there seems to be no end to them.

https://github.com/levymetal/filter-lists/blob/main/filters/...

Thanks! I didn’t realize uBlock can remove specific domains from Google search, but looking at filter definitions it makes sense.

buuh bye Investopedia

Is Investopedia a clone? What is the original?

So both sites lose? That'll make manipulation easier than ever before.

Or does the older site win? Frequently scraped / already popular sites will love that.

It kinda sounds like you think that Google doesn't have the resources to have a human decide that stackoverflow isn't itself a spam clone.

This is a huge problem for a very small number of sites and a non-existent problem for everything else. The list of authoritative, egregiously-copied sources could easily be maintained by one person. The only difficulty comes from determining when to trigger the system.

They have the resources, they don’t have the willingness. Google’s MO is never hiring a human to do a job well, when a computer can do the same job badly.

Derank the newer page, you don't need to derank the entire site.

This doesn’t work in general because people can copy content faster than google can check who was first. Also sometimes you actually want the derivative.

Let’s start with big offenders. If a website copied tens of thousands of comments from SO it gets penalized. We don’t always need to find a perfect system to improve our experience.

Using that rule it would be very easy to derank any website.

Not if the list of A sites was manually curated. Like stackoverflow and wikipedia. This isn't hard for shooting down a ton of sites.

Then people could derank sites by copying them to stack overflow.

And quickly they'll get deleted again: spam postings on SO are generally very short-lived.

Reply on news.ycombinator.com