Google Search Really Has Gotten Worse, Researchers Find

[rcjordan]

https://www.404media.co/google-search-really-has-gotten-worse-researchers-find/

Researchers: from Leipzig University, Bauhaus-University Weimar, and the Center for Scalable Data Analytics and Artificial Intelligence

[BoL]

Seen that one on HackerNews, I bet they're missing a bunch of aff sites as they're not rendering the pages.

I'd suggested on the HN thread a solution is a bunch more search engines, at least there's then some variation in results potentially. Looks like edge are somewhat into the idea https://www.techradar.com/computing/microsoft-edges-new-search-feature-has-me-genuinely-considering-moving-over-from-chrome

I've been liking the idea of 'pressure groups' forming their own algos. Something like Google's index since every other is lesser, but the algo is defined by cohorts you vote for. So if I like Th3Core's ethos, it positively influences the algo of stuff Th3Core likes, for me. Add in more groups etc, take your pick. Add a Pagerank type thing to the groups and their followers.

It does seem like there's more complaints about G quality and their index is the best that's going. Maybe the pendulum could swing back to human curated content? Problem then is how those places get found.

Alt search engines invariably have a hard time given bot whitelisting, MITMs like Cloudflare defining acceptable activity and large sites only allowing the likes of G to crawl them at a speed to keep up to date with their content.

[Brad]

More from The Reg:  https://www.theregister.com/2024/01/17/google_search_results_spam/

>solution is a bunch more search engines

Of course I agree with this.  Infoseek, AltaVista, Excite, Lycos, Inktomi, Webcrawler, plus directories Yahoo, ODP and Looksmart made it hard to optimize one page that would rank across them all.  And I think we need about six major crawling search engines in addition to Google and that are not trying to clone Google's serps.

But over the last few years I think we also need a second class of search engines that lean toward less commercial pages, favoring pages that don't have excessive advertising (or network advertising) and pages that don't have affiliate links.  I'm not sure if these non-commercial engines should be part of the six major engines I proposed above or in addition to them.

>pressure groups forming their own algos

This sounds interesting BoL.  It would be neat to see at least one built that way as an experiment.

It does seem like a lot of new crawling/indexing search engines are starting up.  It's a mixed bag.  Some have already failed, others don't seem to be bringing anything new to the table but some appear to be trying to be different.  It's a bit like watching a horse race just to see who can make it to the finish line.

[BoL]

>It does seem like a lot of new crawling/indexing search engines are starting up.  It's a mixed bag.  Some have already failed, others don't seem to be bringing anything new to the table but some appear to be trying to be different.  It's a bit like watching a horse race just to see who can make it to the finish line.

And a lot of bending of rules, like Brave (what's their UA? Gbot rules etc). and DDG just an inflection of marketing.

I could see an advantage towards DMOZ style human curated lists but it requires following beyond the usual info discovery of default search and social media.  Alt search engines do seem like an answer as an alternative but there's always the issue how how visible they are.

Personally it seems like the younger generations (at least the objective thinking ones) will wonder what is the point of looking on the web at all because there's so much crap, which would undo all our hard work the past 30 years :-) I guess for me being a teen in the late 90s and involved witnessing SEO over time it's just a necessary cat and mouse game. But it seems far too gone towards Google's version of events.

[Brad]

>Brave

I'll say it because I said it before, IMHO there is something that strikes me as dodgy about Brave.  Nobody has detected a crawler yet they claim to have over 10 billion pages indexed by monitoring volunteer humans clicking on links in Google and Bing.  That doesn't add up, humans don't work that fast.  They have no real algo other than to clone the search order displayed by Google, if Google goes down who they going to copy?  I read their AMA threads on HN and Reddit and their replies were all copy/paste identical.  Worse while they said something press release-ish they never truly answered the question.  They always dodged, without seeming to dodge.  None of it built any trust with me which left me asking how private is Brave's privacy?