Google's PageRank patent has expired (2019)

The type of people get into locksmithing are like those who do computer security for fun - they like to figure out how things work and how they can be broken. Which means that every locksmith wants to figure out how to game the system. And the first thing that they all figured out is that people tend to select whatever locksmith is closest. So they all went and pretended to be in a million places around the neighborhood in hopes that THEY would get selected.

That was the case a decade ago. And it was a nightmare. Glancing at a locksmith search now, Google somehow cleaned that up a lot. But it doesn't surprise me that whoever is on top now is someone who figured out how to game the current system.

On Google, I treat my search terms like a Venn diagram if I want good results.

Eg: want an article about the texas blackouts, but not the ones a decade ago?

Type "texas blackout npr 2022" minus quotes.

But if you do that on other engines, it may be MUCH more literal, the literal intersection of those terms, and I need to do the opposite: use as few terms as possible, possibly paired with using the site: operator, intitle operator, or other things.

>It's made it harder to search for bits of poetry, quotations, or song lyrics, especially.

Yeah to be completely clear, my default is DuckDuckGo, then very rarely I fall back to Google, but often if I'm doing that it's because I didn't want to trouble a librarian -- they talk about privacy, but I had a series of unfortunate events when I told one I want to use books as much as possible because I absolutely don't want some of these tech bros to know what I'm looking up.

(That dichotomy of folks who know information science and those who have critical thinking or coding skills needs to end, now. I'm an alumni of one of the highest ranked schools of information science in the world, and I will not be figuratively or literally extorted into a PhD to get roles others get with a bachelors.)

The signals that matter the most: 1. Anchor text (and all variants of smearing and distinguishing between high quality and low quality anchors) 2. In aggregate, which pages got clicked on on any given query (and all smearing variants -- using ngrams, embeddings, ...). At Neeva, we use it for retrieval and scoring. 3. Query understanding signals mined from the query-click bi-partite graph and the query-query session refinement graph. 4. Page summarization signals built on top of 1 and 2 and body text. 5. To a lesser extent, query-independent page quality signals

Whether you use term-based retrieval or nearest neighbor (embedding) retrieval, a heuristic combination of signals or LambdaMart, whether you calibrate to human eval or clicks, whether your topical relevance function is hand crafted or uses a combination of deep learning and IR signals are all details past that.

tldr; there's a lot of craft in a search ranker, and no one silver bullet. Definitely not just PageRank.

There was loads of SEO back in 2000 even! It brought alta vista to its knees, the number one search engine of the day.

Google got started, grew, because it filtered all that SEO spammy junk.

Do you have specific stats to back this up? Number of SEO pages vs good ones?

Or are you just presuming?