Comment by kilburn | Hacker News Reader

Comment by kilburn | Hacker News Reader

Parent Full thread

I don't know about others, but when I think of the "good old google days" I'm _not_ expecting the results for your example queries to be any good.

In those days querying took some effort but the effort paid off. The results for "history" just couldn't matter less in this mindset. You search for "USA history" or "house commons history" or "lake whatever history" instead. If the results come up with unexpected things mixed in, you refine the query.

It was almost like a dialog. As a user, you brought in some context. The engine showed you its results, with a healthy mix of examples of everything it thought was in scope. Then you narrowed the scope by adding keywords (or forcing keywords out). Rinse and repeat. As a user, you were in command and the results reflected that.

The idea that the engine should "understand what you mean" is what took us to the current state. Now it feels like queries don't matter anymore. Google thinks it knows the semantics better than you, and steering it off its chosen path is sometimes obnoxiously hard.

> The idea that the engine should "understand what you mean" is what took us to the current state. Now it feels like queries don't matter anymore. Google thinks it knows the semantics better than you, and steering it off its chosen path is sometimes obnoxiously hard.

Bingo! If you cede control to Google, it _will_ do what it's optimized to do, and not what _you_ are looking for.

What is optimized to do says nothing.

Optimizing for open text queries means dealing with a massive search space, the thing is choosing a subspace where to search and that is the part engines have to refine, how that is done is a different story. Some people may agree to use their location, search history and visits to online stores to do so but some may not.

This is why, in the good old days, my favourite search engine was Alta Vista. In its left margin it had arranged key words like a directory tree that could be used to further refine the search. So my ideal search should do something like this if I type in a generic term: provide me with relevant information about the general topic and than help me to refine my search. The way of Wikipedia to provide a principal article and a structured disambiguation page is the way I would prefer.

I admit, my evaluation of the search engine was just a simple test how much I can get out of the results for some generic key words in the first place. A more detailed evaluation should, of course, look deeper. It was more of a test balloon to see if this search engine raises any hope that it could be better than Google with regard to my own (subjective) expectations of a decent result set.

I get what you mean, but part of the whole initial appeal of Google was that it gave much more relevant results initially than Altavista or the other options. That was why Google put in the audacious "I'm feeling lucky" button.

Yeah but it's from that same philosophy that Google Search is useless as it optimises for the first result.

There is no search engine that searches literally for what you asked and nothing else. Search is shit in 2021 because it tries to be too clever. I'm more clever than it, let me do the refining.

This was the result of two things. mapreduce. using links rank the pages.

Using links to rank the pages is not really possible any longer because of seo spam links.

>"I'm feeling lucky" button

My brain got so used to ignoring it I completely forgot it's a thing. I'm also unclear what it does? On an empty request, it gets me to their doodles page and with text in the box, gets me to my account history landing page.

It automatically redirects to the first search result.

Right, not sure why it wasn't working yesterday as opposed to now, I swear I wasn't doing it wrong (or how I could've).

Reply on news.ycombinator.com