The search help page has been heavily updated since the implementation of the new elasticsearch search engine. Most of the Ninja search techniques that were considered "advanced" have been included as reference.
For me, effective searching spawns from using a combination of search criteria. For example, mixing search terms with tags and user specification has been very effective for me. The difficult part is remembering some of the key components your attempting to filter on in your search.
I recall egreg writing up what some commonly refer to as a short 'course' on the inner workings of
\ooalign with symbol construction.
For this I use the search:
user:4427 course code:"\ooalign" which yields
The only 3 relevant results include the post I'm after, as well as two others referencing the "course on
\ooalign by egreg". Adding
is:answer wouldn't help here, but knowing that egreg doesn't have the student badge might help elsewhere - knowing your crowd and your content helps.
High-voted questions gather the most attention, are the ones you most likely will remember and what you're probably after, so you can even sort by votes (or add a lower bound on the score of, say, 20).
You're working on a document written using
memoir, but would like to use the power of
titlesec to format the sectional units. However, you're experiencing problems.
[titlesec] [memoir] [incompatibility]
This yields a single result:
Compare this to not using tag searching and merely searching for the terms
memoir titlesec incompatibility
A result of 9 posts may not be too bad to wade through, but still is more than the targeted result when using tags.
As a side note, the mixing of search tactics heavily depend on appropriate housekeeping of posts. Our tag editors do a tremendous job of keeping things organized, which can help in making search an effective tool.