A couple of weeks ago someone asked the same in chat, and I gave some tips (see transcript). I'll try to whip that up into a more or less clear guide below. Still in writing.
Disclaimer: all examples will be answers of mine, as obviously they're the ones I know best. Sorry if it looks like I'm showing off - this is not the intention.
- Learn the way of Google operators
- Get to know the big sites (and the recurring themes)
- Don't limit yourself to English
- SFF is a good resource too
- Lead by example!
Learn the way of Google operators
Feeding the search bar a ton of keywords seems like a reasonable idea, but "just" words don't limit the number of results - far from it, generally. Fortunately Google has a couple of very useful tools to refine the searches. In case of story-identification, my favorites are the following.
" " exact quotes
Will scour the entire indexed Internet looking for that very set of words. Probably one of the most useful ones.
Great addition to the " ". It replaces a missing word (or a couple of missing words) so that you don't have to go through all possible wordings of
the back of people's necks / the back of their necks / the back of the infected's necks etc. Wildcard will search for sites containing a phrasing starting with
back and ending with
necks, with something like 1-10 words between them.
Case in point is this answer to 50s/60s movie with round disks on the back of people's necks, for which
back * necks did the trick.
- the minus
Tricky little fella. It will exclude the word/set of words you give it from the search... But as some words/names are very common, or very linked to a particular work, you might miss out on a useful result.
I'd suggest using - to refine a bit, have a clearer idea, then removing it once you've thought of better keywords.
For instance when answering 2010s novel series about a supernatural prison inside a tiny cube I first excluded Dashner (author of The Maze Runner) because keywords like
maze brought up too much of it, but had to remove it afterwards. See the full explanation in chat, starting here for the complete breakdown of how I found that one.
site: the site limiter
A very useful one too!
It will search for whatever keywords you typed, in one site and this site only.
site:goodreads.com is quite useful to search for books, same with IMDb etc.
It's useful to know how sites are structured, as
site: can dig down into the site's hierarchy as well,
site:imdb.com/title will not look at user lists, as their URL begin with
imdb.com/lists/. I find user lists to overlap a lot, which make them a noise in the results, hence the need to exclude them. For fanfiction for instance,
site:fanfiction.net/s/ will look into the works themselves, and exclude user profiles/lists etc.
It's also especially useful to search for extracts in way too long fanfictions rather than loading each chapter and manually CTRL+Fing each page. Once you think you have a hit, you can go for
Jean-Jacques house elf tiger site:fanfiction.net/s/123456 and see if there's a chapter of Fic 123456 on fanfiction.net which contains those words.
Google Images helps too
For covers/stills mainly, if OP remembers it being a certain color, you can filter images by color and try to spot potential matches. Including
cover in the search may help for that.
Also helps when OP is looking for a black and white comic/movie.
A recent winning case for me was Middle grade novel about two kids in an underwater city, one coming from the surface; they get involved with a gang where OP drew a Paint sketch of the cover; filtering by "blue" color returned the book's cover.
Best thing about them Google operators? You can combine them to always refine your search.
For instance, you can exclude sites with
-site:www.pinterest.com, because Pinterest is often noise in the results. Take it further with our beloved wildcard!
-site:pinterest.* which will kick off ALL extensions of Pinterest, may they be .fr, .com, .co.uk etc.
A rather short query, if it combines them well, can yield interesting results, for instance the
movie "back * necks" site:imdb.com from earlier.
Some further info on Google operators here.
Get to know the big sites (and the recurring themes)
Each media has its "unique" characteristics: movies will have taglines, transcripts, books may have more quotes available online...
You might also want to try and spot which works are commonly IDed, and which tropes they cover. (and you can discover great reads/watches while you're at it)
See F1Krazy's answer for TVTropes.
The Internet Speculative Fiction Database offers a very useful advanced search tool.
IMDb is really well done:
- There's a tagline section for each movie for instance. In case the OP only heard the tagline (and has a good memory, granted), Googling
taglines "from the woman" site:imdb.com can do the trick.
- Or if the OP thinks one movie was SO close to theirs, but not exactly, it might be that the story is a remake/retelling etc. Then, the "Connections" section can be of assistance. Note that the "Connections" section has several subsections: Remakes, sequels/prequels, spoofs, references... So if someone describes their movie as having a scene "that seemed to be a reenacting of the Hippogriff flight in HP3", it might be worth taking a look in the References section of Prisoner of Azkaban.
Sometimes what appeared to be a movie was actually a TV episode, so it's worth looking in those. I've noticed that a lot with people remembering "that movie" which turned out to be a Twilight Zone/Outer Limits episode - there are about 250 of them, that may be worth looking into.
A lot of them being transcripted on the Internet Archive/Project Gutenberg, the operators
site:gutenberg.org can help.
A couple of commonly asked short stories: "The Profession" by Asimov (13 results!), "The Road not Taken" by Harry Turtledove (14)...
Know how sites behave. The main wikias for Marvel and DC for instance are rather complete but they're so complete, it sometimes overlaps in the results. Which is why to track down the exact issues in a comic run a that lasted only 43 issues you'll have to tweak the query to make sure the names appear as featured character, by providing their alias and exact name, in quotes:
"toad mortimer" "husk paige" "wolverine and the x-men vol 1" site:marvel.wikia.com. And read through all 19 results to see which ones actually match.
Note on Wikias: Wikias... aren't always reliable. As much as possible, back every info you fetch from them with the original. If you do, the how you get the original is up to you and your country's law about copyright.
Golden rule of fanfic-ID search: use incognito browsing! Seriously, some of them are really bad and you don't want your browser's system to affiliate you and your cookies with those. REALLY!
For the common sites,
site:quotev.com/story will limit to the works.
Anime (and to a lesser extent cartoon, but mostly anime)
Anime-ID is one of the most painful identifications there are, because from what people forget, it appears that the anime production is 40% mecha, 40% summoning monsters, 20% something else. (these statistics are mostly headcanon, but still.)
There are tons of summon/mecha/robots animes and I can't really provide guidance to search those (other than "good luck, have fun"), but some of them tend to come back a lot:
It's good to know which ones come up the most, because this way, you can read the new "summoning monsters" anime-ID, cringe at how unsearchable it looks, remember the 4-5 most commonly IDed ones, and go browse their Wikias to see if they involve the one "unique" bit in the new question.
For the record that's exactly what I did with Huntik. I identified it once in April (had some luck with Google), and self-plagiarized this answer into three other answers. So 4/5 Huntik answers are from me, just because I went to check if "that thing I IDed once" fitted. It's enjoyable :-)
I mainly use Goodreads but I don't really know how it's structured. I'm open to suggestions!
Google Books works too, although that's not something I use much either so once again I'm open to suggestions.
Don't limit yourself to English
OP say/imply they watched/read it in a non-English language (even though it may have been produced in English) and your English keywords don't yield anything? But you don't know said non-English language?
This is the 21st century, and Google translate/DeepL/WordReference have improved greatly.
- Open a Google Translate/DeepL tab and a WordReference tab (or any online dictionary). WordReference is there to translate word-by-word. Not always the best course of action (don't tell that to your Spanish teacher! ;) ) But useful for Google.
- search for keywords' translation with WordReference, throw them into Google - set Google's language to the non-English language!
- you'll probably end up on a foreign-written site which you have no idea what it's saying. If it looks like it could be a review of the book you're looking for, paste the page's link to the Google Translate tab. Proceed;
- repeat with different keywords and operators until you've found it, or gotten sick of it and rage quit. :)
- Alternatively, learn some foreign languages, but you know, effort.
As you can see those aren't especially complex queries. The rationale behind all that is that if it was translated into a foreign language, there must be a documentation of sorts somewhere... And if it wasn't translated to begin with, then you won't get a lot of English results!
SFF is a good resource too
At the time of writing, there are 12,946 story-identification questions, 1,155 of them being duplicates. Meaning: what you're looking for may have already been answered here.
I suggest browsing
[story-identification] keyword1 keyword2 to start; don't include a media tag as a lot of old questions are only tagged with story-identification. Don't put too many keywords, because SE search sucks and you might miss on some. You can try searching the site through Google which has a smarter search engine.
Searched for something, came up blank, and getting a bit annoyed? Favorite the question. Check if it has gotten an answer - you will likely memorise it for later, and if it doesn't get an answer, maybe you'll eventually think of another set of keywords.
Browse old stuff! And re-read them even if "you've searched for them a month ago"! You will have learned new techniques and new works in the meantime, that might prove useful. Plus, sometimes you might encounter two unanswered story-IDs which look complementary - maybe they're the same thing and the info from one can help you kill two birds with one stone.
Lead by example!
Per my answer to Can somebody walk me through how to research these Identify these story-ID questions? (yes, it's technically a dupe, but we're too far gone for that), lead by example. If you find something by searching the internet, please include the successful query, as well as a quick commentary if the thought process behind it isn't obvious to grasp.
This will help future researchers! No need to copy-paste the whole answer - see the link above for further details on that. If anyone's curious here are my winning queries (easy to search as I reuse the same "found with the Google query..." template).
One should feel free to ask other users how they found the thing. Whether they reply or not is another matter, but it's always worth a shot!
To be continued...