I am curious about the problem from the perspective of a startup idea / nonprofit website. What would you look for in an authoritative source that doesn't exist now? And which topics were you looking into?
It's not really that they don't exist, but it's silly stupid to find them now.
For instance - ddg, google, doesn't really matter the search engine - typing in something like "ohtani batting average with runners in scoring position with 2 outs" used to be so easy it'd practically pop up in the quick answers with the exact link to the ESPN table or dudewhoneverleavesbasement.com that has tracked every baseball stat for the last 150 years.
Now you just get basically the same results for "ohtani batting average" or "ohtani news." It's impossibly frustrating. It's like that with everything now.
I guess that's more of a problem with search than anything else, but honestly, search is the main way people navigate the web. If I want anything authoritative, honestly, I'm probably going to wikipedia to see a source list and then checking out authors in the local library. Anything dated post 2022 I consider inherently unreliable.
I think the issue is, dudewhoneverleavesbasement.com stopped working on his website and his potential successor is more likely to be trying to turn that information into an interesting youtube video.
I think this is definitely an issue but not the only one. In the game knowledge part of the web, there used to be personal sites written in plain html of some guy’s faq who figured out every single thing in a final fantasy game and google would draw on stuff like that. Now you get AI SEO spam heavily monetized drivel that contains a lot of inaccurate or outright false info.
I think the small sites like that are either mostly gone or simply not ever indexed.
I've had the same bad experience with search as you've had, but my conclusion was that AI tools are just going to replace "using a search engine" in the near future. Google is more and more unusable to me.
That doesn't exactly solve the authoritative problem, but theoretically it makes it more centralized (one AI company vs. innumerable websites.)
What worries me is as soon as we accept AI results with no fact checking (or ignore it) they will become the means to directly manipulate people without any recourse for judgement about veracity. Like phishing emails with perfect spelling and well formatted HTML.
On the one hand this is definitely a potential issue.
On the other hand I’m not sure how the current state of the web is really any different. We just rely on Google or whatever search engine to “fact check.”
You're looking for https://stathead.com/baseball/ but it is, for better or worse, a subscription service. You can get the raw stats from https://www.baseball-reference.com but you'd have to make your own filters to get splits like that. I have no idea if these surface highly on search engines (I guess not), but thankfully I've used primary sources for statistical information and raw data for decades and simply remember what they are without needing a search engine. I'm good as long they continue to exist, I guess.
Historical data is available in a number of places and is free to use for whatever purpose, commercial or otherwise.
More recent data is not. Major League Baseball has wide open APIs that give you a huge amount of nearly real-time data, and runs a site where you can bulk download data. For personal, non-commercial purposes only. They will go after you for running a commercial service and demand licensing fees.
That's basically why everything polished is now behind a subscription, while you can DIY all you want.
The authoritative source has to be .. authoritative. And it also has to be timely, i.e. updated in real time. This is hard. Being an authoritative source for certain sub-types of information is definitely a viable business, but only if your information is being used by other people to make business decisions. See, for example, the Bloomberg terminal.
