> The Internet is full of terrible cooking information, and if you don't have any authoritative sources at all, it's hard to separate the good from the bad.
The internet is full of terrible information in general. That's why PageRank was invented. Is there something unique about cooking information that means PageRank won't work with it?
YES. Recipes are a long-tail topic. The sites that do best in PageRank are the ones with the most recipes. The sites with the most recipes tend to either be overtly user-generated or slyly repackaged UGC. Moreover, as Patrick will explain, most people who search for prepared dishes aren't actually going to prepare them, which means Google has little incentive to ensure the tops of the SERPs are good.
I'll find the eHow "how to cook a steak" that was pegged to the top of a SERP in a bit.
Another counterexample to your belief about the miracle of PageRank: try searching for symptoms some time.
There are many (non-tech) long-tail topics that I've successfully searched for using Google, so I don't think this is a generalizable statement.
> The sites with the most recipes tend to either be overtly user-generated or slyly repackaged UGC.
I can't speak as to the average user, but I am usually able to easily ignore/sift through such sites. They pop up frequently in all kinds of searches these days, and is one of the clearest signs of Google's results going downhill.
That said, I have no personal experience looking for recipes online, as my cooking knowledge was largely learned from my mom, and I've had no reason to turn to any other sources so far.
> Another counterexample to your belief about the miracle of PageRank: try searching for symptoms some time.
When did I suggest that I believe that PageRank is miraculous? It has its pitfalls, just like any other algorithm.
And I've actually searched for symptoms several times and successfully self-diagnosed (confirmed later during a doctor's visit). So this example of yours doesn't hold up. My success there may have been swayed by my professional biomedical knowledge though.
I had a cold a few weeks back, and a coworker found me a search term that prominently featured cerebrospinal fluid leakage as a possible cause for runny nose. Just to establish how far apart you and I are on the usability of the Internet for this.
Oh, it's definitely possible to find all kinds of wacky and incorrect theories at the top of a SERP. Remember that PageRank depends on internet users being (mostly) correct. And oftentimes users are not. Take any political topic as an example. But this is generally solvable by phrasing your query correctly.
But you can still sometimes see this problem with medical topics because heavy-handed government regulation has largely stifled innovation in the medical industry. New technology adoption in the industry is extremely slow, and from my own professional experience, I can tell you that getting doctors to do something as simple as type their notes (so they're available to patients online) instead of handwriting them is akin to pulling teeth.
The protectionism that has resulted from constant lobbying by the AMA has resulted in lots of doctors being able to avoid the adoption of new technology, so discussion of medical topics online by knowledgeable individuals is rarer than in other fields. Perhaps this will change as baby boomers age and medical costs skyrocket, but I'm definitely not holding my breath.
Also, there is of course the issue of most people searching for this stuff not being very knowledgeable, so it creates some confusion on the part of the searcher, what with all the new vocabulary. As for your cerebrospinal fluid example, it's very possible that I see those sorts of results, but I just tune them out due to how absurd they are.
Edit: you were probably talking about CSF rhinorrhea[0]. As it says:
> Most cases of CSF rhinorrhea occur after major accidents where the bones of the face and skull experience significant trauma.
Ergo, if you haven't been in a major accident recently, disregard. Doesn't seem that complicated to me.
There's a couple of issues specific to cooking and recipes that I feel are worse than other general information. Lots of home cooks and unvetted recipes use unsafe cooking practices--not cooking meat long enough or getting it hot enough (or storing things cold enough), it's common to gloss over or skip important steps, online recipes they can often leave out important details that would be caught by an editor. I recipe I made the other day just asked for 2 cans of tomato puree--didn't specify the size of the cans. My mom has a few excellent recipes that just list the ingredients and you mix the amounts to taste.
None of these are problems for people comfortable enough with cooking and know what you're attempting to make, but if you're trying to learn best practices I think it's easiest to find a reputable cookbook for reference. I agree with that the author said, many cookbooks are terrible. I think it's going to get worse with self-publishing and ebooks (I'm definitely not knocking them--but I think in this case it's going to lower the quality).
To your specific point, I agree with the sibling comment about it being an aggregate for the whole site and people linking to stuff they have no intention of making.
Wow do I ever disagree on undercooking. American cooks have exactly the opposite problem, and the literature suggests our guidelines are way out of whack too.
The internet is full of terrible information in general. That's why PageRank was invented. Is there something unique about cooking information that means PageRank won't work with it?