Wikipedia:Wikipedia is not a spelling checker

Wikipedia is not a spelling checker. It is not a substitute for a reader's ability to spell. Redirects from typos are OK, but redirects from misspellings are evil.

Did you mean Pea Sea? Imagine the amount of redirects we would have to create if we were a spell checker. And the WP:SURPRISE they would find themselves in when they wound up at Cache instead of Cash.

Wikipedia aims to provide readers with many ways to find the information they are looking for. Through systems such as Wikipedia:Redirects, Wikipedia:Categories, Wikipedia:Portals, Wikipedia:Infoboxes, Wikipedia:Navboxes, Wikipedia:Projects, Wikipedia:Disambiguation pages, Wikipedia:Set index articles and, finally, the Wikipedia search engine, editors make a great deal of effort to try to get readers to the information that Wikipedia has. Others provide software to make searching Wikipedia more intelligent, less error-prone, and more accessible than ever. Projects such as Wikidata allow things to be found in other languages, and projects such as Wikipedia:Pages needing translation mean that even foreign-language articles in English Wikipedia are often translated, voluntarily, for an English-speaking audience.

Indexing Wikipedia intelligently takes a lot of time by a lot of people. A spelling error that might present a reader with a "Did you mean" result from a search engine is not to be blamed on Wikipedia. It is the fault of the reader who cannot spell. The reason some editors dislike allowing the growth of Category:Redirects from misspellings is that, to correct them requires an educated guess of what could possibly be the intended target. Even if the target is obvious, it masks from the reader the fact he or she has made a spelling error. If the search term was copied from another work (their final thesis, for example), that error may then continue to persist in that other work.

As it stands, {{R from typo}} and {{R from incorrect spelling}} are categorised as the same thing: but that is a mere technicality, and editors should be careful to distinguish. We all make typos, but some of us can spell. There may come a day when we split those categories.

Of course, we encourage and allow variations in varieties of English. They are neither spelling nor typos. But downright misspellings and typos can often make it harder, not easier for others to search. Just because one reader created a redirect because he or she can't spell or type, does not mean every other readers will: they may have spelled and typed accurately, and be looking for something else. They will more likely end up with a surprise to find themselves not where they intended. And we, as analysts, have no way of tracking that, because although we can see the page hits, we have no idea whether they were hit intentionally. We do not even have an heuristic on how long someone stayed on the page if it was the "wrong" page. We can only use common sense. Despite our various expertise, often it is better to delete a misspelled redirect. A reader presented with a search page at least knows he or she did somethin wrong. A reader who gets sent to the "right" page neither knows nor cares; one sent to the "wrong" page gets puzzled and frustrated.

In short, redirects from misspellings are evil. They do not encourage readers to learn the correct spelling, they either confirm their misspelling or leave them frustrated. It is better to give the honest answer, "we don't know what you mean", with a set of search results, and leave readers asking themselves "Hmm, I must have got something wrong, why do I have the search results?". Better than them shouting "Hey! I didn't ask for this page! How do I get out?! Stupid Wikipedia, I wanted HSBC, not HBSC!".