This is an essay on the deletion policy.
|This page in a nutshell: These are some arguments that have successfully saved articles from deletion in the past, or otherwise supported one's cause, and therefore, may support yours.|
|Arguments to avoid in|
|Arguments to make|
When notability is in doubt, and that is the reason given for deletion, the very best way to counteract that is to demonstrate notability. Especially if the topic sounds obscure or trivial or gives the appearance of original research without the references, proving otherwise will solve this problem.
If you feel you do not have what it takes to improve an article itself when it is up for deletion, you can recommend sources that others can use. Google News, Google Books, and Google Scholar are good places to find sources.
On the other hand, if the page does not appear to be notable, and you believe it should be deleted, the best way to get the page deleted is to prove that. Simply having no references on the page may not be grounds for deletion; you will have to demonstrate that none can ever likely be found. As for articles with a single external link to the subject's own site or MySpace page, this may very well be self-promotion (as in the case of the garage band).
An AfD discussion is not a vote. It is a discussion of whether policies (and broadly accepted guidelines, such as many of the topic-specific notability guidelines) allow or disallow the type of article. Referencing policies, and where appropriate, guidelines is what will make or break it. Even if ten editors state an article should be deleted, and one editor states the article should be kept, but the one who wants it kept gives a good argument citing policy, while the other ten give none, this is sufficient grounds for keeping an article. In the case of guidelines, which carry less weight, it may be less clear cut, but basically the more support an argument has from well-accepted guidelines that reflect Wikipedia community consensus, the more likely it is to prevail.
When you make your comment on an AfD board, familiarize yourself with as many Wikipedia policies and guidelines as possible. There are so many, it may take time to know them all. Keep in mind that what you are saying is not a vote, and without citing one or more policies and/or guidelines, agreeing with someone else's citation of a policy, or rebutting someone else's citation of a policy, your comments will have little if any weight against the consensus formed by others and the decision made by the closing admin.
An essay, unlike a policy, can be unilaterally written by one person. Though it can be edited freely like any article, it is not subject to challenge or scrutiny like a policy (if people strongly disagree with it, they tend to write a competing essay). Generally, any editor who knows anything about writing an essay has enough wiki knowledge that they will be making good edits. Most essays are written in good faith by users who are coming up with valid interpretations of existing policy, often based on their own experiences. Many essays cite policies within to support the author's cause, and some essays do eventually become policy. Especially if the essay was written by someone other than the one who wrote the article, or even if it was written by the same person, if it was written before the article was proposed for deletion, it should be taken into account when cited in favor of a cause.
Notability is not temporary. A single event that receives coverage only for a short period of time and never again is usually not notable (though there are exceptions to the rule). If there is significant coverage for a long period of time, and the subject becomes a permanent fixture on at least some notable members of society, the subject is more likely to be notable.
The WP:BLP1E argument is often cited as a reason for deletion. This can be counteracted simply by showing that the subject has been involved in more than just one event.
Stubs are permitted, but many stubs have been proposed for deletion on the grounds that they are dictionary entries, and Wikipedia is not a dictionary. Even if the page you create is very short, and you mark it a stub, the one referenced statement you provide to show it goes beyond a dictionary definition can be valuable.
If you want to save an article on the basis that it is more than a dictionary entry, the best things you can do are to add some sourced encyclopedic information to the article, and to demonstrate that more sourced information does exist.
Many articles start out looking imperfect, sometimes really terrible, and may be really great one day. By showing the potential an article can reasonably have in the future, this may be a good reason to favor keeping it.
Wikipedia is not about guess, guess, guess. Even if an article does not appear to have sources making it notable, being linked directly from a significant number of other articles in a manner that the links in those articles show that the reader of those articles would want to know more about the blue-linked subject shows that the information the article contains is valuable in defining and providing information on a subject already described in multiple articles. Even the bare mention in other articles demonstrates notability. Deleting the page would then create red links in a lot of other articles.
For this purpose, lists, disambiguation pages, see also sections, and certain types of templates (such as hatnotes or navboxes) are given less weight, and it is best to name at least 3 articles that contain links to the page from the text itself.
Being orphaned does not mean an article is not notable and should be deleted. Many articles about notable subjects are simply too difficult to de-orphan. But when other factors favor deletion, linking may help.
Long articles have many problems. They can be overwhelming to read. They can be slow to load in older computers and in many mobile devices. They can be hard to edit. It is for this reason that it is a long accepted practice to split a long article into two or more smaller articles.
Exactly what information to split into separate articles and how to split is done on a case-by-case basis. Generally, it should be done logically, and the subarticles should all be linked from the parent article in a way they can easily be found.
Often, the result of splitting is that the subarticle does not appear to meet inclusion criteria, either because it seemingly does not meet notability guidelines for a standalone article with its sources alone, is sourced by primary sources only, or is otherwise not judged as content worthy for inclusion in an encyclopedia. Nevertheless, common sense says that the information contained in the article does belong on Wikipedia.
If this is the case, one arguing for the article to be kept should identify in their argument information like the title of the parent article (if the has not already been done), the length of the parent article before its division, where the information in the article up for deletion would be located if it were in the parent article, and why it should belong in an encyclopedia at all.
The examples above share a characteristic. Each is clear, concise, and focused which will gain more positive notice from the closing admin than a long impassioned essay lacking specifics. Such lengthy comments will not outweigh other editors and can harm your credibility in the process.