Whenever you simply enter a URL in a browser, the browser sends a GET request to the server. This also occurs when you click on a hyper link. If you do not have a proper implementation of doGet(), then regular browsing is not possible. If you're lazy, simply call doPost(request, response).
Because GET is used when you enter a URL or active a hyper link, it is very useful for passing parameters without the need of a form; anything that appears after the ? in a URL is the query
string, and it forms the parameters. The only advantage POST as an HTTP method has is that the parameters are not sent as part of the URL. This does form a very little bit of security (you can't see the parameters in the address bar), but more importantly you can use as many parameters as you want. With GET, the browser is usually limited in the URL length, and if you add too many parameters the browser will simply refuse the URL. I've seen this happen with Internet Explorer 6 quite regularly.
Although I mentioned that POST provides a little bit of security, it is not secure at all. The only security is in hiding the parameters in the browser, but they are still sent plain-text across the Internet and can be intercepted by anyone. To prevent this, use POST in combination with the HTTPS protocol.