It’s always been perfectly acceptable to me to have an underlined text link in the center of a large graphic, suggesting that the user should click the text link when, in reality, the user can click anywhere on the entire image to the exact same effect. In other words, the text link is something of a red herring.
The idea, of course, is that some users will notice the text link and click it, while others might notice the finger cursor as they roll over the greater image area. The primary goal is to get the user to just click the damn thing, and no harm is done by the subtle inaccuracy of the fake text link.
I’ve always said that foolish consistency is the hobgoblin of information architecture (I am still the 12th hit on Google for my SIG-IA take on the Emerson quote). This is a great example, I think, of an instance where blatantly breaking the rules makes for a more usable interface.
Do you disagree? Do you think that this practice is an abomination?
And let’s not get into the web style proscription against linking from the word “here”, which I pretty much agree with. Click here for more on that hot topic.