(Foreward: This is expressly for academic purposes only.)
Suppose I write a simple robot (e.g. Java.robot) that allows me to automate a very simple task, i.e. continually perform a mouse click in the same exact part of the screen over and over.
Since I wouldn't be using something like Mechanize in Python, is there any way the website could still detect that a non-human was inputting mouse clicks? Would a savvy website owner have a reason to put in this kind of detection, e.g. if the rate of mouse clicks per second is too high to be a human? In that case I could just add an arbitrary delay between each mouse click (e.g. 2 seconds), then it should be impossible to say "Aha, this user is using a robot and therefore we can flag / ban him."
Thus my question is:
If I use a robot to perform keystrokes / mouse clicks at rate which is decidedly human, is there any way the website that I completing forms in could conclude it was a non-human agent, and therefore possibly take action against my account?
Note that I'm not programmatically opening the website--that is something I manually do, then knowing the layout of the widgets I just have the robot make the appropriate clicks / keystrokes in the right cycles / patterns.
This type of cheating is common in browser games and web advertisement, and companies have measures to detect such behaviour. There are different approaches, from statistics on the clicking interval and timing, as well as tracking mouse movements inbetween of clicks.