I know the discrepancy between Skritter time and real time has been discussed, but I'd like to revisit this issue. I believe that Skritter time should be eliminated and the clock should count real time.
Like many users, I set a daily goal of a definite amount of time studying. (This is in no small part because the review bar is hopelessly broken--as far as I can tell, it's impossible to get to zero reviews due, so using that as a goal would just be masochistic.) In an effort to meet that goal, I very often find myself rushing through reviews I was wrong or so-so on rather than drilling down, rewriting the character, visualizing mnemonics, and reading example sentences. Perversely, I probably end up spending more time learning the character in the long run than if I had spent the extra few seconds then. Intellectually I know I'm only cheating myself when I move on past a shaky character to keep the clock running, but that seems like the natural psychological response.
Furthermore, even if I did have the psychological discipline to drill down on every character for as long as it takes, the current Skritter clock would make it impossible to stick to a consistent study regime when doing that. (Well, not impossible; I suppose I could look at my watch at the beginning of a session and stop when I've reached an hour of real time, but that gets difficult when I'd have to keep track of multiple sessions, at home and at work. Besides, keeping track of time like that is a textbook problem that's hard for a human but easy for a computer. The website should be doing this for me.)
It simply makes no sense to me that the types of activities I mentioned are treated by Skritter as if they had no value. This should be changed, and (accepting Nick's invitation to vigorous debate about "where you wish we'd spend our development efforts") I think it should be a priority. At the very least, there should be an option to display a real-time clock with the amount of real time spent skrittering that day. Fixing this will become even more crucial when the example sentence system is overhauled, as I expect to want to devote more time to reading example sentences then, and that time should be counted.
Postscript: Upon being referred to the FAQ on this subject by the Professor, I see that "We plan to improve the AFK tracking at some point, so that it's smarter about when you are and aren't studying. When we do this, we'll also try out a mode where it keeps counting for a few seconds after you answer each prompt (though not too long)." This is, frankly, even more baffling. Once the AFK tracking is improved, Skritter should know whether I'm away from the keyboard. If I'm not, the clock should be running. I don't see why it would count only for "a few seconds after" I answer but "not too long." If it knows I'm at the keyboard, that time should be counted. Anything else seems like the software trying to make what I think is an inappropriate qualitative assessment of the value of time spent on core study versus peripheral activities like example sentences or mnemonics.
Please discuss. I'm particularly interested if anyone sees value in keeping track of "skritter time" as it is currently understood. I can't see a reason why I would want to keep track of just the time spent trying to think of the answers to the exclusion of keeping track of total time spent studying.