Kahneman and Tversky, the guys who first really began to probe human cognitive errors, found in their research that there was a systematic human tendency to either under or overestimate the expected value of a reward that varied as a function of time.
I'll give you an example. Do you want a dollar today, or 10 dollars in a year? Most people will say a dollar today. How about a dollar today, or 10 dollars in a week? I would take 10 dollars in a week. But does this make sense? I mean, in the first case, I make 9 dollars fewer. Yet, in the second example, I suddenly flip my preference.
It turns out that human motivation is heavily influenced by expectations of how imminent the reward is perceived to be. People overestimate the value of the reward if the reward is imminent, and increasingly discount the value of the reward, the further away it is in time. In other words, your perceived utility of an outcome increases with temporal proximity.
So playing skyrim now is more valuable than an A on your paper until temporal proximity increases the value of that A on the paper, which is when you stay up all night finishing it. One way to try to work around it is to give yourself an immediate reward for forward thinking goal oriented behaviour. You can, for instance, create a situation where you will reward yourself with a tub of ice cream when you've put in 4 hours of work. How can you get around it without resorting to this? Have high level executive functionality. Which to some extent can be exercised by simply practicing doing things you don't want to. So, start small, do 5 minutes of dishes per day. When you have no problem keeping up this...