Basically the only thing that I found relevant or interesting about my accounting class in college was the concept of sunk costs. Put very simply, a sunk cost is money or resources that have already been spent and cannot be recovered. You generally should not factor those costs into decision-making after they have spent, because they are rarely relevant, in stark contrast to what seems logical.
[Note: I may occasionally butcher economic terminology here, but the theory should be relatively sound. Always listen to your professors.]
In my most recent example, my sunk cost is the amount of money I spent on my car—or rather what is now my previous car. I specifically paid a premium on the car to get it brand new: it had only six miles on the odometer when I purchased it. That premium was somewhat difficult to accept, and I thought quite a while about whether I'd want a new car or whether I would accept a used car for considerably less. But eventually I decided it was worth it. It seemed logical to me at first over this past week that since I have already paid this premium once, it would be crazy to do it again and buy another new car to replace it. But what I paid for the first car is no longer relevant. It's a sunk cost. In fact, since I decided last time to purchase a new car instead of a used one, and I did not feel buyer's remorse after doing so, it makes more sense to buy a new car again this time around. If I buy a new car again this month, the important thing for me to remember is that I'm not paying that premium twice; I'm just paying it once. I should only consider the amount of money it adds to this car purchase, though I can then weigh that value against the amount of satisfaction I received a year ago from buying a new car, assuming I would receive that same satisfaction a second time.
It actually reminds me a lot of the common misconception that if you flip a fair coin nine times in a row, you have a much higher chance of getting tails the next time it is flipped. It makes so much sense at first, but people who have taken a basic statistics class know quite well that the odds are still 50/50 on the tenth flip.
The money spent on that last car is only tangentially related to the decision of whether to buy a new car or a used car this time around. It could have been five hundred bucks or half a million; it doesn't matter. I could have bought nine new cars before this car purchase or none. What matters is that in a week, unless some sort of miracle happens, I will have no car, a certain amount of money, and a desire to purchase a new car. In almost every way, it is an entirely new decision, and the only way in which the previous car purchase impacts things is in the experience it gave me.
So in making my decisions about what kind of car I'm going to get to replace my beloved Civic, I am trying to be as rational as possible, and not let history unduly influence me. In a week or two I will need to buy a car. It will be almost exactly like the situation of a year ago, just with an ability to make a larger down payment. How much money I "wasted" by getting a new car instead of a used one is only relevant in that it helps me decide if I will be happy doing the same thing again. The amount that my last car depreciated over the course of a year is also not very relevant. What matters is that I now know what the experience of buying a car from a dealership is like, what it's like to have a new car, what I liked from my Honda Civic (I heart the digital speedometer so much), what I like in the way a car feels, and those sorts of other non-finance facts.