Five Keys to a Good Prediction

When trying to make accurate predictions about the future, there are five key points I try to keep in mind:

1. Physics - The laws of physics must be obeyed. While breaking the laws of physics as we currently understand them can create exciting new story opportunities in science fiction, it makes for poor predictions about the future. Technology can and will get dramatically better, but it will always be subject the the laws of physics. If you are going to predict everyone will own jetpacks, you need to take into account the energy required to produce enough thrust to move a person and the theoretical limit of different energy storage methods. Is a travel method really practical if you can only use it for 15 minutes at a time?
2. Perfect is impossible - When dealing with an incredibly complex system and a large number of actors, expecting perfection is simply not reasonable. Mistakes might be rare, but they will happen, so they must be factored in. If your prediction depends on something working perfectly, it is not a good prediction. If you are claiming society will adopt giant fleets of flying delivery drones, you need to determine how many people killed by faulty drones falling from the sky society will accept so we can get a 12-pack of razor blades delivered to our doors in less than an hour. People seem to be willing to accept a surprisingly high level of accidental death and human suffering in the name of convenience, but there are limits.
3. People will always be people - If you expect everyone to behave completely reasonably, rationally, and honestly, you are going to have a bad time. Some people are always going to be jealous, delusional, egotistical, sadistic, self-destructive, or just plain mean or stupid. While only a small number of people will fit any of these descriptions at any given time, they tend to have a disproportionately big impact. Many good systems were ruined because a few jerks refused to follow the rules. A good prediction must make allowances for these groups.
4. The past is precedent but not inevitable - History has a remarkable way of repeating itself. Looking for historic parallels can both help shape a prediction and give it validity. The great thing about the future and technology is that truly new, unique things will happen, so it is still possible a prediction without any historical precedent will come true. However,  if after an extensive search you can’t find any historical analogies, you should seriously examine all the logic and assumptions behind the prediction.
5. If it can be done someone will try to do it - The human race is full of enough eccentrics, visionaries, thrill seekers, freaks, trolls, geniuses, and idiots that if something technically can be done, someone will eventually try to do it. They are not guaranteed to succeed, but they will try. Sometimes the results will be wonderful and other times they will be disastrous, but if it is possible it will be tried. 

*By Aeronutronic Divison of Philco Corp, under contract by NASA [Public domain], via Wikimedia Commons