Originally Posted by
Pannonian
If you want to call yourself pragmatic, use the engineering model.
1. Do not use drastically new solutions, unless it is thoroughly tested, first in theory, and then in practice via small scale models. If you can't provide proof of concept, assume revolutionary new solutions do not work.
2. Use tried and tested solutions, adapted to the problem at hand. The closer a tried and tested solution is to the defined problem, the more reliably you can assume its implementation to be. If the context is completely different from that of your chosen solution, do not assume that it will work.
3. Peer review by acknowledged experts. There will always be differing views. But if an overwhelming majority of experts on the subject hold an opinion, it is more likely than not that their received opinion is the correct one.
4. When implementing, start with as small a scale as possible. Always leave yourself a way to reverse your actions.
5. If evidence suggests your model is wrong, reverse your actions and return to the baseline.
Brexit fails every one of these points, extravagantly.