Far-left policy or far-right policy? Wait until the perfect moment or take action right now? Plan everything out or just go with the flow? In this world, heuristics often conflict each other.

One of the possible solutions to resolve this conflicting heuristics and choosing the right one is this: ** stay in the middle (since the appropriate solution is usually easier to reach from the middle), see the terrain/context, and then swing a little to one side or another depending on the feedback given by the terrain (move closer to the appropriate solution). **

For example, in most situations we must not wait until the perfect moment, nor blindly take action right now. Instead, we should first decide that we will until the moment with an acceptable probability of success (stay in the middle). We must see the situation to spot such moments (see the terrain) and then take action. If we find out that such moments are hard to come by (feedback from the terrain), then we might need to lower the standard of required probability of success (swing a little to one side depending on the terrain). On the other hand, if we find out that such moments occur frequently enough, then we might need to raise the standard of required probability of success (swing a little to the other side depending on the terrain).

A meta-problem of heuristics with the same solution:

  • Heuristic 1: Sometimes, we need to start from the middle and swing to one side or another depending on the terrain (as stated above).
  • Heuristic 2: Sometimes, we need to start from one extreme and move to the other side depending on the terrain (contrary to what has been stated above).
  • The solution is the same:
    • Stay to the middle in most situations.
    • Recognize that you may need to start from one extreme (not the middle) based on the extreme rareness of a situation, your past experiences with said problem, the risk and reward structure of the conflicting heuristics (e.g. it’s better to overestimate your death-match opponent than to underestimate him), etc.
    • Decide whether to lean to one extreme based on the factors stated above, or stick to the middle.

ASCII art for illustration (Hunting the Invisible Ant in a Dartboard, 2D version):

‘A’ here represents the invisible Ant or the Appropriate heuristic to use. ‘D’ here represents the Dart or Decision you made. The closer you hit the dart to the ant, the louder the ant will scream (this is the feedback from the terrain). The process above can be illustrated as such.

|-------------A-----| usually, the appropriate solution is easier to reach from the middle
|---------D---A-----| so, we start from the middle and observe the terrain
|-----------D-A-----| and then try to move as close as possible to the solution
|------------DA-----| by swinging a little bit to one side based on the feedback

Final Warning: Situations can change. Just because the target ant is at the right side right now, doesn’t mean that it won’t go to the left the next moment.