# ASAP and ALAP In decision making we either want to act as soon as possible (ASAP) or as late as possible (ALAP). We want to act ASAP when the information that we gather will not be very useful in improving our outcomes. These are examples where the feedback we get of just quickly making the decision and seeing what happens is very valuable (see taleb on tinkering and smal, convex risks). On the other hand, we want to act ALAP if the information we gather may drastically improve our outcomes. ### 1. Heuristics #### 2.1.1 Decision Matrix In order to reason about the amount of time spent on each given decision it is helpful to have a few heuristics in our toolbox. Recall the [decision-matrix](decision-matrix.md). Below, we have a rough guide for the time that should be spent for each type of decision:  Decisions that are *reversible* or *inconsequential* (lower left hand quadrant) have the following property: moving slow and letting them linger is **the biggest risk**. Gathering too much information becomes a **drain on resources**. The best example here is that of buying a mattress online. Their are an overwhelming number of options and near infinite research that can be done to attempt to make the best decision. However, it is by definition an incredibly reversible and inconsequential decision. Mattresses almost always have long return policies, and they cost less than $1000. Interestingly, in a sense the best way to gather more *quality information* in this type of a scenario is to *make a decision*. This is because the most useful information when it comes to buying a bed is "does it feel comfortable and do you sleep well on it?". The best (and realistically only) way to gather this essential information is to make a decision, buy a bed and try it.  Decisions that fall in the upper two quadrants require us to spend more time **gathering evidence** and **deciding**.  For decisions of this type the biggest risk is moving too fast and missing something. The next question we must ask is how much time should we be spending? #### 2.1.2 Time to Decide vs Usefulness of Information This leads us to a new heuristic that will help decide when to gather more information and when to act on the information we have. Our heuristic is as follows: > `Time to Decide` is a function of `Usefulness of information`. As the usefulness of information (as it pertains to the decision at hand) increases, the time to decide should also increase.  It is worth noting two things about the diagram above that leave it to be slightly inaccurate: 1. Time to Decide is a *function of* the Usefulness of Information. When I say "function of" I am meaning that the direction of causality goes from Usefulness of Information to Time to Decide. The above diagram has it reversed. 2. The function is not linear. It is important to not get stuck in the middle (for instance taking a medium amount of time to decide with moderately useful information). So, for decisions that fall into our upper quadrants we want to decide *as late as possible (ALAP)*. --- tags: #decision-making #heuristics links: [Decision by Design MOC](Decision%20by%20Design%20MOC.md) [When-to-Decide](When-to-Decide.md) created: 2020-11-25 modified: 2020-11-25