We should think of our beliefs and the evidence we engage with as if we had a little homunculus tv courtroom in our brain adjudicating whether to admit evidence into the record. Obviously, this is incredibly difficult to pull off in real time, but it’s a nice thought experiment to pause and consider the weight of a claim being made.
This idea came to me while watching a YouTube video covering the recent downfall of a famous hustle influencer, where the presenter made an observation that she (the presenter) would normally not take people’s personal lives into consideration when judging their professional work, but the case that the influencer sold conferences and products marketed as relationship coaching courses under the pretenses of having a great marriage was swiftly undermined by her (the influencer) getting a divorce approximately two years later.
I was impressed with this statement by the presenter – she was right! Under normal circumstances, the personal life of a person shouldn’t bear weight on something like this, but given the fact that the evidence under consideration was whether someone was misleading about their personal life and getting others to pay for her “expertise,” it would be grounds to consider this piece of evidence as relevant or bearing weight. My homunculus courtroom judge ruled that the testimony was admissible.
This is a silly thought experiment to anthropomorphize cognitive thought-processes that are otherwise just a black box to me. I suppose it’s a little farfetched to think that we have this much control over our beliefs, but maybe the next time I listen to a claim (or gossip, or something that doesn’t jive with my experience… or claims that I want to be true…), I will remember my homunculus courtroom and think twice about the claim’s believability.
I was reflecting on Seth Godin’s musings about the number of moons in our solar system. The initial assumptions we use to make predictions about our world can sometimes be orders of magnitude off from truth.
We as humans don’t like to be wrong, but we shouldn’t be overly concerned with our initial assumptions being off the mark. After all, if we knew the truth (whatever “truth” happens to be in this case), there would be no need to start from initial assumptions. It is because we are starting from a place of ignorance that we have to start from an assumption (or hypothesis) in order to move forward.
The problem lies with whether we realize we are making assumptions, and how committed we are to holding on to them. Assumptions made about the physical world can often be value-neutral, but assumptions that intersect with the lived experiences of people always come pre-packaged with history that’s value-loaded. It’s fine to make an assumption that your experiences are shared with others, but that assumption can only be carried so far. At some point, you have to acknowledge that there will be a lot missing in your initial assumptions that need to be accounted for.
The lesson then is this: when working from an estimation or prediction, be careful with your initial assumptions. It’s fine to begin with your own experiences, but always put an asterisks beside it because your experience is likely not universal. We must guess, then check. Test, verify, then revise.
Aiming at truth is a noble goal, but we should settle for asymptotically moving closer towards it as it more likely reflects reality.
I have a trick for finding parking at work in the morning. The trick I use doesn’t guarantee that I’ll find a good spot every day, but it does prevent me from wasting time driving up and down lanes when there are no spots available. The entrance to the parking lot at work is at the far end of the lot, with the building on the opposite side. This means that when you start your search, you begin at the furthest point away from the building and your search pattern will take you towards the building.
In terms of strategy, this means that the spots with the highest probability of being empty are both the furthest from the building and the closest to you when you begin your search. This obviously makes sense from a safety perspective – if the cars were entering the parking lot closest to the doors, then pedestrians would be in greater danger of getting hit and traffic would always be impeded. However, this means that it’s hard to determine when you enter lot where empty spots are among the banks of cars. Due to poor lines of sight and the number of large trucks used by students, you often won’t see an empty spot until you are a few feet away.
If you rely on this strategy for finding the closest parking spot to the door, you’ll waste a lot of time driving around except in cases where you stumble across a spot (which I estimate would be a low probability event). I’ve started using a strategy to avoid searching for those spots and reduce wasted time in randomly driving around.
My strategy attempts to address a number of constraints:
My parking utility is maximized when I find a spot close to the door. This reduces the amount of time spent walking, which is good for inclement weather, icy conditions, and because I’m usually running late.
My parking utility is diminished when I waste time circling the lot searching for ideal spots. Instead, I’m seeking a satisficing outcome that balances maximizing utility and minimizing search time.
I’m competing against other actors as they also drive around seeking empty spots. These people are usually students, who are also usually running late or seeking to reduce their walking distance.
Keeping these considerations in mind, this is the strategy I employ in the morning.
First, I’ve limited my parking search to one of the three lots. By reducing my options, I can make quick decisions on the fly. Lot 1 is directly in front of the door, and since I arrive before the majority of the students, I find that it satisfies my needs most of the time. If Lot 1 is full, I move to Lot 2, and finally Lot 3 being most sub-optimal.
Next, on my way to the entrance of Lot 1, I scan the first row of cars for empty spots there. Since I drive passed it, it allows me to quickly eliminate it if there are no spots, or at least gauge where the spots will be relative to any additional spots in the second and third rows of the lot.
Then, I use a trick to quickly assess the likelihood of empty spots. I look at the shadows of the cars and pay attention to noticeable gaps. When I enter the lot, I can see down the second (middle) row. If I see anything, I drive towards the gap and usually there is a free spot (except in cases where someone has driven a motorcycle and not parked it in the motorcycle-designated lot). If I see no gaps in the shadows, I move on to the third row and repeat the pattern.
The majority of the time, this gives me enough information quickly to know whether I need to drive down a row. There are two limitations to this strategy: first, it relies on there being no cloud cover, and it doesn’t allow for east-facing shadows to be examined. This is not a perfect strategy, but my goal is to maximize my parking preferences while eliminating my wasted time driving around the lot examining each parking spot hoping to stumble onto an empty spot. Using this strategy balances these two interests and generally gives me a satisfactory outcome quickly.
A final consideration I use is to notice cars leaving the lot when I enter, and noting where they are coming from. That is the fastest indication of where a parking spot is on the busiest days when I’m competing against other cars looking to park.
All of this occurs within about 15 seconds of me driving up to the lot at work.
If you have reached this point in the post, you might be wondering why I spent so much time explaining how I find a parking spot (is this really the best use of a blog???). I think this example of setting up a solution to a problem is a fun way of explaining how I ideally like to approach a problem. I try to consider what outcomes I’m aiming to achieve and work backwards to consider options that would fit those criteria. In doing so, I have to consider what input I need to let me quickly assess a situation and make a decision by eliminating extraneous options.
It’s important to know when you need to be right, and when you need something to work well enough most of the time. For instance, if this were a higher-stakes situation (say, I was doing surgery), I would want a strategy that would be the equivalent of finding the closest spot to the door every time. Instead, I know that my goal is achieved if I reduce the amount of walking time and reduce the amount of time and fuel spent hunting for an optimal spot.
When coming up with a strategy, I knew that hoping to stumble across an empty spot would be a net increase in my search time. So, I found a way to quickly gain information that would eliminate many non-options. Rather than looking at the cars themselves, I instead look for gaps in shadows – an indirect indicator of outcomes I want. It’s a simple heuristic that eliminates the need to confirm that cars are occupying spaces all the way down the long row.
While the strategy will not save me time in 100% of cases, it does shift the outcomes to a net decrease in search time, which meets my goals and gets me to work on time (most of the time).
A Paramedic's Notebook 