- We tend to do a lot of tasks that balloon out of initially guesstimated proportions.
- Anecdotally, maintaining a log of what I’ve done has helped me combat the fatigue of feeling like I’m not doing enough.
- I think there’s enough science to explain the reasoning behind this.
Let’s explore 3 statements backed by science.
- Zeigarnik effect tells us that the unfinished tasks are readily recallable by the brain than the finished tasks.
- In his book Thinking fast and slow, Daniel Kahneman talks about the WYSIATI effect (What You See Is All There is), where we normally make judgments about things, based on the information that is readily available to us.
- Anchoring tells us that individuals prioritize the initial information offered (read: readily recallable) in their decision making than more accurate, slower to arrive information.
Putting these 3 things together, I’m going to say that “we attribute more attention to unfinished and ‘to be done’ tasks than the ones that are finished”.
- Unfinished tasks are readily recallable.
- Readily recallable information is easy to come by when forming automatic judgments.
- “How did your day go? / How is it going?” leads to a very automatic judgement.
Always putting the focus on what’s to be done, discounting what has been already done leads to what I call: Todo Fatigue. When it sets in, it feels like you’re running in place.
When we’re doing complex tasks with long feedback loops, presence of intermediary positive reinforcement will keep us motivated. Generally more interesting the problem, harder it is to forecast and break the task down ahead of time.
Sometime ago, I tried writing down the things I did in retrospect, against the things I had to do. In essence, they became “done lists”. A done list is simply a todo list written after the fact. Or, a todo list that gets added with the tasks done in retrospect.
When a todo list generally looks like this:
A done list would look like this:
This helped me in three ways:
- An objective anchor: When I had to answer the question “how did your day go”, it had a objective anchoring point that had accurate information on what I did.
- Accurate weight of the task: It helped me readily answer the question - “Why was X not solved?”, and granted the task with the appropriate weight of the accomplishment.
In the previous example, solving x is much much more difficult and time consuming than bringing milk. Log of things done (done list) communicates that. Forecasting the scope of solving x wouldn’t have been very successful.
How I maintain done lists
I use a simple bullet-journal like method. Every day, I produce a page of things that I did. If they are related to a thing that I planned to do in the day, they are grouped under it. All unfinished groupings get carried over to the next day while all the finished (
[x]) tasks do not. For example:
My personal setup
I stick to a collection of markdown files (one file per day - ex:
2020-07-24.md) and have some automation using bash scripts. I sync it across with iCloud sync. Plus, by using some bash scripting like
cat 2020-07-24.md | grep '\[x\]' | wc -l, I can get a count of things I crossed off for the day. But I think you can implement the same using any text based software, or paper for that matter.
By maintaining lists partitioned by days helps me derive a “relative value” of the day just by looking at the lines with a
[x]. I feel that this is significantly easy for my brain parse than counting the number of items with a special tag in a long running file.
If a task has been appearing with a
[ ] too many days, that’s a signal that it is something not really urgent. Therefore, I can move it to a
someday.md file (much like the someday/later folder in GTD). This helps me reserve my attention to the tasks that need to be processed with some urgency.
But the data input for these can be a chore, especially if you want to track items that are sourced somewhere else like in Gmail or a Trello board for that matter. That is where automation comes in. Text based software is attractive because of the opportunity to automate data entry if you’re technically inclined. However, automation and subsequent quantification is a fascinating subject that I’d like to explore in depth another time.