Learning About My Brain From Typing Z-A

Speed Typing Online's Type the Alphabet page

Recently, I challenged my girlfriend to see who could type A-Z the fastest on Speed Typing's "Type the Alphabet" page. She won :( Regardless, I loved how the site presented the results so clearly and colorfully.

I then thought, "Hmm, I can use this site to study how my brain works." I'll just keep repeating this simple task to try out new techniques and observe how my performance changes.


Logistics: I focused on Z-A, since I already typed A-Z with my girlfriend. Let a word in WPM (words per minute) = 5 characters. Since Z-A = 26 chars, we have 26 letters/5 chars = 5.2 words. Thus, (Z-A WPM) = 5.2 * 60 / (Z-A time).

My first attempt was 8.177 seconds (38 WPM). After just 20 minutes mixed with deliberate practice and free typing, I managed to achieve 3.028 seconds (103 WPM), a 170% improvement. Towards the end, typing Z-A started to feel natural, almost like breathing or simply moving my fingers!

Daniel Kahneman describes two modes of thinking in Thinking, Fast and Slow: 'fast' mode, where thinking is instinctual (subconscious), and 'slow' mode, where thinking is deliberate (conscious). These modes perfectly matched my experience before/after these 20 minutes. My brain followed the progression below, which I'll expand on next.

(100% slow mode) --> (familiar chunks + slow mode) --> (more chunks) --> (100% fast mode)

Starting with Familiar Chunks

First, my brain boosted my typing performance by identifying chunks, meaningful groups of letters, that were familiar to me. For example, I recognized "cba" from "collective bargaining agreement", a term I had frequently seen in NFL and NBA news. In fact, "cba" was easier to type than the letter "q" itself, since I rarely located single letters in my daily typing. Remarkably, my brain located existing chunks with no conscious effort on my part.

Deliberate Practice to Build Unfamiliar Chunks

I soon ran out of familiar chunks, though. At this point, I could either hope to build some muscle memory, or I could direct my brain to focus on building unfamiliar chunks. "zyxwvuts" was the perfect candidate, since I needed to type this chunk correctly at the beginning of all my runs.

The process of focusing on a specific chunk for practice is known as deliberate practice. The concept is obvious: why waste time going through what you can already do well, when you can focus on what you suck at? Well, this concept is hard for the human brain to understand and accept. My brain kept wanting to type the complete sequence to get the 20 minutes over with, while I knew in theory that practicing 'zyxwvuts' would better boost my speed. In general, working on weaknesses takes conscious effort -- one must override the brain's tendency to do what it's familiar with.

Here is my final chunk breakdown of zyxwvutsrqponmlkjihgfedcba.

  • zyxwvuts - Deliberate practice.
  • rqponml - No deliberate practice. Got somewhat fluent since I needed to type this segment to complete the Z-A task. Remains my weakest segment.
  • kjihg - kj and hg are neighbors in Dvorak. Fun to type.
  • fed, cba - Common words. Also fun to type.
I type with the Dvorak keyboard. (I'll leave the full story for another blog post.)


After 20 minutes, I reached 3.028 seconds (103 WPM). How did I do when I revisited this task after my initial learning session?

  • After a 10 minute break: 4.618 seconds (68 WPM).
  • After an hour break: First attempt: 5.274 seconds (59 WPM). Second attempt: 3.516 seconds (89 WPM).
  • After sleep: First attempt: 3.311 seconds (94 WPM)!? After 7 attempts (some failures), I was able to get to 2.657 seconds (117 WPM).
  • After a few days: 3.382 seconds (94 WPM). Second attempt: 2.384 seconds (131 WPM).

I've found that retention vs. time spent is not a linear relationship. With just a bit of upfront effort, I developed a skill that I could perform again much later.

Perfectionism and the Fear of Failure

During my 20-minute session, I did not always do better. In fact, prior to my sub-4 second runs, I clocked some slower runs like 6.2 seconds and 7.9 seconds.

While I felt most frustrated on these slower runs, they also best revealed my weaknesses. For example, I found that, for the chunk "rqponml" that I didn't practice, the letters "p" and "l" specifically gave me trouble. These runs kept me honest, showing how my faster runs could be largely influenced by luck, whenever I happened not to miss "p" and "l".

Thus, an effective workflow is: find weaknesses through failure -> eliminate weaknesses with deliberate practice -> test yourself again to find failures -> repeat.

I couldn't help but think how much of a disservice grades do to learning in school. Unlike my insignificant failures in this study, bad grades leave a permanent mark on a student's record -- and these records can be vital for future prospects like college admissions. Students, fearful of recorded failure, go out of their way to avoid failure. Once they do encounter failure, the pain from such failure repels students from reflecting on how they can improve, which may lead to additional failures over time, establishing a negative feedback loop.

One way to work around the fear of failure in school is to test yourself constantly and honestly before you have to take the official exams that go into your record.

My Takeaways

First, I'm amazed by how much I learned firsthand from this task. While I've already had prior exposure to some insights through sources like Thinking, Fast and Slow and Learning How to Learn (which I highly recommend!), these ideas now feel personal to me.

Second, I realize that strong feedback systems work well to counteract a lazy brain. As I've gotten older, I've started to become more self-confident about my abilities. "Wow I did a great job!" "Of course I know everything!" In reality, my brain has just simply become lazier, because it starts to avoid looking at its weaknesses. Self-confidence = lazy brain. The best way to combat this laziness is to provide it irrefutable evidence of its shortcomings. Speed Typing's slick UI provided that evidence over and over again. I couldn't have learned to type Z-A in 20 minutes without it!

How fast can you type A-Z? What about Z-A? Comment below! You can also reach me on Twitter at @dickson_tsai.


Popular posts from this blog

Tutorial: SQL Analytic Functions

Lessons from Teaching My SAT Reading Series at schoolhouse.world