Sensemaking as the Heart of Expertise - Commoncog

Blog Comments
1

A few weeks ago I wrapped up a short series on sensemaking. The explicit goal of that series was to give you better methods to make sense of a potentially disruptive new technology — which at the time of publication (and probably for a few more years yet) will be about AI. But the series was about sensemaking in general, and the ideas we explored together are applicable whenever you have to make sense of new developments in investing or business. This means that the methods may be adapted to make sense of politics, or social change, or — god forbid — war.

This is a companion discussion topic for the original entry at https://commoncog.com/sensemaking-heart-of-expertise/
2

Btw, if you want a concrete example of how this feedback might look like in a military context, there’s a brilliant anecdote from @Latham_Turner here:

https://substack.com/@lathamturner/note/c-250572755

All Navy aircraft train at NAS Fallon — where Top Gun really is. For months we train every mission. The whole air wing in the air at once. 20 to 30 aircraft, each flying its piece of the plan. Usually a strike package going in-country to hit a target and come home. Simulated enemy aircraft. Real bombs.

The real training is the debrief. Every pilot who flew, in one room, for three hours. Sometimes until 1 AM. Often times over a beer or two.

There’s a giant screen showing every airplane and missile in the sky. The replay runs minute by minute.

“Stop frame. Eagle 13, what was the call you made here? What did you see?”

The pilot stands up in front of everybody and repeats the call. Even if it was wrong. Even if it got a plane killed. He says what he saw. Then the weapons tactic instructor walks through what he saw. No judgment, just the expert explaining what he was looking at.

It’s brutal. Every mistake, every missed recognition, every bomb dropped is critiqued in front of all your best friends.

But this is what learning from expertise actually requires.

State what you saw. See the outcome. See what an expert saw that you missed. Update your model.

I’m looking for ways to teach this to my kids. It’s applicable to investing, business, art, music, science. Our education system offers no opportunities to learn it.

Which seems like a miss.

9 Likes
4

Great training technique. Very NDM

2 Likes
5

Thanks for the shout out. I think it was the best training I’ve ever had (and that includes SERE school, which is it’s own kind of training). I assume that training style was probably learned from some of the NDM literature Jared. I don’t think Naval Aviation built that on its own.

5 Likes
6

Relevant: How to 'git gud' at Games (Faster Than Everyone Else) | Raise Your Game

Spend some time searching around for a top-tier player you enjoy watching who talks through decisions as they go. This lets you pause, predict what you would do, and immediately compare your thinking to theirs. That instant feedback—“I would do X, they did Y”—will help you quickly build an intuition that matches that of a top player.

This form of practice is also more time-efficient than playing. All games will have some downtime. Whether it’s time spent looking for a match, unskippable animations, or going through the motions of finishing a game that’s already decided. You can skip right through these when reviewing VODs.

3 Likes
7

YES! I was just reading that yesterday, and noticed that bit immediately.

I had a similar experience to this recently — I’ve been playing Slay the Spire casually ever since my daughter was born, a habit I picked up during my paternity leave. But a few months ago @DRMacIver blew my mind when he told me: “Novices try to build the best deck they can given the cards that they win from battles. Expert players build the deck they need to complete the level, and pick their path through the map accordingly.”

This is going to make zero sense to anyone who doesn’t play Slay the Spire, and I’m a little lazy to explain what this means … but David had apparently picked this up from watching a lot of skilled players on YouTube, and then managed to synthesise it down to a simple principle.

Which … wow. Talk about sensemaking as the heart of expertise — this totally changed the way I approached the game, and the way I approached getting better for the game!

2 Likes
8

Of course there is an overlap of Slay the Spire enjoyers and Commoncog people! The entire roguelite genre is built on making good choices from limited options, in the face of uncertainty about what challenges await you. And roguelites are more fun because there’s limited possibilities and you get to keep retrying if you fail.

I’ll take a stab at it: Slay the Spire is a deckbuilding game with a series of battles. After each battle you have the option to add one of three cards to your deck. Over the course of the game, you pick your cards and try to assemble a good-enough deck to win all the battles along the way.

At the higher difficulty levels, it’s easy to pick cards to win earlier battles, and end up building a deck that has zero chance to beat the final boss (and so, you ultimately lose). Winning decks tend to be ones where all the cards reinforce a common strength - which usually means deciding not to take “good” individual cards if they don’t play well with your deck → skipping card rewards often, and removing cards that don’t fit in well.

And then the rest of the game is map path decisions (picking a path with more battles for more rewards, vs. one with more healing for safety); knowing what possible enemies exist and how to beat them; and then the actual in-battle decisionmaking.

5 Likes
9

I have no idea what Slay the Spire is, but this sounds a lot like effectuation!

2 Likes
10

That’s basically the premise behind @parconley’s The Best Tacit Knowledge Videos on Every Subject

My own contribution to the list is Cracking the Cryptic which is a Sudoku YouTube channel.

4 Likes
11

Had a few hours of housework yesterday, thought I’d catch up on CommonCog, and so I did the whole exercise. 3 cases, voice memos, listen to Cedric’s reaction, another voice memo. Also inadvertently listened to the other Estee Lauder case, not having realized there were multiple.

Cedric’s reactions were very interesting, especially because they often pulled in things that were not written about in the case. When I took MBA classes, I noticed that the resolutions to case studies often involved things that were left out of the case, and it’s part of what led me to start calling MBA classes “LARPing.” Here though, they served merely as enrichment; plus, there had never been a claim that all the answers to an assignment would be in a case, especially as these cases are relatively short. They also mentioned a lot of things I had totally missed, such as the Ample Hills personal guaranteed loans and ensuing personal bankruptcy.

My initial reaction to the Ample Hills study was focused on the things dwelt on in the case study: lack of proper sales forecasting, operational difficulties. And then I got smacked in the face with “It was their cashflow and especially their interest payments – operational problems are fixable and just invite investment” where the interest payments were barely mentioned. When I’d heard all but one location were positive EBIT but the business wasn’t, I had not been thinking of interest paymetns.

And that primed me to think about the next two case studies from a cash flow basis, and perhaps seeing things that aren’t there.

There’s an Amazon review I really hate of Ben Horowitz’s “The Hard Thing About Hard Things,” one of my favorite books. The reviewer says “So all your revenue is from one product and one customer – diversify! Why can’t this silly CEO see a business lesson that’s obvious to me, a teacher turned forklift driver.” As if saying one word can magically conjure up a second product line when the first is struggling.

Reading these case studies, I found myself imagining solutions that I fear someone else could classify in the same category. Okay, so they wanted the Disney location which necessitated they build a factory which necessitated they open a lot more locations. Okay, well…couldn’t they just have a smaller offsite production space instead, or like a random kitchen somewhere in Orlando? The food distributor had major expenses maintaining their warehouse during COVID…couldn’t they just turn off all the lights and AC, or start leasing it for use as a field hospital, or maybe if they’d had a plan for a sudden 2x revenue drop it would have been enough to survive the 10x revenue drop and maybe it would have told them to get two adjacent warehouses instead of one so they could scale down easier. And maybe Estee Lauder could have done a smaller launch with fewer computers, or done better testing to find PMF and proper marketing – and why would “only $900k after Christmas” mean “almost bankrupt” anyhow?

But I don’t know to what extent any of those were real options. Hearing Cedric’s take on the food distributor – that it’s not clear they made any mistakes but still died – was a sobering smack of a business lesson which is obvious in the abstract yet hard to internalize.

(As an aside: I disagree with the idea that no-one could have seen COVID-19 coming. The timing was unknown, yes. But Contagion came out a decade before, and I distinctly remember listening to Homo Deus near the start of COVID and getting to the section where he said “Based on current trends, the risk keeps increasing of a zoonotic virus causing a global pandemic and societal shutdowns.”)

I must say though: the most important thing I noticed on second listen of the cases was the reference to my hometown in the Ample Hills case. The “Ooey Gooey Butter Cake” flavor is almost certainly based on Gooey Butter Cake, a St. Louis specialty.

4 Likes
12

In the article @cedric says

The core idea is ridiculously simple.

  1. Experts can see things that novices can’t.
  2. If you can teach novices to see the domain like the experts, you will accelerate expertise.

I’m starting to see flashes of the sense making concept of how experts see things differently. At work they asked everyone to come up with ideas on AI automation and most of them were some form of a chatbot on top of a curated knowledge base.

Meanwhile I looked at it and immediately thought about data systems (pipelines, schemas, databases) and how they all come together to create something far more useful.

It made me realize how even with a tool like Claude Code the ideas and questions you feed it are limited to what you know or have experience with. Even if you somehow get an expert answer or a full blown system design, you still won’t know how to evaluate it, you won’t know what it takes to build it and operate it properly.

I experienced this when I was building software with agents and I was like look at Python code that seemed correct and worked perfectly while my manager, a much more experienced developer, would look at it and see all kinds of problems or suggest better ways to do it.

5 Likes
13

Oh, yet another example of this learning approach in the wild. This one is thanks to @TylerD, who recommended this interview with legendary investor Reese Duca (all bold emphasis mine):

And so I come to UCSB thinking I’m going to be an engineer. And then I take an economics class from a professor, this fella by the name of Herb Kay. And he had done his PhD at Stanford in economics. And the PhD was about companies and business models that had started during the Depression and had been able to survive and thrive in the Depression, World War Two, and then the Consumer 50s. Because the timeframe he was looking at was 1930 to 1960. And what were the business models, and what were the attributes?

And later in the term, this is after class, he said, Reece, I received a new grant. I’m gonna hire a new research assistant. I can tell you’re engaged in the class. You read everything, you ask the right questions. And if you’re interested, it’s yours. Tell me before tomorrow morning at 6:45, but it’s yours. And so I thought about it for about 30 seconds and I said I’d do it. So I ended up being his research assistant, his reader, his grader, and his personal investment researcher for the next three and a half years, which was like one of the most incredible gifts any student could ever have.

What I say to people who talk about Herb Kay as being an extraordinary economics professor: What he was was an extraordinary investor masquerading for about 10 years of his career as an economics professor. And what he had done to develop his investment strategy. And his investment strategy was focused solely on public investing. He didn’t invest in private companies. But the most important interaction I had with him was on S-1s. And he would, every two weeks or so, he’d give me two or three S-1s. And he’d tell me, Look at the following three things. Read it from cover to cover and come back with the right question. We’re gonna get on the phone with the investment banker, we’re gonna get on the phone with the CEO, we’re going to talk to a couple of customers. And so I’d go back to my dorm room, or my apartment room, and I’d spend time reading these S-1s and I’d come back, and half the time he would reprimand me for missing this and Why didn’t you think about this, and I’d leave a little bit with my tail between my legs, but I realized my learning curve was going through the roof as it relates to my role as an investor.

When I finished high school, I had $2,000. I had saved this money, I had worked in a butcher shop, and I’d done a teeny bit of investing. And so I arrived at UCSB with $2,000. I spent the three and a half years with Herb Kay, I leave UCSB, I have $7,000. So after UCSB, I applied to business school, I applied to a couple business schools, got into all the business schools, and decided to go to Stanford.

4 Likes