Author: MIKE

Are you tired? Is that really the question?

MIKELeave a comment

“When faced with a difficult question, we often answer an easier one instead, usually without noticing the substitution.”

Daniel Kahneman

We probably do this a lot more than we think. Consider how many decisions we face in a day, how many are automatic, and how many switches go unnoticed.

It happens to everyone. In a recent post, Scott Alexander writes about an individual randomized control trial for a sleep supplement. Alexander wanted to know if the supplement helped him sleep better, or if it was a placebo effect.

Placebos are powerful and cheap. If they get the job done, all the better it seems. But Alexander is serious, smart, and curious—so he tested this idea.

A friend disguised the supplements and sugar pills in oversized capsules, flipped a coin to determine the order, and placed the camouflaged pills in a monthly pill planner. To test the effectiveness, Alexander recorded hours of sleep and subjective ratings on how he felt. The results shocked him.

But not for the reason he expected.

I think the active ingredient here was not letting myself look at the clock. Without external cues to tell me how tired I should feel, I was forced to rely on how tired I actually felt, which in many cases was “not tired at all”. 

Scott Alexander

In removing the alarm clock, Alexander removed the easier question. If it was before nine, it was too early. Obviously. Or not.

These kinds of tendencies work great for a lot of things, but occasionally work too much and we get interesting results like this.

Join 999 other subscribers

Photo by Aphiwat chuangchoem from Pexels

The Psychology of Rotten Tomatoes

MIKELeave a comment

In an article for Wired, @SVZWood writes about Rotten Tomatoes, the website one in five Americans use to inform their movie choices. Wood’s article showcases three helpful ideas:

Simple algorithms. Rotten Tomatoes has become a successful business and top website based on a simple algorithm—counting. Two movie curators read reviews each day, judge them as positive or not, and then count the totals. If a movie has more than sixty percent positive, it’s marked as fresh.

Numerical authority. There’s something about people, numbers, authority, and precision we just can’t get over. Wood captures this, writing, “There is an authoritative allure in the site’s numerical scores….(people) reflexively—and nonsensically—trust a fresh sixty percent Tomatometer over a Rotten 59%.”

In an age of big data and algorithms and random forests, it’s helpful to keep in mind that simple systems work. One example, football.

Numbers feel secure. They’re a rationalization blanket in a world of unknown things under the bed.

Jobs-to-be-done. Reviewers don’t like the binary nature of fresh or not. Wood notes that, “there is no Underripe or Overripe tomato.” In the terms of Bob Moesta, this is the difference between a supply-side orientation and demand-side orientation.

In JBTD theory, the goal is to find what customers want to do that rather than what a supplier wants or thinks the customer wants. Reviewers (supply) want to share their nuanced take rather than a 👍 or 👎 via their Twitter account. The customers (demand) just want to know if a movie is bad, potentially good, or very good. This plays out in some interesting ways that Rotten Tomatoes has plucked.

For example, a review which is glowingly positive and a review which is slightly positive are both coded as positive. This leads to extreme scores. This is a feature, not a bug. It’s the information people want to know.

Join 999 other subscribers

Morey thinking like Marks

MIKE1 Comment

It looks like Daryl Morey is thinking like Howard Marks. But, he probably always has.

On the Wharton Moneyball podcast, Eric Bradlow talked through why the Houston Rockets might think playing “five guys a large number of minutes that are all six-seven to six-four” to end the season and into the playoffs was a good idea.

At first this seems like a mistake.

Basketball has defined positions. Point guard, shooting guard, small forward, power forward, center. We’ve always done it this way.

Oops.

That expression is a red flag. Instead we should think from first principles, which is what Bradlow and co-hosts Adi Wyner and Shane Jensen do. If the Rockets play shorter players that don’t fit the traditional mold, what will happen. Bradlow summarized:

“Given they’re shooting threes and the higher variance of the rebounds, a lot more balls hit the floor because it’s not just dropping straight down so you might want faster players who can get to the ball quickly.” (though maybe not)

“And now you’re getting the other team to possible take less threes. That’s a good thing about the Rockets.” (by exploiting size mismatches down low)

“The other thing you might argue is that the other team is going to crash the boards and that might create more fast break opportunities.”

Wharton Moneyball podcast 2/12/2020

Small ball strategy might be something smart coaches pursue. Zach Lowe wrote that Toronto’s Nick Nurse will play weird defense, and will sometimes tempt the other team “into inefficient one-on-one battles.”

Two principles Marks hammers again and again is the importance of being right and different and the importance of second level thinking. Though it’s really no surprise that Morey and Marks take different paths to make similar results in unique domains.

Danger Word Search

MIKELeave a comment

My kids love Ellen’s Game of Games.

My kids love the mini-game Danger Word. The premise is for one player to get their teammate to guess a word. The wrinkle is that the winning word (sofa, ocean, etc.) is similar to the danger word (couch, sea, so-on). If a guesser says the danger word, they get smothered and covered. If a guessers says the winning word, the other team gets hit with it.

It’s an interesting game because the best clues have multiple levels. They give information about both the danger word and the desired word. One winning moment was when a woman gave the clue ‘otter’ and her partner guessed ‘sea’. It was a winning clue that also avoided the danger word of ocean.

Knowing something is this, and not that, will become more important as things move from single-issue-tangible to many-copies-digital.

  • YouTube is an amazing resource but the search options are mostly by name. The most challenging queries are for people who share names with pastors. One side effect from weekly sermons is regular content and what better way to reach the flock than via YouTube.
  • Websites used to have a .gov, .org, .com structure that hinted something was this and not that. In some recent research I found a dot-org domain that looked pretty legitimate. However, going to the About page showed that it wasn’t what it appeared to be.
  • Visual evidence used to be clear. Screens though are shaded windows. Rather than asking is something manipulated, the default is now to ask how something is manipulated.

In Ellen’s game, the goal is to give a precise clue.
In YouTube searches, the goal is to search a precise query.
In website scours, the goal is to find truthful facts.

As more information goes online there will be more, not less, statements about being at war with EastAsia. The skill we’ll all need, my daughters included, is separating one category from another. Danger words from winning words, helpful queries from unhelpful ones, real from fantasy.

Made Up Start Up: Upcoming Streaming

MIKELeave a comment

Movies are awesome. The next time you hear someone regret remakes, remixes, and redundancy please remind them to respect your review. No one complains about meatloaf, sequels are comfort food.

However, I don’t always remember who is who, who is alive, or who is in love. Here’s the pitch: a streaming service that offers a limited run of a limited number of movies based on the time of year.

Wharton professor Jonah Berger researched how ideas spread and part-of-the-reason ideas spread is timing. People are contextual operators. We need cues for action. How many times have you seen a trailer, heard about a restaurant, or found out about some music only to never followup. You need a nudge.

The Upcoming Streaming businesses would be just that. The week before Harry Potter X comes out, the last two movies from that universe are available to stream. The Rock has another Jumanji-like movie coming out, and so another Rock movie is available to stream. Actually, the Rock always has movies coming out. He’s a mainstay.

For movie studios this is a no-brainer. Some studios spend more marketing a movie than making one. This streaming service would be inverted marketing. People pay to get excited for the next movie. Free money.

There’s hope for this idea as smart contracts become more popular. Tyler Cowen notes that Hollywood and whaling were the VC OG. If Silcon Valley is into crypto and blockchains then maybe the theater business isn’t far behind.

Another advantage is that this business doesn’t interfere with the competitive advantage of existing streaming services. If a movie is offered in more than one location, few customers would cancel Netflix just because a movie or two each year were available somewhere else.

But wait, there’s more! This business has scarcity. If movies are only available for the week before a release people will feel the need to act now. It can’t hurt having psychological influences pushing this idea along.

Upcoming releases wouldn’t be the only available source of inspiration. How many studios would love to pump their old rom-com libraries for one more run as they return to our lives as a special Valentines week lineup? How great would it be to stream only the Thanksgiving episodes of Friends over the Thanksgiving holiday?

If someone knows Michael Ovitz, have him get in touch.

Tim Harford’s Arguing Advice

MIKE2 Comments

Tim Harford joined Tyler Cowen to speak about his career, production function, and why dice are something he couldn’t live without. Harford said this about debates:

“I still think debating is underrated. People think that it’s a very elite thing, practiced by elite people from very posh schools and very privileged backgrounds and favors a certain kind of education. Those things are all true. At the same time, in a debate you are protected by certain rules, given space and time, and people are not allowed to interupt you except in certain formalized ways.

“Once I became an adult, and entered certain corporate spaces and corporate meetings, I became aware that all the old men were talking all over the young women. They’d interupt and wouldn’t let these younger people get a word in edgewise. I realized that debating protects anyone.”

Tim Harford

A contest has explicit rules. A culture has implicit rules.

For organizations to ‘argue Zell‘, they must have leaders willing to be challenged and leaders who communicate that. Career risk depends on the culture.

Like strategy, marketing, or ethics, culture is something organziations do. “Unless you set it, it’ll just be what it is.” The easiest way to create a specific culture is to hire well. After that, incentives matter.

Solve for the Equilibrium: Marathon Lottery

MIKE2 Comments

When supply equals demand—in economic models, not Alchemy—the price is calculated, equilibrium is achieved, and angels sing. What economists do we aren’t sure. Rarely does this work because value is subjective. Our relativistic nature is seen when it’s less grating for a CEO to make 100 times their employees than for a peer to make double.

However solving for the equilibrium is a helpful. This Phantom Cowen would lead us to ask about how we might design a system people can’t buy, favor, or cheat their way into? How do we fix the game and avoid equilibirum? It’s been done with Jeopardy and horse racing.

Enter the NYC marathon.

The NYC marathon is super popular. The race welcomes 53,000 people, the most in the world, but demand is triple the supply. If this were a business, the organizers could raise prices. Pricing power! But profits aren’t the goal.

“At New York Road Runners, it is our goal to give everyone on the planet both a reason to run and the means and opportunity to keep running and never stop.”

NYRR Mission

Rather than raise prices, the NYRR had to solve for the equilibrium another way. In Planet Money episode #962, host Kenny Malone explains how the running club allocates their scarce resources. How they solved for the equilibrium.

What’s great about this example—beyond the process—is the creative ways NYRR solved their problem. They’ve added value without addressing the race.

Escaping College and Commodity Competition

MIKE1 Comment

When one product is similar to another there is competition. Businesses which compete on price lead to low-cost-high-volume winners. Thankfully, being dissimilar isn’t difficult. Many brands (Advil, Coca-Cola, Harvard) differentiate commodities by framing the context.

Differentiation avoids the market mechanism. But that doesn’t mean things will be easy. Being the low-cost provider or being different both take extreme amounts of work. That said, jobs-to-be-done thinking may lead to a slightly easier slog

The JTBD idea is updated on Twitter, but Michael Horn’s TAG brought up an easy idea for escaping competition. Instead of asking who are your customers, ask what do customers want?

Typically businesses count things. This is easy data. Plus, the more math the less career risk: a bit of math in itself.

Through their book, Choosing College, Horn and Bob Moesta encourage people to think about new categories. To count less and consider more. Get past demographics (high school graduate) and to demands (level-up).

What is the demographic profile of someone who eats at Panera Bread? That’s a hard question. However, what’s the JTBD of someone who eats at Panera Bread? That’s easier. Ron Shaich explained that Panera is “the kind of space where you want to sit and do an interview, a place for a bible study group, for a team meeting.” 

People hire Panera for group space and bagel spreads. That’s the JTBD. That’s the differentiation.

Being different is easier—though not easy—because it offers many directions. If a business competes on price, that’s it. If a business finds a new JTBD they might have that market all to themselves, temporarily.

Is Jack Reacher a Bayesian?

MIKE1 Comment

We’ve covered why Jack Reacher would be a good financial advisor. In A Wanted Man, Reacher once again proves his mettle. To set the scene, we’ve got two bad guys (we think) on the run from our two heroes (we think). It’s rural Iowa and Reacher and Hero Two need to figure out which way the troublemakers fled.

Reacher, a man of action, says they have to act fast.

“Now? They could be in a million different places.”

“So it’s time to gamble. Before I get taken off this thing. Or supervised. One or the other is sure to happen first thing in the morning. That’s what maximum effort means. So suppose the two guys are still on the road?”

“But which road? There are a million roads.”

“Suppose they stayed on the Interstate?”

“Would they?”

“They’re probably not local. They’re probably running home right now, which could be a big distance.”

“In which direction?”

“Either one.”

“You said they might be traveling separately.”

“It’s a possibility, but a small one. Statistics show most paired perpetrators stick together after the commission of a serious crime. Human nature. They don’t necessarily trust each other to deal with the aftermath.”

“Statistics?”

“We find them to be a useful guide.”

Part-of-what-makes the Reacher books so popular is his fists the size of supermarket chickens. Part is his modern day sleuthing. Reacher is curious and he notices things. If he took the stairs at 22B Baker Street—two at a time, of course—he’d know how many there were. Part of the reason is Reacher’s knowledge of base rates.

Julia Galef explained Bayesian thinking this ways: “It’s important to pay attention to these snowflakes of evidence that maybe weigh very little on their own but if you pay attention to them and notice how they accumulate, their collective weight becomes enough to break a tree branch.”

Median and Average Meanings

MIKE4 Comments

There’s a story about Bill Gates’s wealth and height you’ve probably heard. If not, let’s quickly share it here. If Gates were to walk into a room of you and twenty-five friends, the difference in average height and median height would be small.

However, the difference in average income and median income would be large. Gates’s wealth raises the mean because it’s relatively ginormous. Even in a room of millionaires, Gate’s presence changes the average from one-million to four-billion. That’s a lot.

This story is helpful to keep in mind because averages hide nuances.

In 2016, the average student loan debt was $37,000. However, the median figure was $17,000 and one-fourth of all borrowers owed less than $7,000. Part-of-the-reason the average is so far from the median is because of graduate school, one-in-four post-graduation debt-holders had more than one-hundred thousand dollars in debt. My wife attended medical school, and I can attest to the amount.

This median approach might paint a rosier picture.

Retirement accounts show the same point, only in the opposite direction. The average balances is $100K whereas the median is $24K.

“Often an average is such an oversimplification that it is worse than useless.”

How to Lie…

Average can become an If/Then word. If the average is presented, then we can attempt to find the median as well. Bill Gates will be our patron saint here. Imagine him, staning with a Diet Coke in hand, reminding you to find the median too.

Another fun part of you and twenty-five friends is “the birthday bet”.