A Working Theory

In order to get good at anything, one must accept being bad at it for a while. Being bad at things, but still doing them to learn and improve is an essential skill. It is skill that I now try very hard to cultivate in myself. To get meta, I used to be very bad at being bad. Being incompetent would make me unhappy and be very discouraging. When I viewed my self worth as a function of my ability, it was devastating to have no idea what I was doing. By the time I wrote this, I had successfully divorced my self identity from the outputs of life.

My self identity realigned around life’s inputs. Learning became an essential input. In addition, I saw myself progress from being bad to functional at a few skills. That was encouraging. School, internships, family, friends, works, etc. force us to be bad at things, but still do them. After a while, I saw how much growth was possible. Instead of thinking of myself as bad, I thought of myself as a beginner. Being a beginner means a very high rate of learning.

Learning  is a core value for me. My goal is to make sure I learn something new every day. What attracted me to data science was the vast, ever growing, amount of information in the field. Every night, after work, there is always something new to learn about data science. Cooking (my favorite hobby) also offers me the opportunity to constantly learn new techniques.

I’ve struggled with severe should pain the last 18 months. After trying everything else, I gave yoga a shot. There are few more uncomfortable experience than an inflexible person trying yoga for the first time in NOLA heat without a water bottle. It bruised my pride, but helped my shoulders. I’ve stuck with it for the last month. My shoulders are the best they’ve ever been since the pain started. My posture and ability to hold yoga poses have improved. Even so, I still fall pretty often.

I’m not bad at yoga; I’m just a beginner.

“It can do anything”, is rarely something knowledgable people say about a tool. This is especially true of quantitative tools. The people who best understand a tool, also understand its limits. That doesn’t mean they are bad at using a tool. Often, they know what other skills will complement the limits of the tool. They can then master that complimentary skill.

For example, Bill Gurley dissects the Discounted Cash Flow (DCF) model used in company valuation here. It might be surprising to see an investor speak critically of the workhorse tool of company valuation. In my opinion, that’s what makes him a good investor. He understands that a DCF is a mathematical framework of connecting inputs to outputs. However, the really tricky part is figuring out the inputs and the huge amounts of uncertainty around them.

Another great Bill Gurley example is this post on the Lifetime Value (LTV) model. He explains how to calculate LTV, but to be very careful about the assumptions that go into it. It can very easily lead companies to spend a lot in acquisition costs that never gives them an ROI.

An example from a different person, is this transcript from Charlie Munger. He says that everyone in business should understand accounting. He goes on to say,  “But you have to know enough about it to understand its limitations—because although accounting is the starting place, it’s only a crude approximation.” He also says everyone should understand the limits and quirks of human cognition. He recommends studying this in order to understand motivation, marketing and to lower the chance of being manipulated.

Every Data Scientist has read through some study or paper that grossly abuses Linear Regression. If you don’t know the assumptions being made by the model, you can reach ridiculous conclusions. It’s actually very rare in business to have a simple linear model that meets all of the assumptions that can both explain and predict the underlying system. Of course, it can still be a useful tool as part of an ensemble or for subtasks.

Whenever something is being quantified, there will always be a loss of nuance and context. Of course, that doesn’t mean that quantifying the unquantifiable is bad. In fact, it’s usually necessary to reduce the complexity of the problem. However, it is important to understand what information is lost. Those who respect what was lost also usually know what was retained. The people who push the border of a tool forward usually know where the border was before they moved it.

Whenever I see someone do something that looks difficult or near impossible, I always wonder what their emotions are like. When Barack Obama gives a speech seen by millions or when an NBA player shoots a buzzer beating shot in a playoff game. What are they thinking? How are they not crushed by the weight of the moment? Are they unaware of the pressure or do they not feel fear?

Well, I can’t answer how they feel. However, I can reflect on my own emotions as they relate to difficult situations. The first time I did public speaking, it was stressful. So stressful, I wondered if I was too stressed to perform well. After the first few times, I built confidence and was able to perform. Sometimes better than others, but I rarely failed because my emotions overcame me.

I’ve never stopped feeling some fear before public speaking. It is part of the cadence. The fear never proves founded. It’s just an emotion. Over time, it fades into the background. My experience having that emotion and not having it be predictive has put it into context for me.

I was recently talking with someone who’s job has become much more stressful the last few months. He said that he’s just adjusted to swimming in warmer waters. Emotions are like water temperature. As long as it’s not lethal, you’ll eventually get used to swimming in it.

When Breath Becomes Air is simply one of the best books I’ve ever read. {spoiler alert}

I’ve read a number of fantastic books about life, death, mortality and human’s search for meaning. Some of them make great points and offer compelling insights. However, they are all written for an audience. They are all bound by the need for logical consistency.

When Breath Becomes Air is different. It is written by someone in a mad rush as his death creeps closer. Paul acknowledges his contradictory and emotional reactions to being diagnosed with terminal cancer. He explores his own mortality while staring it in the face. His book does achieve great fame and touches many people. However, he does not know that’s how it turned out. He didn’t even know it got published. He writes for himself, his family and daughter. How often can you read a book that was so fulfilling to it’s author that they spent their dying breaths on it?

Interestingly, Paul doesn’t explicitly acknowledge he chose to write the book with his remaining days. His wife (Lucy ) explains this in the epilogue. Instead, Paul just examines life and mortality. The reader doesn’t have to bear the weight of knowing this book offered Paul meaning in life. The reader just feels like they are in his head. He reflects painfully articulately on his life, his unfulfilled dreams, his physical decline, his memories of patients deaths and his own mortality. Like most books about death, this one made me consider life.

Thank you to Brittany Gorevic for recommending When Breath Becomes Air to me.

Monopoly is one of my favorite games. It is just the right mixture of probability, deal making, strategy and numbers. Monopoly is a game of working with luck and other people.

My key philosophy is: Do things that maximize your chance of winning. Most people play to avoid losing. However, nobody cares about 2nd place. With 4 players, everyone starts the game with only a 25% chance of winning. Playing risky increases your chance of winning (which you care about) and your chance of losing early (which you don’t care about). Below are some tactics that I have found maximize a player’s chance of winning. Of course, many also increase the chance of losing quickly, but that doesn’t matter.

Tactic 1: Get as much real estate as possible. The first stage of the game is all about acquiring a monopoly while stopping other people from doing so. If you can’t get a Monopoly, get real estate that can be traded for a Monopoly. Trading helps the two people who made a trade. Trade until you get a Monopoly and don’t encourage others to trade. Early in the game, people overweight cash for real estate. People also feel more comfortable trading a Monopoly to you when you have no cash; another advantage of being real estate rich & (see next tactic) cash poor.

Tactic 2: Mortgage and build aggressively. Rent doesn’t increase your chance of winning so ignore it. You should have a lot of real estate from tactic 1. Once you have a Monopoly, mortgage properties to build on your Monopoly. Build in a concentrated way on one or two Monopolies. Investing until you’re broke in a Monopoly slightly increases your chance of losing, but it dramatically increases your chance of winning. I like to focus on a cheaper monopoly that you can build up quickly.

Tactic 3: Create a cash cow (a cheaper monopoly that is fully developed), and build towards a knock out punch. In the middle of the game your properties should be generating cash for you to reinvest. If you don’t have an “expensive” monopoly, you can should try to get one. Use the cash from your cheaper monopoly to help you acquire an expensive monopoly. With enough investment, these eventually become knock out punches. The cash generators can be low cost properties with hotels or higher cost properties with a few houses. Only mid to high cost properties can be knock out punches. If you have a cash generator, but not a knock out punch, consider giving someone a free pass in exchange for a card that give you a knockout punch monopoly. (similar strategy to scaling a startup)

Tactic 4: People will be terrified of your Monopoly. To avoid losing if you’re poor from your reinvestments, trade free passes with people that have Monopolies. Again, trades help both traders and implicitly hurt both other people. Trade often.

Tactic 5: The red & orange are both the best for their sides of the board. However, it’s more important to get a Monopoly first than it is to get a specific Monopoly.

Tactic 6: 7 is the most likely number to be rolled (~17% chance), followed by 6 and 8 (~13% each). When someone is 7 spots away from something you own and either 6 or 8 away from another, build. There is a 30% chance they hit you and you get an immediate return while building towards a knockout punch.

This is a “Science of 1” post. I have experimented a lot with my caffeine consumption. Until the age of 20, I never drank coffee. When I studied abroad in Spain, I started drinking coffee for the cultural elements of it. After getting a summer internship,  I started drinking coffee daily to help me focus / wake up in the morning. In the beginning, coffee had a very large positive effect on my ability to concentrate. Eventually, I found myself drinking 2+ cups in a day with a very marginal impact. I wasn’t sleeping as well though.

I began working full time almost 2 years ago and decided it was time for me to find the right coffee drinking pattern for me.  There was an article that cited a John Hopkins study that rang true. It claimed that most of the positive impact of caffeine was withdrawal being relieved. The theory was that if you controlled for overall caffeine consumption, coffee did not actually improve any cognitive functions or mood. Basically, daily coffee drinking lowered people’s baseline mood and cognitive function and then restored it when people drank it. I could not find the cited study. However, I wasn’t looking for peer reviewed proof. In fact, something could credibly pass science’s tests for the general population, but still fail to help me. Instead, I experimented with different caffeine habits on myself. A few things quickly emerged:

1. Drinking coffee after lunch makes it harder for me to fall asleep.
2. Each day, I’d need to drink a little more coffee than I did the previous day to get the same effect.
3. If I went a little while without drinking coffee, it would once again have a strong impact.
4. I drank black coffee long enough to enjoy the taste and never considered milk or sugar an option.
5. I hate feeling like I “need” to drink coffee to just wake up.

Through some experiments, I settled on four days as the maximum number of days per week I can drink coffee. At four days per week, I feel a noticeable boost on days that I drink coffee. On days, I don’t drink coffee, I feel fine and don’t particularly miss it. The easiest days for me to skip is Saturday, Sunday and Monday. Like most habits, it was hard to do it at first, but now feels like second nature. When I do drink coffee, I drink it early in the day to minimize its impact on my ability to sleep. Sometimes, I wait an hour or two after I get in to drink coffee. That hour is reminder that I don’t need coffee to function. It’s just a boost.

One reason evolution is amazing is that incorporates knowledge into our species. However, it is slow and a lot of people live suboptimal lives waiting for our biology to adapt. Fortunately, humans, unlike other species can adapt to the world quicker than evolution through science. We can observe the world and learn from our environment consciencely instead of waiting to see what confers a survival advantage.

This is great. Science is great for pushing humanity forward quicker than evolution would. However, we usually require that science is reproducible across the population. We are also trying to understand the average impact something has on people. However, as a single person, the average doesn’t offer any guarantees for me. In fact, it can be a burdensome constraint. I’ve decided to experiment on myself to make decisions for myself. Things that are measurable, happen frequently and have a low cost for being wrong work the best.

For example, diet is constantly analyzed ad naseum by the science (and psuedo science) community. I’ve read so many different studies on what diet is best. There was a temptation to look for statistically significant, reproducible studies on what diet produces the most day to day energy. Instead, I just experimented with a bunch and picked the one that makes me feel the best. Of course, that doesn’t mean that my decision generalizes to the whole population. However, that’s the point. It doesn’t have to generalize. In fact, there’s no reason to think that the one diet is best for every one. I also don’t need the optimal diet. I am willing to satisfice. My diet only impacts me. I’d rather have a pretty good decision that I am sure works for me than a possibly great decision for the general population.

Experimenting on myself solves a number of problems:

1. I don’t have to worry about reproducibility. No need for me to sort through conflicting studies or to worry about convincing other people.
2. I am confident whatever I am doing will actually help me achieve my goal. High confidence in outcome helps my willpower and follow through.
3. If I am likely different than the general population, this will account for that.
4. No analysis paralysis.

This “science of 1” theory only works in certain cases. Things that carry a high cost or occur rarely do not work and I rely on general science. For example, I can test which diets make me feel the best day to day, but I can’t test which diets are healthiest over the long term. Generally, this works best for day to day impact and not long term impact. However, it has been very helpful and freeing to experiment on myself for myself.

Also, a variation of the Slow Carb Diet makes me feel the best day to day. Of course, that proves nothing about the long term impact of the diet. However, there’s some solid science to back up the idea that highly processed foods and sugar rich diets are not good for you. Given that, it’s very unlikely that I’m making a worse choice than the default american diet for my long term health. However, I am nearly certain that short term, this diet makes me feel much better.

Life is a series of events that we go through. Often, we try very hard to control what those events will be. Or at least we try to make sure certain positive events are part of the list. Those positive events could include having a company we work for IPO, winning a prize, getting married, buying a boat, etc.

In order to maximize the chance of us experiencing one positive event, we often willingly subject ourselves to many negative ones. This could mean studying for a test in order graduate college, working long hours to get a promotion. Of course, the terms positive and negative are subjective. However, we can all substitute some positive and negative events into this equation based on our own definitions.

How do relationships with other people fit into our calculus on experiencing positive events? This gets complicated, but our relationship with various people usually pulls us through many of the positive and negative events we experience. A family member could get very sick and need a tremendous amount of emotional and financial support.  A childhood friend could Vinny Chase and pull his friend group along for fame, fun and business deals with Mark Cuban.

How are we supposed to navigate the complicated, connected maze of people, circumstances and the pursuit of happiness*? Well, I have no idea how to answer this question generally. However, nothing has been more rewarding to me than sticking with people regardless of the circumstances we pull each other through. Sometimes those have been very happy, sometimes they have been very difficult. However, picking people and sticking with them regardless of their present circumstances has been one of the most meaningful things I’ve discovered so far in life. In many ways, that is the promise of family: sticking with people whatever the cost. With friends, you can pick them; with family, besides a spouse, luck assigns them; you pick some coworkers and get others from luck. Not sure how to sort this out, but choosing the right people and then being loyal to them regardless of the cost seems to be the most meaningful thing I have found in life, so far.

It is hard to know, but I suspect that over the decades this will become a cornerstone of my life’s philosophy as I refine what it means.

*whatever the hell that means

Sometimes, it feels like each day is roughly equal in it’s impact on the entirety of life. Like a game of basketball, sometimes you are up or down. Sometimes by a little or a lot. However, the pace is relatively stable. The score only changes by 1-4 points per play. Even giving up the maximum four point play can be erased in two possessions.

However, is this an accurate reflections of life? Is the impact of each moment roughly equivalent? Perhaps, life is more like soccer. Occasionally you score a goal. Occasionally someone scores on you. A 90+ minute game hinges on a few moments. In a low scoring game, a single play could determine a win or a loss.

Which is a better metaphor for life? Looking at bibliographies of strangers, most people’s lives are reduced to the best and worst thing that the person did. “Well he was a great senator and brought us jobs, but he also got caught up in that scandal.” or “Sure Woody Allen made great movies, but there was that whole pseudo-stepdaughter relationship.” Rarely is the day to day mentioned. Did he order the lobster when someone else ordered a burger at lunch knowing they would split the bill? Did he remember people’s names? Did he have road rage? i.e. the things that fill most of our days and judgements of other people.

Of course, how people typically judge a stranger’s life is not necessarily correct or important. Nor is it the point. It is simply an example of an activity where the average doesn’t matter. The only thing that matters is the minimum and maximum. Summer interns are usually remembered similarly. Did they do something awesome and did they screw up anything badly?

Venture capitalists are usually judged (financially) by their best investment. Their batting average doesn’t matter (at least for early stage VC’s). A private equity investor is trying to avoid an investment that’s a total loss because they will never shake that black eye. They are focused on avoiding a worst case scenario.

There seem to be quite a few activities that have these properties of irrelevant averages. The tails are the story.

Most of life feels like a game of averages. It seems to be the way that our brains are wired. However, a single youtube video can now garner 100 million views and launch a career (The Chainsmokers). We like to believe that consistent effort and hard work lead to a fair reward. In many ways, reducing a lifetime of events into two moments, seems unfair. Worse, if we follow the soccer analogy, goals can happen at any moment. Constant vigilance is required to not miss a setup for a maximum life moment.

Perhaps, it’s worth deciding if an activity will be best-worst. If it is, logic dictates that we ignore the middle, avoid catastrophes and try to exploit the smallest chances of a breakout win. Of course, life can be a lot more pleasant by shifting some best-worst activities to a steady, more controllable focus. For example, defining life happiness by your reputation to strangers is extremely fragile. Any moment can ruin you. Defining your happiness internally is not subject the variation of a best-worst process.

On the other hand, sometimes a best-worst view can be freeing. Most of life’s decisions will not dramatically affect the rest of life. Overpaying for dinner, getting stuck in a broken subway car will not matter. Meeting someone you marry will be a huge deal for many people. Getting a small increase in salary won’t really matter. Breaking into an industry you love will matter. So would being convicted of a serious crime.

To use a poker analogy, if the ante and pot size for most of life is really small, ignore it. Just make sure you win the big hands if they are 10,000x more important than the typical hand. In fact, use the small hands as a way to position yourself to win. Accept a lot of small hand losses to get information about how other people play. Perhaps lay a trap by establishing yourself as a person who recklessly bluffs by losing some of the small stakes hands. Then when you have a huge hand, people might think you are bluffing and you can win enough to just leave the game thereby remove the chance of losing the huge winnings.

When I look back on any activity that I’ve done, I always look for ways I could do it better. Usually the first question is: did I do a good job on that activity or do I need to completely rethink my approach?

The biggest clue I get is the outcome of the activity. For example, let’s say I tried to predict a company’s profit for a quarter. At the end of the quarter, I get to compare my prediction to what actually happened. Let’s say that my prediction for profit was correct. It would be tempting to just say that my system worked and requires no tinkering for predicting the next quarter’s profit.

However, let’s say that I broke down my prediction and it turns out that I overestimated revenue and overestimated costs. The net result was that my prediction was correct about profit. However, it was only correct because I made two mistakes that happened to cancel each other out.

The above example is trivial and easily quantifiable. However, offsetting mistakes are a constant threat. It makes it hard to trust all results that I observe. Let’s say that I am running late for an important meeting (mistake 1) and to make it worse, I don’t even tell the person that I am running late (mistake 2). That person happens to be even later to the meeting than me and never realizes that I was late. Since, they never knew I was running late, it’s almost like I made no mistakes. My two mistakes would have offset each other.

In software engineering, sometimes it is tempting to only test the input and final output for a piece of software. One of the big dangers is that offsetting mistakes can hide between input and the correct outputs. Instead it is better, though more tedious, to write unit tests that for each part of the software.

I try to approach my life  with the same lens. Just because something had a positive outcome, doesn’t mean my preceding actions were “correct”.  They could have been wrong in offsetting ways. I try to break my actions into small pieces and evaluate each in isolation, but that is far from a perfect strategy. Awareness of this issue definitely forces me to be more reflective, seek more advice and question whether I can do things better.