#39 Data-Driven? Probably Not.

Most self-proclaimed "data-driven" people I've met are... not.

[full tweet here]

The second foundational principle inside the Guardian Academy is "Frequency of Exposure"


Because The Guardian Academy is built on a foundation of:

- Knowing yourself
- Appreciating when bad things don't happen
- Getting better results in life

One of the biggest risks to all three of those things is not understanding "frequency of exposure."

Here is a short video I made for TGA explaining it:

[6 Word Updates]

If you think you can control your emotions, think that some people also believe that they can control their heartbeat or hair growth.
-Nassim Nicholas Taleb

We tend to believe that we have more control over our emotions than we do. We also tend to act on momentary feelings and justify our actions with data to hide our impulsivity.

Data And “Data-Driven” People

Hot take/observation:

The more exposure to data the worse the decision.

It doesn't make sense when you hear it out loud, but it makes perfect sense if you pay attention.

Humans are impulsive and emotional.

They're also smart.

Dangerous combo. Emotions lead to impulsive decisions which are then justified with data. I'm not convinced that justifying emotional decisions is the best use of data.

What Is The Purpose Of Data?

To answer this question, let's go back to our core concepts. One of them is:

We only take asymmetric bets (to the upside).

Couple that with the concept of repeat exposure (probability):

  • We aren’t going to bet a dollar to win a dollar (symmetric). There is no reason to take that bet. If we made this bet an infinite number of times we would end up in the same place.

  • We also aren’t going to bet a dollar to win fifty cents (asymmetric to the downside). If we made this bet an infinite number of times our losses would grow, on average, with each bet.

  • If a bet has a higher expected value than the cost of making the best we may take it. If we were to take it an infinite number of times, our total gains would grow over time.

All decisions we make are bets. We always want to make sure that we are taking bets where 1) there is asymmetry and 2) it is to the upside

This is what we’d refer to in TGA as “Engineering luck”.

Thus, exposure to data must have more upside than downside. In other words, we expose ourselves to data when the upside of the exposure is greater than the downside of the exposure.

There is only one scenario where the upside to seeing data is greater than the downside:

When it informs us of how we can change our behavior to get better results

That's it.

In our world:

The Only Purpose Of Data Is To Inform Behavior

The question to answer is:

Am I willing to change or modify by behavior based on what the data says?

If not, there is no upside. If there is no upside, there is definitely no asymmetry to the upside. In this scenario, we can only incur an emotional loss.

Here is an example.

You think to yourself:

“I wonder how much I weigh?”

First, we would ask:

If you don't like the number you see, are you willing or able to make changes to your lifestyle right now?

If the answer is no, nothing positive can come out of stepping on the scale. Because right now, nothing can change.

There is, however, a chance that you don't like what you see, which would be a negative emotion.

  • No upside.

  • Potential downside.

Stupid bet to take, in my opinion.

Another example:

Imagine you are in this investment: 15% return with 10% volatility per annum (From Nicholas Nassim Taleb).

If you checked it yearly: 93% of the time you’ll have a positive exposure.

Monthly: 67% probability of positive exposure.

Daily: 54% probability of positive exposure.

Minute: 50.17% chance of positive exposure.

If you are viewing on the minute time frame, half the time it’s up, half the time it’s down. The problem with this is that a loss hurts more than an equivalent gain feels good. Even at break-even on paper, you have incurred a massive emotional loss.

Winning strategies can cause us to incur an emotional loss and make bad decisions.

Making emotional decisions based on an inappropriate frequency of exposure is not being data driven. It’s being impulsive and justifying it with data.

Many people bail on winning strategies due to an inappropriate frequency of exposure.

Here's What I Do:

1) I track everything that I can track. When the time comes to potentially change the behavior, I want a trend over a long period of time to make decisions.

2) I track it, but I do not look at it. I do not look at data unless it is specifically to inform a decision or behavior.

Before taking a peek at something I just ask myself:

"If I don't like what I see, am I really willing to make the changes necessary right now?"

Sometimes the answer is no. So I don't look.

Since I am unwilling to unable to change the behavior, there is no benefit. I can only incur an emotional loss.

Appropriate Frequency Of Exposure

So what is the appropriate frequency of exposure? This depends on your Solvable Problem™, macro belief, and understanding of yourself.

Solvable Problem™

If you have a goal that you need to fund in six months you will have a different strategy than a fifteen-year target. This is one of the reasons longer time horizons have a higher success rate. Your target and time to target inform strategy. Strategy informs time preference. Time preference informs the appropriate frequency of exposure.

As seen in the example above, the investor will make better decisions by looking at his portfolio once a year than once a minute.

Macro belief– What is your belief in? If you believe that real estate in Arizona is going to 10x over the next ten years, it doesn’t make much sense to look at your real estate portfolio every day.

The exposure might cause you to abandon a strategy developed in a state of sobriety.

An inconvenient truth - if you need to look every day it's because:

- you don't have a fundamental belief/understanding of your strategy

- you are hoping to see something positive; cheap dopamine.

- a combination of both

Remember: a loss hurt more than an equivalent gain feels good. Seeking a small emotional peak could lead to a far greater emotional valley. Seeking cheap dopamine always has the same risk profile:

Best case: small emotional win

Worst case: massive emotional loss/deficit

Asymmetry to the downside.

Seeking dopamine without effort is always a stupid bet.

Understanding Yourself

Some people have inoculated themselves to data. They have no emotional response at all. Most are impulsive.

Pay attention to your response to data. Ask yourself if "knowing more stuff" or "having more data" is actually serving you.

The Impact

I have seen business owners on the verge of losing tens of millions of dollars because of their obsession with more data. Most of the time, we are able to re-orient them. Sometimes they can't shield themselves from the data or override their impulsiveness.

We've seen a few blow themselves up or abandon winning strategies due to the emotional decisions they call "data-driven" decisions.

Relationships, fitness, investments, and business decisions are all impacted negatively by an inappropriate frequency of exposure.

Remember that the deficit is an emotional one. If you create an emotional valley for yourself, it will negatively impact EVERYTHING you touch.

Something to think about.


Collect this post to permanently own it.
Nic Peterson Stuff and Things logo
Subscribe to Nic Peterson Stuff and Things and never miss a post.