"Skate to where the puck is going."
Wayne Gretzky
Pop Quiz Time.
Two twin brothers see the same alarming headline about cholesterol on the same day. They both rush in to get their annual check-up.
Brother 1 (Blue) has a total cholesterol of 245mg/dl, which is considered high
Brother 2 (Red) has a total cholesterol of 215 mg/dl, which is considered borderline
If the goal is to get cholesterol into a healthy range, which brother needs to make a more drastic change? Below 200 mg/dl is considered healthy according to Dr. Google.
Answer: It depends.
It's tempting to make a snap judgment and recommend Brother Blue behave more like his brother. This data point suggests that Brother Red is "healthier".
**This is where I tell you that I am not a doctor, your doctor is smarter than me, listen to them. This is about how I think about data, not about what you should do.
Personally, I wouldn't make any behavior changes based on this data point.
Here's why:
Imagine that the two brothers have gotten their bloodwork done every ninety days. Ninety days ago Brother Blue's total cholesterol was 255mg/dl and Brother Red's was 205 mg/dl.
With that information, the data point becomes a line.
Given the line, are we still convinced that it's Brother Blue that is in the most danger without a change?
I don't like to make decisions based on lines either. **I'm still not a doctor. Your physician's prescription overrides my weird perspective on data
So let's use the science of hindsight to dig a little deeper. What I want to see is a trend. 180 days ago, Brother Blue had a cholesterol level of 259 mg/dl and Brother Red's cholesterol level was 198 mg/dl:
Okay, cool. A trend.
I'm still not a doctor but I'm also not so sure that Brother Blue should change his behavior at all. It seems to me that he is on track and can continue doing what he is currently doing. In fact, changing his behavior could break this trend. Especially if he started behaving more like his brother. Generally speaking, if you're trending in the right direction changes are more likely to hurt than to help.
Brother Red has a "better" cholesterol level today, but look at the trend above. The "better" data point is a point that makes up a less favorable trend.
Given this trend, I would wonder if Brother Red needs to make the most dramatic lifestyle change. Again, I am not a doctor. This is not medical stuff, it's data stuff. It's not a suggestion to do something, it's a suggestion to gather more data before doing something.
Now...
Replace "cholesterol" above example with any metric you use to measure your progress. I prefer metrics that help determine whether something needs to change. Weight, Personal income, total revenue, profit margins, total profit, number of employees, lipid panels... it doesn't matter the metric, it's the same story:
"Skate to where the puck is going"
A single data point can only tell us where we are relative to other people at that exact moment in time.
It does not account for where have been or where we are going if we continue what we are doing, uninterrupted. The first set of data above showed us that Brother Red was in a "better" place than Brother Blue.
Brother Red vs Brother Blue is irrelevant. Where is Brother Blue-headed? That is all that matters to Brother Blue. And I am of the opinion that the answer to that question should inform his behavior.
The only thing worth measuring is Brother Blue today relative to Brother Blue yesterday.
With the intent of informing Brother Blue how to be in a better place tomorrow.
The question to ask is not "Who is healthier, wealthier, happier than me right now?". The question to ask is:
"Am I trending in the right direction?" or "Given my behavior, is time passing going to get me closer to or further away from where I want to be?"
That's it.
I believe that comparison can be useful. Comparing our current self to our former self informs us of the efficacy of our plans and behaviors. Comparing trends can help us understand which behaviors are more likely to get us to where we want to be, over time. Comparing ourselves to others using single points of data has zero utility*.
*Has zero utility to me. It is very possible that I am lacking the skill to do something useful with a single point of data.
Given more time, it appears that Brother Blue's cholesterol will fall into a healthy range. Brother Red may rise into what is considered the high range.
It would be a shame for Brother Blue to stop what he is doing and start behaving more like Brother Red. At least based on this single data point. To be clear, I am still not a doctor. Both brothers may be in serious need of lifestyle change for various reasons. I am not suggesting otherwise. I am showing that the single point of data in the example above does not inform me of anything useful.
Everyone's got their own style. How you choose to make decisions is up to you. I can only share what I do, which is this:
Important decisions are often behavior based. I make behavior-based decisions based on trends, not lines or single data points.
When I find myself playing the comparison game, I try to override my monkey mind. I do this by comparing individual trends. If there is no trend (the data has not been recorded, for example), I will wait until the data is collected. Waiting is hard, but it's better than going in the wrong direction. Keyword: restraint.
I think a lot about things I might want to change in the future. If it's important, I'll set up a way to record the data so that when the time comes, I'll have a trend to help me make decisions.
Do I make decisions without a trend? Of course.
I don't share this so that you have more things to do, like tracking more data. You're probably doing things right now that, if you kept doing them, would get you to where you want to be.
My hope is that in sharing this, you think twice before letting your monkey mind take the wheel and convince you to abandon those things.
And one more time...
This is just how I think about data, behavior, and time. I am still not a doctor. Your doctor is way smarter than I am, listen to him/her.
Onward
Nic