Comic Book Lessons in Analytics — The Joker
HA HA HA HA HA HA HA HA HA
Weren’t expecting this one, were you?
I’ve written a few articles on the analytic capabilities of comic book heroes. Batman is often called the World’s Greatest Detective, and he earns the title through constant investigation and brilliant insights. Other heroes demonstrate the power of intelligence and analysis.
But in most such things, there is balance. Knowledge and ignorance, darkness and light, yin and yang, Holmes and Moriarty…
Sorry. Moving on…
What are you talking about?
Let’s look at that last one for a minute. In almost every way, Holmes and Moriarty are equals. They are observant, clever, incredibly intelligent, erudite, and ingenious. It is their morality (and maybe ego) that defines the difference.
Well, Gotham is not London. And the Joker ain’t Moriarty.
Batman is generally recognized as the best detective in the DC Universe. His grasp of analytics is inhumanly impressive, not to mention obsessive. He has been studying the art relentlessly since he was 8 years old. Bruce Wayne is an intelligent man who has spent his life studying people, including himself.
There is one man who knows Batman better than he knows himself.
I mean this one.
The Joker has not studied analytics (as far as we know). He has not been working on his detective skills and crime-fighting techniques (as far as we know).
Half the time, he’s not even sure what he’s doing (so he says). And yet he still manages to match the Dark Knight time after time.
Do you have a point, or do you just like Batman?
I like Batman. But, yes, I have a point.
Most analysts are drawn to the field because they like to study the ‘building blocks’ of things, primarily in the form of data. They study the best techniques and they learn from the masters. They find the patterns and relationships that drive analytics and make it useful.
What happens when there is no pattern? When the dots are scattered? When the lines don’t meet? We make regression models to ‘account’ for the random values. We exclude variances that we can’t squeeze into a box.
The Joker is an example of this problem given form. And he knows it. From his perspective, he has been trying to make Batman understand this fact for years.
Batman cannot accept this premise. He really is convinced that there are patterns underlying everything, even random circumstance. This obsession powers his relentless analysis and can be incredibly useful, but it can also be limiting.
Still waiting for that point…
We see patterns in everything because that’s how we understand anything. We needed to communicate, so we created language.
We needed to track time, so we noted that the sun rises and sets on a fairly regular basis. Then we had to create a model (the calendar) to manage the fact that it was only a ‘fairly’ regular basis. Our calendars started to fall out of sync with what we expected, so we adjusted the rules for calendars.
I mean, “leap year”? Really? Just add one extra day every four years? But it was too late by then, and the whole year concept still worked for the most part. We just had to keep making slight changes. Like deciding we wouldn’t add a day on century years. Unless the year was divisible by 400. Now we still add ‘leap seconds’ on occasion, just to keep that clunky model in line.
Maybe I should have chosen Calendar Man for this topic. Another day…
Anyway, analysts have to look for patterns and relationships and meaning in order to move from ‘analysis’ to ‘synthesis’ to ‘insight’. Just like Batman. Which leads us here…
You can catch yourself feeling forced to find a pattern that isn’t really there (or at least not properly defined), and ignoring or excusing the data that doesn’t fit into it.
Or worse (and probably just as common), you can get caught in the trap of seeing a pattern where none exists and totally bending the data to meet the theory you’ve been told to prove.
You wouldn’t be the first.
There are wild cards. There is randomness all around us.
That doesn’t mean our job is impossible or hopeless.
It means that we should recognize the existence of randomness and acknowledge it. It means that we do not force our conclusions onto the data.
Dude, you’re all over the place here. What are you saying?
It kinda got away from me, yeah.
I actually have two points, but I was hoping you’d see them by now.
- No matter how good you are, the data will always escape your toybox.
- When it does, you can build a better box around it.
Find the patterns. Create the models. They are still useful. They are still necessary.
Just accept the limitations. Don’t try to pretend you can account for every variance. Learn from each encounter.
Oh yeah- and never face the Batman over a chess board. Or the Joker anywhere, ever.
