Beware Too Many Statistics
A good argument only needs one or two
Beware lots of numbers. Most significant assertions, claims, predictions etc can be explained logically with 1-2 charts, graphs or statistics. more than that is overkill. That’s because most statistics are ordinary. And most statistics are ordinary, with no interesting observation, because most of life is ordinary.
Most of life falls into normal distributions. This distribution, when plotted, takes the shape of a “bell curve”. In the middle of the bell is the highest point, showing it is the most common outcome (frequency of outcome on y-axis). The farther away an outcome is from the middle, the less frequent it is. Those areas at the end that are infrequent are called the tails. The tails are where the significant numbers are.
Most areas of life are normal and fit the description above. For example, most days are pretty normal, and days that are similar to average are frequent too, but it’s very infrequent that you get married, win the lottery, have an epic adventure etc. These days are in the tails of the distribution.
But remember that these numbers are very infrequent by nature. There aren’t that many of them. So what are the odds that an argument made with 10-15 meaningful charts and graphs has no mistakes. And by mistakes, that can mean the math is wrong, the interpretation is wrong, or even the application of the interpretation is wrong.
A good argument doesn’t need all these graphs. And a bad argument will use lots of graphs on purpose to trick the reader into thinking the argument is good. So beware too many numbers. But respect the numbers that matter.
Question of The Day
If you invested someone else’s money, lost it all and had to send them a memo explaining your decisions in hindsight, would you explain using one chart or ten?
Your Friend,
Noah BigNerd Sochaczevski
PS. If you work in a highly mathematical or scientific field, this may not apply to your work. That’s a more fringe case, I know little about but am happy to learn.