Quote:
Originally Posted by cannon7
Unfortunately I don't think the field of statistics is well understood, even in analytics circles. Why else would the Replication Crisis be so widespread? |
Exactly. It is part of our evolutionary DNA to want to see patterns. It is how we think and perceive the world.
When patterns appear in numbers, we don't just see them, we
want to see them. And we are pre-disposed to draw conclusions.
And that is fine for some things. In fact it is vital for some survival instincts. But it causes big problems with statistical analysis. I have been involved in research for decades in finance, and behavioral finance, where replication crisis is a major issue. Because the financial rewards are so potentially lucrative, people seek patterns that they can exploit and profit from everywhere.
And because we are over-confident in our own abilities, we are quick to believe that which we are quick to see. Who has time to bother replicating our findings? Patterns are patterns, let's go!
And the problem is rampant even among professionals.
In sports, where we have no real incentive to
prove our work, to replicate our findings, it is far too easy to simply jump to the conclusions we seek.
Player A has higher CF% than player B, therefore player A is a better player.