The past matters. People just don’t know why it matters, and they’re bad at using that information.
The past matters when looking at large numbers. It’s awful when looking at one example. Here’s what I mean: Let’s say that out of every 100 graduates of School A, 80 are now talented contributors in their workplace. Out of every 100 graduates of School B, 40 are talented contributors. Someone shows up applying for a job at your company, and their diploma says “School B” on it. What do you know about them?
That’s right – nothing!
You know absolutely nothing about this person based on that one fact. But humans are bad at understanding that. We look at statistical effects and assume they apply universally and causally. So we not only assume that everyone that comes through our door from School B is a bad fit simply because we’ve evaluated it as “worse” than School A, but we also assume that attending School B versus School A is what made them that way. Neither of those things are true.
In fact, one of those beliefs being false is what proves that the other one is, too!
If School A made people into talented contributors, then 100% of graduates would be, not 80%. And if School B made someone rotten, then no one who graduated from there would be any good instead of 40%. The reality is that people are how they are for many, many reasons and nothing is universal.
So from a large, organizational standpoint, you might say “We know that more graduates of School A tend to contribute at a higher level, so let’s prioritize getting a higher number of interns from School A,” and that would be reasonable. But only if you also say: “But let’s not rely entirely on one statistic; let’s make sure we also have our own way of evaluating merit based on multiple factors so that we can confirm that what we’ve heard is true, perhaps even determine why, and not miss the good talent coming out of School B as well (or let the flashy credential of School A blind us to people that might be in that 20% minority).”
People aren’t exactly good at doing that. But hold yourself to a higher standard!
The Opportunity Machine 