Last year, I shared some facts about March Madness. For example, the chances of a perfect bracket are  at 1 in 9,223,372,036,854,775,808. For those who don't wanna do the math that's 9.2 quintillions. 

This year, 16-seed UMBC (The Chesapeake Bay Retrievers) beat 1-seed Virginia, and left no perfect brackets after only 28 Games. 

As we came into the Elite Eight, nobody picked all eight. Almost nobody predicted No. 3 Texas Tech making it this far … and there were 17,304,320 digital brackets this year. 

Is there any way to accurately predict something with this many potentialities and this much discord?

Different factors can feel like they provide an edge … knowing the history of teams, their ranks, how often they've been to the Final Four, or how great their coach is … but even that knowledge can lead you astray. You'd think an NCAA analyst might have a better shot at a perfect bracket than your grandma or co-worker, but several of the highest-ranked brackets this year were guesses. 

via ESPN

It's fun guessing when you're making bets on college basketball, but what if this is your livelihood …

Prediction is hard … especially about the future.

The way people fill out their brackets often mimics the way investors pick trades or allocate assets. Some use gut feel, some base their decisions on rank and past performance, and some use predictive models … and some of us get disappointed when Duke loses.

 Apparently, humans aren't as great at prediction as we think we are.