Following the Trump victory, there’s been a lot of news of finger-pointing in the Democratic party. Numerous thinkpieces have been written on the many portents indicating Trump’s election, and how they were obvious. Indictments of the rest of us for expecting a Clinton victory have also appeared, as if it were a crime to expect a victory that didn’t pan out.
We hoped for a landslide. We would accept a marginal majority. For many of us, a Trump victory was (still is) unthinkable.
Wikipedia calls it a black swan event, and they happen. The website fivethirtyeight.com even acknowledged that a Clinton victory had an eighty percent probability (based on the factors they measured). Still twenty percent is one in five, not likely but not uncommon. We’re going to get surprised.
This is, ultimately, a problem with wanting the safety of assurance. Gödel’s incompleteness theorems while they apply specifically to mathematics (axiomatic systems that contain basic arithmetic) they also apply nicely to the real world: We can’t rule out paradox, and we cannot rule out improbable events. There will always be black swan events in any complex system. And yes, it hurts when the consequences of them can be measured in lives and resources.
We’re always going to have the potential for hail-Mary sports plays, phenomenal weather, surprise terror attacks, rampage killers and holes-in-one. Likewise, we’re always going to face the risk of unexpected election results. And we’re always going to be able to see how this unexpected moment was visible and obvious in retrospect when considering the right factors (which were, before the fact, submerged in a sea of other factors).
Black swans, by their nature, take us by surprise. We need to roll with them. When they can prove to be frequent enough and calamitous, we need to prepare for their inevitability, as we do (sometimes inadequately) regarding natural disasters.
And yes, we can take joy in analyzing what happened and how processes evolved to their eventual outcome. We might even be able to add new factors to consider in the future, thus improving our ability to predict. But we’re never going to be able to include every factor ever. The chance for unforeseen consequences will always be nonzero.
The last thing we need to do is shout each other down for why our predictions so thoroughly failed, even the end result is painful.
It turns out some things are just very difficult to predict.