In Greek mythology, Cassandra was given a dual gift and curse: that she would accurately predict the future, but that nobody would believe her prophecies. She became a tragic figure when her prophecies of disaster went unheeded. In modern usage, a Cassandra is usually just a pessimist: somebody who predicts doom and gloom, whether people pay attention to them or not.
We know that people are generally rubbish at accurately predicting risk; they seem to constantly over-estimate just how often things will work out. This is usually due to either the planning fallacy or optimism bias (or both; they’re very closely related). However, while that is by far the most common mistake, and certainly the one that’s gotten all the attention, the opposite is also possible. Yesterday I caught myself doing just that.
I was considering an upcoming sports game and found myself instinctively betting against the team I typically cheer for (that is, I predicted they would lose the game). However when I took a step back I couldn’t immediately justify that prediction. The obvious prior probability was around 50/50 – both teams had been playing well, neither with strong advantage – and I am certainly not knowledgeable enough about that sport or about sports psychology in general to confidently move the needle far from that mark.
And yet, my brain was telling me that my team had only maybe a 25% chance of winning. After much contemplation, I realized that by lowering my prediction, I was actually hedging against my own emotions. By predicting a loss, I was guaranteed an emotional payout in either scenario: if my team won, then that was a happy occasion in itself, but if they lost then I could claim to have made an accurate prediction; it feels nice to be right.
With this new source of bias properly articulated I was able to pick out a few other past instances of it in my life. It’s obviously not applicable in every scenario, but in cases where you’re emotionally attached to a particular outcome (sports, politics, etc) it can definitely play a role, at least for me. I don’t know if it’s enough to cancel out the natural optimism bias in these scenarios, but it certainly helps.
The naming of biases is kind of confusing: I suppose it could just be lumped in with the existing pessimism bias, but I kind of like the idea of calling it the Cassandra bias.