Technically, they did predict it: 538 had 35% chance of Trump winning and emphasized the uncertainty factor, while other outlets had Clinton's chances as high as 99%.
If the results had been different, our prediction would have been correct.
First, that comment is about the polls, not their model.
Second, it's called a polling error. It's actually a thing.
This election happened to be extremely sensitive to polling errors due to the nature of the electoral college and the fact that a number of key races were extremely tight.
When an average poll has a 3% margin of error, and you factor in state-level correlations (and the fact that the state level polls tend to have more variable quality than national polls), a slight error results in a huge swing.
Which is the entire point they're making, if you're willing to set aside the snark and read what they wrote (since it's actually, you know, interesting and informative).
If you're talking about Taleb's "black swan" style uncertainty, that's specifically the kind that cannot be taken into account.
If you're talking about events being correlated instead of independent, that's what set 538 apart from other poll-aggregators: Nate gave Trump a relatively high chance of winning, because a polling error in one state would likely indicate the same polling error in every other state.
As I understand it, FiveThirtyEight's models reflect current and past uncertainty, but do not properly take into account future uncertainty. This greatly undermine's their models for use as a prediction, a probability, or a forecast.
At one point, FiveThirtyEight was forecasting/predicting that Clinton had a 66% chance of winning the election. If there were infinite parallel universes that veered off from that moment in time, their implication is that she would win the election in two out of three times. Taleb is saying that is nonsense, that there is more variability than that.
Of course, Taleb's model wouldn't drive page views or get people excited. FiveThirtyEight gives voters a reason to check-in daily on how their candidate is doing in polls, but it is misleading as a forecasting tool.
Eh? Taleb was criticizing the variability of 538's estimates and said they should not have been making such drastic changes to their estimates. Taleb waffles between Bayesian and Frequentist and in this case he's being Bayesian.
He didn't criticize them for underestimating uncertainty, but for being too uncertain.
Hilarious defence of their work in recent tweet
"Clinton came within 2 points of 307 electoral votes, in which case polls would have been right in 49 of 50 states. "
ie. If the results had been different, our prediction would have been correct.