Experts, Polls and Journalism

During the recent US election night I followed the BBC coverage from Times Square, backed-up by the real-time New York Times dashboard, and all the social-media feeds I could handle. (So did Robbie.)

The BBC coverage was excellent, probably second to none. Andrew Neill, Katty Kay and Emily Maitlis are as good as journalism gets, so we can maybe forgive the utterly naff virtual-reality graphic fills dumbed-down by Jeremy Vine, adding absolutely nothing to the proceedings. Emily had everything at her fingertips already, even if the content and the ticker strapline was sometimes cautiously behind the real-time chatter out there. Validation and verification is the reason we have mainstream media after all. And I repeat, it doesn’t come much better than this team.

For most of the night, at the left side of the panel sat Norman Ornstein. A political science commentator with about as much credibility and authority as anyone could have in that role. My main agenda, aside from the politics of the US election itself, is the less sexy topic of epistemology. Knowledge itself – how do we know anything and how do we decide to act and/or communicate based on what we think we know.

I posted this Facebook (and Twitter) comment, early on, ie before the prediction graph flipped (below).

“Norman Ornstein sounds like a real expert.
I so hope he’s right.
(Nothing to do with US election –
just on behalf of experts everywhere.)”

Part of it is statistics – still less sexy – how we handle and interpret the data we do have, and my authority of choice there would be Nassim Nicholas Taleb. But part of it is even deeper than that – the origins and basis of each individual “bit” of data – the models and methods and tools being used by the surveyors and the surveyed. You know they say garbage in garbage out, but we’re dealing with the psychology of game theory and the like here. Where’s the garbage? Good voters are no more beyond a good joke than a good journalist. Where’s the skin in the game. It’s not simply a matter of more tools and more analysis, it’s more a matter of the right data and the right analyses. Less may be more.

It’s possible Ornstein is, in Taleb terms, an “Intellectual-Yet-Idiot” but I sincerely doubt it. But, we are all idiots if we don’t carefully unpick what’s going on here.


Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.