Red herrings

Should we include or exclude them?

Was it Colonel Mustard with the lead piping in the Study? Or was it Miss Scarlett with the rope in the Conservatory?

I was in Hatfield yesterday, teaching my Arguing with Numbers course to a group of people from Hertfordshire and West Essex Integrated Care System. One of the questions we grappled with was this: when you're using data to find out an underlying cause of something, and one of your investigations leads to a dead end, should you include or exclude that dead end when you write your final report?

The argument for excluding is that it simplifies your eventual argument. This is particularly relevant if the eventual argument has multiple cause-and-effect steps, like a sort of chain reaction. If you interrupt the flow of the chain reaction by wandering up the cul-de-sacs and back again, that'll make it harder for the audience to follow the argument.

But there's also a good reason why you should include a dead end. It's the same reason why writers of murder mysteries include red herrings in their storylines. The very process of discovering that something is a red herring is itself a piece of useful knowledge that helps the reader follow the investigation.

Sometimes we find our underlying cause not by hitting on the right hypothesis but—instead—by discarding the wrong hypotheses. As Sherlock Holmes famously asked Watson in The Sign of Four: "How often have I said to you that when you have eliminated the impossible, whatever remains, however improbable, must be the truth?" I think that's how I play Cluedo. And I think it's how Coleen Rooney worked out it was ......Rebekah Vardy's account.

[17 May 2024]

If you want to get notifications by email of new content on the Kurtosis website, please email and ask to be added to the email subscription list.