Pranks as natural experiments

Ken Levine, as previously mentioned, is an enormously accomplished writer and director of sitcoms going all the way back to MASH, but he started out in radio and still keeps a hand in. A few years ago he was co-hosting Dodger Talk with Josh Suchon on KABC, Los Angeles where part of his job was to deliver local traffic reports...

But I always wondered – was anybody actually listening to these traffic reports? One evening, late in the season, the Dodgers were in San Francisco and I was at the station preparing for my big minute. I was hanging out with Howard Hoffman, the production director, and I suggested a way to see if listeners paid any attention. He laughed and said, “you wouldn’t dare.” (This is where that paragraph on pranks pays off.) I gave him a sly smile and headed for my booth.

I opened the report by saying, “If you’re going to the Dodger game tonight, there’s a fifteen minute delay on the Golden Gate Bridge, the 880-Nimitz in the east bay reports slow and go from Concord…”

I just gave the San Francisco traffic report. Super straight, as if this were a San Francisco station. And I tagged it with the Sprint commercial.

Howard came into the booth hysterical. Now we waited to see how many phone calls we got. This was 6:45 in the evening, during the peak afternoon commute.

So how many did we get? I bet you’re ahead of me. That’s right. None. Not a single one. Zero. The big goose egg. No one from the station ever called me. No one from the Dodgers. Nothing. 

The following year there was no traffic. I hope [the sponsor] Sprint took that money and used it to buy another repeater tower.

Perhaps it would have been different with a home game, but these results do look conclusive. This got me wondering. I can imagine situations where a prank might lead to useful data. I wonder if anyone knows of any actual examples other than this one.