November 21, 2011

My Complete Track Record

A fan of this blog recently asked about my track record since I started making predictions. I mentioned that I publish an accounting of my track record quarterly, when I announce the True Top 50. He said he wanted to know what my record was to date. I confessed I had never done more than a year's worth of analysis, and then had focused only on hits I had gotten either right or wrong.

Fast-forward a few weeks, and he presented me with a gift: He had actually gone to the trouble of combing through my entire blog and matching up predictions with results!

Long analysis short, here are the results:

  1. Of the 29 times I predicted a test was "likely to succeed," I have been right seven times. That's an accuracy rate of 24% or one in four.

  2. Of the 117 times I predicted a test was "unlikely to succeed," I have been right 104 times. That's an accuracy rate of 89% or getting it right about nine times out of 10.

  3. Of the 36 times I have predicted "bomb," I have been right 34 times. That's an accuracy rate of 94% or getting it right about 19 times out of 20.

For this exercise, my accuracy was determined by a TV spending threshold and a check of retail stores. If a campaign spent at least $500,000 (at rate card) on rated stations and the item made it to retail, it was declared a success.

One other thing: It seems I only predicted "hit" one time in blog history (when Guthy-Renker and IdeaVillage partnered to do Sexy Legs), so they lumped that one in with the "likely to succeed" results and counted it against me (since I was wrong). There is one other time I predicted "hit" (with Allstar's Snap-2-O), but it is too soon to determine the outcome of that campaign.

So there you have it: A full accounting of my track record for anyone who wanted to know. As for my feelings about this record, I don't take much pride in predicting failures accurately. If you think about it, the failure rate for the industry is easily 19 times out of 20, so all you'd have to do is call everything a bomb and you could match my record.

I am a bit more proud of my success at predicting winners.While failing three times out of four might seem like a terrible track record to the uninitiated, any industry insider would love to be wrong that infrequently!

Now if only I had an equity stake in all of those successes ...