Random Thoughts, Predictive Analytics and Apple Watches

Last week my colleague Ajay Khanna and I were at the Innovation Enterprise (IE) Predictive Analytics Summit. This was our second event in the IE series this year, having also sponsored the IE Big Data Innovation event in Las Vegas last month.

In both cases we entertained the attendees, and gave them a chance to win an Apple watch at favorable odds. All they had to do was visit our booth with a playing card distributed from the special Reltio deck, and if they pulled out the same card from a full 52-card deck, they won the watch. So they had 1 in 52 chances to win. (Shout out to Kathleen Calderwood, our stellar Dir. of Field Marketing for her genius game idea) 

With the over 200 attendees at the first event, we anticipated that we would give away at least one watch, but alas there were no winners. Three quarters of the way through the Predictive Analytics event and still no one had pulled out a winning card.

Then Ajay did it. He turned to me and said, “Is there a limit to the number of winners? What happens if we get 4 correct draws?”

Is there a limit to the number of winners? What happens if we get 4 correct draws?
— Ajay "maybe I should not have said that" Khanna

Me: “Then we give out 4 watches, that’s the promise of the game. It’s random. All cards remain in the deck, and there is no reliable data foundation, or previous history to improve the chances of winning … besides that will NEVER happen.”

As they say, if you think of it ... whether it’s the universe of energy in the world, as described in the now infamous book “The Secret”, or just random chance, immediately after the break we had our first winner!

The first gentleman who won, a data scientist was in such disbelief that he demanded to see the full deck to make sure we weren’t fixing the odds in his favor. And even more incredibly, just as I was capturing his details, Ajay was playing the game with another attendee and they correctly selected their card too!

Back-to-back. If you are doing the math, 1 in 52 for the first, combined with 1 in 52 for the second, the odds were 1 in (52*52) = 2704 or 0.037%

Needless to say there was a lot of cheering and hollering at our booth as a mini crowd gathered to see what the commotion was about.

And even more incredibly, the next participants didn’t win, but the second person drew out the card of their colleague as they were standing together. The odds of 3 in a row would have been 1 in 140,608!

So what did we learn at the summit?

  • Random is random, sometimes things will happen when you least expect it
  • However the universe is listening to everything you say or do. If you think it, it will happen
  • People love to win, no surprise, and providing an unbiased reasonable set of odds to win is very much appreciated

But ultimately we had a blast at the event, fun is one of Reltio's core values. We amused while getting the point across that without reliable data with an accurate view of historical facts, any form of analytics would be futile. You might be able to use the data you have to do predictive analysis, but those predictions would be wrong, and you may just as well put your business at the mercy of the random universe.

If you don’t yet have a reliable data foundation, you can either think good thoughts about your data, and hope the universe rewards you with correct insights; or visit us at our next event, and we’ll show you how to stack the odds in your favor. And if Ajay is there, your chances of winning an Apple watch just went up :-).