Better than 50/50

You're walking in the forest and come upon a bridge. A mysteriously cloaked figure jumps out holding two envelopes. "Each of these envelopes contains a different number. You may pick either envelope. I will only let you cross if you can tell me whether the number in the envelope you pick is larger or smaller than mine" 

Despite being put off by the disruption, you take an envelope and ask: "Can I look inside my envelope before making my guess?"

The cloaked figure is surprised because, until now, everyone has ignored him and his envelopes. After some consideration, he says: "Sure."

You open the envelope, thinking: "I don't even know how large a range he's picking from. This is a total 50/50..."

OR. IS. IT??

What I love most about the solution is it doesn't come from gimmicks, hacks, or any external information. It is, in its purest form, a strategic decision framework. You magically improve your odds with something equally random. It's the entire act of doing something over nothing that improves your odds.

So, what's the strategy?

Pick your own random number, and follow a simple rule: Answer larger or smaller, so that your random number would land between the other two. If your random number is 15 Million, and you open your envelope to read 42, then guess the other number to be larger. 42 < 15 Million < Other number. Land in between. 

While the math is too long for this post, it works out that your odds will always be greater than 50%. Depending on the range and distribution your odds can get as high as 99.999% or just 50.0001%. Either way, better than 50/50.

This strategy breaks the symmetry of the problem because we introduced something when nothing else was available. It's a beautiful example of how strategic frameworks create progress. As Product Managers we constantly face questions with seemingly random answers. Sometimes, despite the data, a decision feels 50/50. Things may feel beyond our control - we have no stake, we have no sway, we have no say.

But, left with no answer, we can ground ourselves in our own reference points. When we establish the right kind of a framework as a guide, we regain control. By introducing information into the system that didn't exist without us we can gauge results, measure impact, adjust, iterate and refine.

Things to keep in mind when orienting around ambiguity:

1. Create a Reference Point: You decide what "looks good"
2. Break Symmetry: Don't treat every path as equally valid; Form a hypothesis, test it out. Let the data react to your criteria. 
3. Do Something: Don't lose out to analysis paralysis or anxiety of insufficient data. By all means pick the best metrics using the best data you have, but empower yourself with the reminder that any strategic pillar increases odds more than none.

It's not about claiming to have all the answers. It's all about creating clear rules for how you derive your answers. It may not guarantee success, but 50+x is way better than 50/50.