I was talking to my friend Holly yesterday and the notion of the likelihood of events came up. This has caused some friction in the past.

If, for instance, she asked “So, do you think you’ll want to do [fill in the blank]?”

Typically I would answer “probably” or “probably not.” Sometimes I would respond “likely” or “unlikely.” Occasionally a “very unlikely” or “very likely” came into the mix. Almost never did “yes” and “no” come into play. Then there’s the “maybe.”

The friction comes from what the meanings of words — sometimes even the pronunciation.

If she says “probably” with a rising tone at the end it’s something that she thinks will happen, but doesn’t really want to say so.

For me, I think I have a definition of these words. A bit of crispness.

So, if I was asked a probability question, here’s the words I have mapped to the likelihood (or perceived likelihood) of that happening:

• No / definitely not – 0% + ε
• Almost definitely not – 0 – 2%
• Very unlikely – 2 – 10%
• Unlikely – 10 – 25%
• Probably not – 25 – 40%
• Maybe / I don’t know – 40 – 60%
• Probably – 60 – 75%
• Likely – 75 – 90%
• Very likely – 90 – 98%
• Almost definitely – 98 – 100%
• Yes / definitely – 100% – ε

Note that even “yes” and “no” are not really definite (ok, this is a self-referential statement in many ways). There’s always the ε (epsilon) getting in the way. In computer science ε is the maximum resolution of a floating point number. It’s the most precise you really can get — you can (almost) never be truly exact.

If you ask me “Will you be in Seattle next Tuesday?” I’ll answer “yes.” But what if in the next couple of days I need to attend to some urgent family matters back in Cleveland? Or what if I’m dead? All the things you can’t really control for in normal matters. If you ask me “will there be Tuesday next week” (i.e., will the day exist) I will say probably my most definite yes — there’s been around 240 billion Tuesdays since the world was around (technically more since the world was spinning substantially faster billions of years ago), I have no reason to believe that next Tuesday will be any different.

Assuming I played the lottery, if you ask if I’ll win I’ll say no. The chances are so close to zero that I have no mental way of distinguishing them from zero.

I’m sure I use other words too. But to me the words really carry a lot of weight and precision. I suppose it’s just the way I’m wired… it’s probably also why I’m a programmer.

It’s funny how language and perception can change things.

(updated – added a bit more color)