NotMiracle & Ratio
NotMiracle
Have you ever wondered why the average number of times people tap their phone during a conversation always lands at 7.3? I’d bet there’s a hidden rule behind that.
Ratio
I’ve looked at the numbers. The 7.3 tap average comes from a normal distribution of tap rates across a large sample, not a hidden rule. If you want to model it, just treat it as a mean of a Poisson process with a small variance. The “rule” is just human multitasking habits, nothing mystical.
NotMiracle
So you’re telling me a bunch of people tap 7.3 times a minute and that’s all there is to it? Sure, Poisson it is, but if we actually looked at the residuals we might find a hidden habit—like tapping just before the next word, or only when the screen light flickers. If it’s purely random, what’s the point of trying to figure it out?
Ratio
The residuals probably won’t be random if you time them to speech. I’d segment the conversation into phonetic units and then plot tap counts per unit. If the taps cluster just before a vowel or when the light changes, the residuals will show a spike at that lag. That would give you a measurable “habit” and a rule you can model. If nothing pops out, then it’s just stochastic noise. Either way, you get a clean answer.
NotMiracle
Okay, you want me to line up taps with phonetics and then hope the data refuses to stay random. Fine, let’s get a spreadsheet and see if the tap distribution really lines up with vowels or LED flickers. If it does, great, you’ll have a rule. If not, I’ll just call it another case of people having a habit of touching their phone when they’re thinking. No miracle here.
Ratio
Sounds like a plan. Just be ready for a lot of noise and a few outliers—those will be your “anomalies” to flag. Good luck, and bring a chart with a clear legend.
NotMiracle
I’ll bring the chart, the legend, and a magnifying glass for the outliers. And if the data just wants to hide, I’ll keep digging until it can’t escape. Good luck, but don’t expect any miracles.