Poisson, hubris, and punchlines.

October 2nd, 2021

A while back I took a hard stance. It wasn't the first time I'd taken it, and it occurred to me at the time that it was remarkable my outlook hadn't much changed in the intervening years, despite a great deal of change in just about every category in which change can take place. Somewhere within, a voice, a notion, a guy1 even offered the idea that this stolid, stalwart approach in the face of reasonable arguments from far more experienced parties might just be a glaring sign of hubris.

But what checks had I on hubris, then? I answered to the man, and the man gave my mean little stance a place amongst the mottoes on his header. I cherished the complement, and took even greater pride in the hard line I'd chosen. The fates are cruel and the chances are slim, so grab a good sieve and dig in, indeed. I figured I'd passed the hardest test --to find him. That I'd continually, one way or another, passed all the tests in between --the struggle to give myself over completely, the daily challenges of whatever tasks or feats that were commanded of me. I'd failed countless times and risen again, I'd come dangerously close to permanent failure more than once, as well, but I was where and who I was, with who I was with, and so I had it cinched, that primordial challenge of fate, that gaping unknown of chance.

Until the world broke apart completely and swallowed up all that mattered in life. In a way I'd never imagined it could, though of course it could have at any point, in that way or in any other. It wasn't lost on me that life is short, nor precious, and yet...it was. Wholly, utterly, lost on me; in dark forebodings of the future I imagined crying on my master's shoulder on the death of one of my parents, which I knew would come, and not too far into the distant future. But the forebodings ended there, never fraying, not even in the smallest quantum, towards something as unthinkable or unbearable as master himself being the one to die. Even now I daily find my thoughts defaulting to the assumption that the whole thing is a stupid joke, some sort of charade, just another task-based challenge at which I'm doing particularly poorly. It just doesn't make any fucking sense.

And it occurs to me that I can prove it, if I lean on the Poisson distribution. Here:


So consulting the actuarial tables, one of the many thingseverythings I wouldn't know about if master hadn't deliberately and patiently taught me, we find that the rate of death by drowning in the ocean amongst adults aged 25 - 44 in Costa Rica is 2.3 per 100k population. We'll note that the sample spans 2007 - 2009 and be somewhat displeased with the size but appeased by the relative recency. Now, how many such beach trips did the good man make? I'll perhaps disappoint some by revealing there was no ongoing tally board somewhere showing how many of each particular diversion regularly shared in paradise took place. None, except of course, Trilema, which has and knows everything. So let's see, how many of MP's articles mention going to the beach? Ah, to query and to read over each one, to tally and to remember, to relive the unabashed joy and the exhausted pleasure, to bury that one June day's sickening horror in the bounty, voluptuous and wholly his, of all the times we'd been afore.... Well, that knocked a week or so out of me, in which forty or so such pieces were savored --taking pains not to count recountings from the logs, lest I find myself an excuse to truly never finish writing this. Thusly sated, we can also point out the easy cheat of cribbing the man's own 112 - 212 estimate, and taking the average at 162. Seems far more likely.

Let's find lambda, then. Given 2.3:100,000 from the insurance drones and our gnarly reality of 1:162, we'll knock the hundred thousand down to 43479 (I'm rounding up in favor of a clueless fate, nice as I am), and shove it into those 162 beach trips for a lambda of .0037. You see where this is going, do you?

Now we need k, the number of occurrences. That'd be one. One "occurrence", the only occurrence capable of stopping everything that occurs to matter.

(.0037 ^ 1) * e ^ -.0037 / 1! = .0037 * .996306839 / 1 = .003686335, and where a probability of 1 means 100% likelihood, we come out at 0.3686335%. Not half a measly single fucking percent.

Perhaps it's not so much that I'd taken a hard, an unreasonably hard line. Perhaps it's that it wasn't hard enough, because apparently fate and chance are a good deal crueler and slimmer than I'd ever imagined, to bless me with such unimaginable splendor, and then to take it away.

And no, I wouldn't for a moment change anything about it, save its having to end. I suppose that's the true punchline.

  1. Do you have little guys? Ones that doubt, ones that alarm about something, sometimes useful ones that can even be sent to chase after some forgotten item and return with the answer at some later, subconscious, interval? In my experience, most little guys are obnoxious, and can even be harmful at their worst, but they're not all to be painted with the same brush; some are truly useful, earnestly hardworking little guys! I wish I knew how to give them cookies. []

5 Responses to “Poisson, hubris, and punchlines.”

  1. Diana Coman says:

    While the conclusion might even be entirely correct, the sadness with statistics - even when going by the name of Poisson - is that it's more of a slippery eel than any sort of fish and especially when "proving" something is contemplated. Basically one can "prove" just about anything with statistics, it only takes some search for the "right" numbers really.

    The most troublesome part there is the attempt to apply a statistical model to an individual really - ie perhaps you could attempt to calculate the number of deaths by drowning in CR this year and then use that to estimate the chances of drowning when going for a swim (such chances should not change if it's your 1st or 100th swim that year and this doesn't hold in your model, either) but even so, for each individual considered separately like that, the chances are always 0 until they suddenly become 100%...

    • hanbot says:

      You know, it's kinda funny (to me anyway): I took a while to write this, what with all the "number of times we'd been to the beach" gathering, and then sat on it, feeling certain I'd munged the formula. Eventually I went back through it, satisfied with the operations but with the sudden realization that that aside, I'm not using it for the correct kind of thing. Whatever, I told myself, it's some thoughts, you've put them down; publish. And I wondered if it'd ever come up again. I suppose all this to say is that I'm glad: that I'm wrong, that I can still be naive, and that someone else can see it.

      the chances are always 0 until they suddenly become 100%...

      Indeed. I know I can't actually make sense of it, just grasping at (slippery, as you say) straws.

  2. Row says:

    Well, an arbitrary adjust of a Poisson distribution to disconnected events is just that... us humans trying to make sense of what's essentially chaos.
    I however share a minuscule fragment of your pain, minuscule, but indeed mine. It's so good reading him.
    Be well.

  3. late says:

    Well.. in the words of Goran Bregovic - The fish doesn't think because the fish knows everything.

Leave a Reply to hanbot