It means (well meant, I am out of that general support hell now) something will be broken every day. Yeah yeah yeah 5 9ās and all that but you know with 20K plus servers of all kinds the math says something is going to be b0rk3d every day. Happily most of the time it is an easy fix.
Literally the life and death balance of children. Just kidding, Iām also a coder. hey guys.
Iām in theatre, so it means no probability of getting a job that you donāt need to explain to your parents.
In my past profession as IT support, we developed response timetables for different users, based on a number of factors:
how far up the VIP chain
type of problem and severity
location
ability to solve it in one fell swoop remotely vs talking them through it
all the other stuff we had going on that day
their personality type and our track record with them
With these factors and probably others, we judged to the minute how long we could wait before even responding to their request. This is because, probabilistically, we had determined that for the majority of users if we waited, they would solve their own problem or get a co-worker to help them. Typically on the order of 15 to 30 minutes. Sometimes hours.
From mining our ticket archive, we had compiled all the requests, followed up by, āNevermind! Figured it out!ā and calculated the amount of time between the two. And then categorized the people and all those factors I listed and then would verbally confer as requests rolled in:
āOD, 40 minutes.ā
āYep.ā
or
āOD, only 15 minutesā
āNo way, 5 today, he is having a freak out.ā
OD = Optimal Delay
Iām sure something like this has been on the IT Crowd or The Office or someplace. I wouldnāt know because I never watched those shows. Found them boring because that was my every day for 15 years.
So, using probability to our advantage, we saved ourselves about 25% labor from maximizing our Optimal Delay stat.
yes, but what was the probability of other factors being involved? (I think this counts as the meaning of probability in the relaying of oral historyā¦)
100% probability of unmeasured confounders!
Unfortunately, the same thing it means for the general population: 90% always happens (unless itās a good thing that would happen to me), 10% never happens (unless itās a bad thing happening to me).
I was going to get some elective surgery and the anaesthesiologist told me for a person in my health the odds of fatality were only about 1% and I said, āThey better not be 1%!ā at which point he switched into some sort of okay, I can talk to you like you have a brain mode and explained that the hospitalās lawyers want them to give numbers and most people translate 1% to ācanāt happen.ā
I also saw a study that people actually estimate risk a lot better when you say ā30 out of every 100 peopleā instead of ā30% of people.ā People literally donāt know what probabilities mean.
Even smart people have a very hard time with simple probabilities and especially the intersections of probabilities. You must read Daniel Kahnemanās Thinking Fast and Slow. These biases and many more are exposed and dissected. Such a great book. Totally recommend.
This is Boing. You forgot an affiliate link.
Probability The likelihood of the occurrence of any particular form of an event, estimated as the ratio of the number of ways or times that the event may occur in that form to the total number of ways that it could occur in any form. (HPJ 60)
Probability
See cross section.
(USAEC-1974)
Related:
Probabilistic model See stochastic model. (NCRP 123I)Probabilistic model
See stochastic model.
(NCRP 76)
Edit: if you are talking about the definition of āprobable errorā then that probably matches the first drawing @95% confidence.
Probability in my work?
Isnāt that the scale of 1 to 1 of if iām gonna screw up today?
Whenever I teach probability I get someone who tells me that there is a 50/50 chance that it will rain in the Sahara desert tomorrow, because it will either rain or not rain.
But to be fair, it is not like I really know what probability is either. Not from a philosophical perspective anyway.
But thatās only when the events in the underlying space of states are equiprobable.
"And there is a 50/50 chance you will win the lottery tomorrow.
ā¦
The brain lottery."
To be honest though anything beyond simple probabilities confound me. One single isolated event? Easy. Start trying to figure it out with branching conditions? Iāll be at the bar.
Well, thatās something entirely different. I think weāre talking about āunderstandā as in the ability to appropriately gauge oneās expectations (e.g., not wanting to get a voluntary surgery with a 1% chance of death), not āunderstandā as in āa first year philosophy student wouldnāt call you and idiot for thinking it.ā By the latter metric of understanding, no one understands anything.
Yup. The other day, I saw someone at the corner shop buy 12 pick5 tickets. I asked them if they thought they were going to win. They said yes. I then convinced them to do the math with me, and pointed out that even with 12 tickets, their liklihood of winning was still about 1 in 300million. This is a state lottery, there are 6million people in my state. They expect a winner every 50 drawings. Not every drawing.
20k+ servers to manage?!?!?! Sweet fuck all! In that case, āprobabilityā would likely āimpossibly highā that Iād quit that job in a fuck-new-york minute.
