Ask HN: How to get comfortable with large numbers?
Does anyone on HN have a good resource for becoming comfortable analyzing and relating to large numbers? I see billion dollar valuations and trillion dollar stimulus packages, and I am not sure the best way to interpret it to make it tractable and understandable.
Appreciate any books, lectures, comments, or courses on getting a better grasp on dealing with this.
If anyone is willing to talk about their own experience, I'm interested to know how you wound up in the position to make decisions with lots of zeroes on the end where the rounding error might be in the 10's of millions. And are you comfortable making decisions that large? If so, how did you get comfortable with it?
6 comments
[ 3.6 ms ] story [ 38.6 ms ] threadIt gets wonky when you realize the difference between the top schools and the mediocre schools (as a proxy for talent) is 30-40k annually. That’s nothing in the scheme of large deals. I guess looking a bit more, my question could also be worded to ask what changes or becomes less relevant (or more) as the numbers get bigger.
I’d argue the tools required to budget and effectively allocate $100 are different (or should be) than those needed for $100,000,000. Granted, it’s not 1,000,000 times harder to budget (and your categories are bigger), but there should definitely be more care and thought put into the $100,000,000 problem.
I’d like to know what that looks like or if there are resources for making those decisions well. For example, at that level each 1% advantage becomes a million dollar improvement, so hiring a Harvard MBA to study the problem for one year at $150K salary suddenly makes sense. You don’t do that for a $100 problem.
In my mind it only gets weirder as you get into the Billions and you can begin to do some really weird and anti-competitive stuff.
Its kind of like starting out on a 14" monitor and then getting a 30" monitor. At one point was sheer amazement but after awhile of operating at a larger magnitude your sense of scale adjusts. You get used to it. Adaptation is real. That big monitor just blends into the background and smaller monitors seem quaint and constricted.
I sometimes deal with storages in the petabyte range. The number of drives and equipment required to get a petabyte currently requires significant construction. Under my desk is a 60TB array that was built in an afternoon. That used to be large. Now I treat that as just long term storage for a 16TB SSD array. That was unfathomably huge for a persistent ram disk. The first hard disk I bought as a component I installed was a 400MB disk. I remember when 1.4MB floppy disks were huge. Because I started out with 5.25" which held 360k.
Its all about expectations, utility and what you are doing with these numbers. If it helps, you can make up a unit of your own as a temporary placeholder. Eg if I was thinking about property as a developer then 1.5 million might be an "average property". Then I would mentally think in those terms. 15 million is then just 10 properties.
If 10 million is a rounding error and this still seems weird then you need to scale accordingly.
In reality, all the numbers we use are huge, we just don't see it because of the units we talk in. Eg an ordinary orange or apple is made of millions of atoms. But you aren't overwhelmed because you think of it as one piece of fruit. You feel no problem having and thinking about a dozen pieces.
A slice of an apple can be seen as a rounding error compared to a dozen apples. How many water molecules are in that "rounding error"?