I’m always making claims like “This is my 230th favorite song!” or “That’s the 347th best movie ever made.” I get called out less often that you’d expect.
Thanks for reading! If you’re enjoying my webcomic impersonation, you might also like Her Phone Number is Pi?, Right Before the Quiz, and Student Engagement, Method #47.
When my daughter was younger, she’d say – in total seriousness – things like, “the horse is my 4th favorite farm animal.” But I don’t think she ever made it to double digits. 🙂
Man, horse at #4! Tough judge. Makes me wonder what #3 is. 😉
Ooh, I *like* the “This X is Y standard deviations above the mean” thing. I might have to start using that.
I often say things like “Breakfast is one of the top 10 most important meals of the day” or “Thursday is one of my top 10 favorite days of the week.”
Nice. This is one of the 10 best comments on this post so far.
How about a statement like ‘Mom is in my top ten list of parents’ ?
Heck, she makes the top five!
Umm, bogus.
To find the maximal-so-far (on some scale, call it “value”) object in a sequentially processed list of objects, you don’t have to remember any of the objects. You just have to remember the value of the maximal-so-far, and note whether the current object is higher or lower in value. So provided Blue Guy remembers how good his best sandwich until now is, he can confidently state that this one is better than that one, and therefore the best, even if he’s totally forgotten the sandwich itself.
But to identify a sandwich as the 23rd best, you do have to remember at least the values of the best 23 sandwiches so far. That’s much more challenging.
Yeah, I thought about that. Really, the 4th panel isn’t so much the guy on the right winning the argument through logic, as it is the guy on the left failing to come up with the obvious reply.
Obviously he’s not a computer programmer, then!
You’re assuming that sandwiches have some totally ordered value and that this individual is capable of computing the value for each sandwich, which is very unrealistic.
Most people have access only to an intuitive comparison function — mainly subconsiously processed — that allows them to compare two objects they think about. Better or worse?
Moreover, most people don’t return the same value every time. There is a stochastic element to their evaluation function.
A complicated, random partial order could very well require storage of every data point to be reasonably certain that a newly introduced sandwich has the highest expected value.
Shit just got real.
I understand most everything being said here, but am I the only one who just enjoyed the post? No one? Just me? Okay.
If we didn’t like it, we wouldn’t bother to pick nits: we’d just close the tab and walk away.
I’m loving the debate!
My hunch is that the best working model for sandwich preference would involve a stochastic element, as Talon suggests, but not a significant enough one that you’d need to commit every sandwich you’ve ever eaten to memory. My guess is that a working list of 10-20 would do it (although I’m sure that’s more than most people have). To recall the 23rd-best sandwich, though, would likely require a storage capacity of 100+.
Psst, this is like my 11th all time favorite blog post ever.
All right! That’s 2.1 standard deviations above the mean for “niceness” in the distribution of things people say to me.
n what was the sample size for mean?!! 😉
Hmm… we may have to comb through the footnotes to figure out…