Re^5: Zen and the art of ignoring XP

[Anonymous Monk]

The ever-so-humble tlm (hey, isn't Hubris the perlish virtue, not humility?) doesn't like the comparison with TimToady. :-) So, let's put it in a bit of context. I read somewhere that all it takes for someone to seem a genius is 3%. If someone knows merely 3% more of a subject than you, they seem to be a genius in the subject, even if they really don't know very much. Think of a grade 5 student teaching a grade 3 student math - the grade 3 student would think the grade 5 student "knows everything" about math, when we all know that most PhD's in Mathematics know very little about math. (At least, that's what my manager claims - who actually has a PhD in Math.)

What counts as 3% more "knowledge", though? If we just count hours of education, we get ratios of relative teaching (assuming all students absorb information equally, which is of course false...) that become much higher than 3% for just about all the major stages of learning...

A grade 5 student has 5 years of math education: a grade 3 student has 3, discounting kindergarten and pre-school. The grade 3 student has only 60% of the grade 5 student's education; that's a big difference!!! Much more than 3%!!

An high-school graduate has 12 to 13 years of math education. If we claim 12 years at about an hour per day, with about 10 months/ year spent it school, that's about: 10 (months) * 5 (hours/week) * 4 (weeks/month) * 12 (years/education), or about 2,400 hours of instruction (or around 200 hours/year).

As a math undergrad, I was expected to spend about 30 hrs a week on mathematics (classes/homework) every week. A term was about 3 months of instruction, with two weeks exam prep, and two weeks down time.

If we count exam-prep as "education", we get: 30 (hrs/wk) * 14 (weeks/term) * 3 (terms/year) * 4 (years/undergrad), or an additional 5,040 hours for an undergrad math education.

Now, 1/4 of the degree was expected to be non-math electives, so if we factor those out, we get about 3,780 hours of math education during undergrad, for a total of 6,180 hours of education for a graduate.

In terms of education, a high school graduate has been taught less than 40% of the mathematics that a math grad student has learned. That's one reason that the difference between 3rd and 5th grade seems smaller than between high school and university grads: it is smaller, in terms of relative education. (Another reason is the academic cut-offs: only people who learn math quickly are allowed to take university courses in the first place: this increases the information density...)

If we assume that a PhD takes only an additional four years, at a pace of education roughly equal to undergrad (ignoring increased information density), we get 9, 960 hours of math education: and that's before the PhD settles into his job of researching new things... that makes the gap between math grad and PhD about 62%... or about the difference between grade 3 and grade 5.

And that's kind of how I'm using tlm and TimToady in the same sentence. If we assume, on a scale of 1 to 10, that TimToady's knowledge warrants a 10, and I get a 3, tlm seems to be getting a 5 or 6. (And that's on a logarithmic scale ;->) From my perspective, that's more than 3% - can't tell the difference from here ;->

It can actually seem the opposite: PhDs seem horrible at math to novices, because they can't do simple arithmetic anymore: they only need it rarely, and they use calculators when they do. They can reconstruct all the principles of mathematics starting from basic axioms, though. :-)

Twice, I saw master martial artists at work: both times, I thought what they were doing was trival, and boring. The Tae Kwan Do master's foot was floppy and weak as he flicked little kicks up in the air above his head, and the Ju-jutsu master's throw seemed simple and obvious.

Later, I watched a black belt in Tae Kwan Do surge forward with great effort and concenration: and watched his kick run out of power and fall, despite obvious straining, at no-where near the height of the master's solid little flick-kicks. And suddenly, the master's kicks were impressive.

I watched as six black belts, including my instructor, questioned the ju-jutsu master with obvious confusion about all the intricate details of the "simple" throw I'd just witnessed; and suddenly, the throw was impressive.

I think there's an analogy to coding as well: it's hard to tell if a task is easy; or if a master is making it look easy. Past a certain point, you just have to look at who's impressed by who... you lose the capacity to judge fairly for yourself. --
AC

[willyyam]

I think there's an analogy to coding as well: it's hard to tell if a task is easy; or if a master is making it look easy. Past a certain point, you just have to look at who's impressed by who... you lose the capacity to judge fairly for yourself.

This is paradoxically true with Perl, I think. Initially the novice is amazed and bewildered that the line noise of a given piece of code does something amazing. Later, when you gain greater facility you being to realize that while the code works, it may be taking advantage of Perl's ability to write fast and loose solutions to transient problems, but what's really impressive is the crystalline, lucid gems that are effortless to read and present themselves as paragons of programmer intention. Later still, you see that some of what seems to make complete sense and is useful every day is in fact a manifestation of deep magic. Perl seems to me to be fractally complex in this realm - each epiphany leads me to a new confusion as rich as that I've left behind.