Hacker News new | past | comments | ask | show | jobs | submit login

I'm greatly oversimplifying:

Suppose you have a series of 10 pairs of number coming from the universe, lets call them particles. After crunching the numbers you discover they fit perfectly in a linear function y=mx+b. Yep, two constants, one parameter and you can explain all those particles. Obviously, more particles could exist if they fit your curve, and you in fact discover, in time, 15 more of them!

Wow the model works, it predicted a bunch of awesome stuff. You and your peers name it "Standard Model" since it's so good. It's still just a name though.

Now suppose one more particle comes up and it doesn't fit the line at all. Sure you could "extend" the model by converting it to a Laplace polynomial with 24 more coefficient just to accomodate that new particle, but you can't honnestly say it's the same model anymore.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: