Fermi Questions B/C

[bernard]

[jaymaron]

Fermi problems: https://jaymaron.com/fermi.html
Fermi plots: https://jaymaron.com/exhibit.html
Fermi data on units: https://jaymaron.com/unit.html



The Fermi score function is asymmetric logarithmically. "5" should be replaced with "3", because 10^.5 = 3.16

The Fermi function should be continuous.

The Fermi function should be able to reward accurate answers, like the Metric Mastery function. The cyan function in the plot unifies Fermi and Metric Mastery. There are many cool questions that aren't possible in the Fermi event because of the goofy scoring function.

Caltech has a class on "order of magnitude physics", which is ammunition for Fermi. In the class, the standards for accuracy are high. Accuracy within a factor of 10 isn't good enough. The techniques taught in the class enable you to calculate things to better than a factor of 2, and these techniques should be rewarded.

From the Caltech syllabus for the order of magnitude class: "Emphasis will be on using basic physics to understand complicated systems. Examples will be selected from properties of materials, geophysics, weather, planetary science, astrophysics, cosmology, biomechanics, etc."

Unified scoring function: Let "1" be the correct answer and "X" be the submitted answer.
Score = 3 - log_10 (X-1)
If X<1, replace X with 1/X.
If Score > 6, reset Score to 6
If Score < 0, reset Score to 0



https://jaymaron.com/cobalt/fermiscore.png

[knightmoves]

Unified scoring function: Let "1" be the correct answer and "X" be the submitted answer.
Score = 3 - log_10 (X-1)
If X<1, replace X with 1/X.
If Score > 6, reset Score to 6
If Score < 0, reset Score to 0

A lot of Fermi Questions don't have a "correct answer", and for a lot of other questions, knowing enough to get "the correct answer" is of zero value.

Somebody knowing, for example, the area of a football field to better than 10% is worthless. There are people who play or watch football, and will know exactly how big it is, and there are people who can wave their hands and say "it's sorta that big" and get somewhere close. The former is not of more value than the latter.

[Klebb]

Hello Science Olympiad Community!

I'm back with another big document... this time about Fermi Questions.

Presenting:
A Brief Introduction to Fermi Questions

(my titling skills aren't exactly great)
I've been asked a few times how to get started on Fermi Questions, and over the past few months I've been working on writing up an answer to that question. I hope this suffices!

Read it here: https://bit.ly/3fxpANh

Good luck and have fun with Fermi Questions!
- Klebb

[jaymaron]

Each problem has a customized ceiling and floor for points. The choice of the ceiling score can be chosen so that 10% is the best accuracy achievable. The ceiling score is then "4".

This is a set of Fermi problems where a ceiling and floor is specified for each.
https://jaymaron.com/fermi.html

[Flavorflav]

I'm writing this for an invitational. How many questions should be on it? I was thinking 50, but I am seeing some out there with only 30. Thoughts?