Built on Facts

Randomness for the Weekend

The debate? Didn’t watch it. I’m keeping my “not watching the debates this year” record spotless. Not because I don’t want to keep myself informed but because modern TV debate formats don’t let anyone do anything but go for the best soundbites, zingers, and gotchas. That said I’ll probably read the transcript, but I can’t bear to actually waste the time to sit through watching it live. I’d personally like candidates to debate Lincoln-Douglas style including cross-examination, unmoderated except for a guy with a timer. That I would watch. Anyway, both sides seem to be more or less satisfied with their candidate’s performance so presumably nothing too dramatic happened.

Science is more interesting. Let’s talk about science.

Here at Texas A&M, graduate electricity and magnetism is split into two classes, E&M 1 and E&M 2. The first explores the concrete side of the subject. Configurations of charge and current, dielectrics, the kinds of actual electric and magnetic fields you’re likely to encounter experimentally. I haven’t actually taken that class yet, though I hear it’s pretty tough. As it happens, the two classes are not consecutive despite their numbering and I’ve ended up in E&M 2 without having taken the first. It’s a very different than its companion. Effectively it’s a classical field theory class. We’ve spent a month exploring the covariant relativistic forms of Maxwell’s equations, and we the closest we’ve yet come to a “real world” problem is in our current homework assignment, to solve the fully relativistic equations of motion for a charged particle in parallel electric and magnetic fields. And that’s the whole point of physics: to develop and understand the laws governing nature, and to be able to say something about how those laws manifest themselves in reality. In E&M it’s complicated enough to require a class for each half of that coin. Fun stuff though.

The inimitable Swans on Tea posts a good physics anecdote:

A college physics professor was explaining a concept to his class when a pre-med student interrupted him.
“Why do we have to learn this stuff?” he blurted out. ”
To save lives,” the professor responded before continuing the lecture. A few minutes later the student spoke up again.
“Wait- how does physics save lives?” The professor responded.
“By keeping idiots out of medical school.”

And that’s we we teach Physics 201. Well of course that’s not the official reason, but there’s a reason physics classes are something med schools look for.

Reader exercise: college football rankings are determined by a combination of closed-source computer formulas and human voting. Suggest an objective and fair system to improve on this situation. Mine: use ELO rankings. They’re mathematically rigorous, easily adaptable for the purpose, and computable by anyone. You can know in advance what the rankings will be assuming a particular set up outcomes. This would also help prevent a certain constantly-overrated team from sitting at #1 by virtue of not playing anyone difficult. As of this week it’s no longer a problem for now, but considering pretty much every season since the introduction of the BCS has had major controversy I’d say it’s definitely past time for objective rankings.

Ok, enough Saturday ramblings. Tomorrow of course is Sunday Function day, and I haven’t yet decided on one. I have a few I’m contemplating, so it will be sort of a surprise even for me. Have a great weekend!

Update: Here’s Glenn Reynolds on the Chinese conducting their first spacewalk. I agree with his assessment. Good for the Chinese! How about another race to the moon? Maybe a friendlier one this time!

Comments

  1. #1 BrianR
    September 27, 2008

    “Not because I don’t want to keep myself informed but because modern TV debate formats don’t let anyone do anything but go for the best soundbites, zingers, and gotchas.”

    I generally agree with you, and last night’s debate certainly had some of that … but the format that Jim Lehrer used did try and minimize some of that and it worked to some extent. Both candidates certainly used their time to state their talking points and slogans (especially in beginning and end) but there was also some decent substantive exchanges (at least relative to presidential debates of recent years).

    My wife and I don’t have cable and watched it on PBS … that way we are shielded from all the ‘sporting event’ crap that the major and cable news stations were doing.

  2. #2 Alex
    September 27, 2008

    They changed the format this year. As a former LD debater, I have to say I was much more satisfied with last nights debate than anything from the ’04 or ’00 seasons. If you want to learn a bit more about the format, see: http://www.debates.org/pages/news_092108.html

  3. #3 dWj
    October 3, 2008

    http://homepages.cae.wisc.edu/~dwilson/ is even simpler than Elo, and is easy to explain to football fans, but is not stupid the way some rating systems with those characteristics are.

    Elo, as the chess world has done it, is unidirectional; if you play someone who was previously untested, and in the next round they beat someone who’s really good, you don’t get credit for the new information that this player was actually pretty good. If you simply reiterate the season over and over and try to get elo ratings that are consistent, any team that goes undefeated will run off to infinity. I’ve been playing with something elo-like that simply imposes a quadratic penalty on any rating being too far from zero, but I wouldn’t care to try to sell it to an average football fan. (Incidentally, Alabama is #1 in the system at this exact moment, but I’m still playing with parameters. Utah, Northwestern, Vanderbilt, and Kentucky are in the top ten (as are more expected teams like LSU). Any system will get more reliable as the season goes on.

  4. #4 Baktru
    January 2, 2011

    Yes I am slowly reading through this blog.

    So I am still a way off current posts. I wanted to note though, don’t use ELO, use the Glicko system. It is statistically more sound than ELO. I’m not saying ELO is bad, but it has some issues (as shown by the chess federations needing to intervene every now and then). Glicko uses three parameters continuously, rather than one and a half in ELO (with the different K rating for new players). Current ability, certainty on the ability and ability variance or consistency. I.e winning against the best known team and losing against the worst team in succession will increase the ‘inconsistency’ value and has an effect on the calcs as well.

    Not as simple as ELO, but it seems to work better.

    B.