Sabbath Piano

This is a Unclassical Piano, a 14-year-old kid who's uploading video to YouTube. Well, "video" -- it's really just him playing piano, so the major point is the audio. The pieces (he has 16 online) are all covers of classic rock tunes. And, uh, they're pretty good. He's not working from sheet music. He's listening to the songs and working out the arrangements himself. That's incredibly cool, and he's going to learn a heck of a lot more that way.

It also makes the interpretations a lot more interesting. These are songs that have been played on the radio constantly for the past 30 years or so. I couldn't guess how many times I've heard them. Now for my tastes, he does miss some crucial components of the songs. Obviously he has to make choices what to include -- he's not going to replicate a 4- or 5-piece ensemble on the piano. Sometimes, he makes choices that skip little bits that I think are crucial to the songs.

And I think that's incredibly cool. Despite being overplayed, these are great songs that have some interesting parts. And it's easy to get set in your ways in listening to things. It's great to hear them from a fresh viewpoint and catch the elements that stood out for him.

For some reason I've been growing more interested in classic rock again. Part of that is because local radio is pretty much nothing but, at least for what I might be interested in, and most times I won't take the time to set up the iPod for short trips. So I'm spending more time thinking about these songs -- the classic rock canon, if you will.

And the thing is, they're pretty good. It's easy to get sick of them because they're so incredibly overplayed, but there's a lot of very good music in there. I sat in the driveway on Saturday listening to the last few minutes of Stairway to Heaven. That's almost the poster child for overplayed, overindulgent 70s arena rock. But you know? It's a pretty good song. It features great work from all the band members. Plant doesn't even screech until the last minute or two.

There's a lot of stuff like this. It's easy to rant against classic rock, because that's just about all you can hear. Almost once a week I'll flip stations to hear the same song in two places at once. That's why WTTS is so disappointing. It's a good radio station and has the most independent playlist in the area, at least of anything bigger than a high school station. They're willing to play songs that don't get airplay anywhere else. And yet they still spend the majority of their time on stuff that I can hear ten different places on the dial.

It's a real annoyance, but it doesn't make these songs bad songs. It's easy to conflate "overplayed" with "not good". I'm going to try to get out of that habit. I'm not going to stop buying new(er) music. But I do think I'll go back and add some of the older stuff to the mix.

Polling, Polling, Polling

It's now exactly a week after the New Hampshire primary. So I suppose I'm completely too late to comment on the results. However, they're not as interesting to me as the fact that Clinton's win was very much unpredicted by the polling. Much has been written about this and I don't have any particular insight into why the polling was wrong. What's more interesting to me is whether polls are all that useful, and whether there's a trend towards people actively working to subvert them.

I have no data, only anecdotes. As with any US citizen over the age of 18, I've responded to numerous polls and surveys, through mail, phone, and online. I remember for instance being about 19 or 20 and keeping an Arbitron (?) radio listening diary for a couple weeks. Even then, I remember my first thought was "what radio stations or DJs or programs do I want to help?" I didn't create a completely false diary; most of what I wrote down was pretty much what I was really listening to. But I did make a point to listen to programs that I thought deserved attention, and if I caught myself spending time on something I disliked -- or if I got mad at something I heard -- I made a point to not record it. And I thought about who needed help: I didn't worry about recording data for shows that were already wildly successful, but I did for things that I was worried would be cancelled.

Why'd I think that way? There are a couple of pretty useful defensive mechanisms here. When a stranger asks you to do something, it's not a bad habit to ask both "what do they get out of it?" and "what's in it for me?" I wouldn't recommend making decisions based only on those questions, but it's worth knowing the answers even if you don't make use of them. So with Arbitron: what did I owe them? I think they were paying me $5 or something, which was nice but certainly not a complete compensation for the work I did for them.

I have used variations on this theme in many surveys and polls since then. I doubt that this demonstrates any particular insight, and I wouldn't be surprised if many people follow the same approach. More generally, I would bet that people just don't care a lot about giving thoughtful answers. After all, this person on the phone is asking for ten minutes of your time. What do you owe them?

As our culture gets more jaded and over-saturated with targeted marketing, I think this kind of thing will only increase. Polls will get less reliable and people will start to question their value.

Does it matter?

Extreme Pareto: the 98/2 Rule

Oh, I so really don't plan to make a regular habit of this. But it is the Tuesday of the New Hampshire primary, making the second major milepost of the nomination process. And despite my earlier whining, I really am excited about the election. There is a very good chance that in November we'll elect a much better president than the one we have now, and that can only be a good thing.

I am however a bit surprised by the apocalyptic coverage. The net effect is that the primaries will be all but won or lost today. Statistically you could easily write off these the delegates and still win the nomination. The kicker, as usual, is the media. Candidates who don't do well in the first two contests are seen as not worth covering, and that's a huge handicap to overcome. If the media and punditry anoints a winner -- and probably a challenger -- it's then much more difficult for anyone else to raise money or get coverage. So it's not that I don't understand the mechanics, but still it seems so crazily premature.

How crazy? As of tonight, delegates are essentially allocated for two states, Iowa and New Hampshire. For the Democrats (as an example) that's 67 pledged delegates out of a total of 3,253... or about 2%. (The percentage is even lower if you look at unpledged delegates but I don't pretend to understand how they land on a candidate.) To put it in perspective, Tennessee -- the 18th-largest delegation -- decides 68 pledged delegates. So if Iowa and New Hampshire weed out all but two candidates from either party, that's based on a tiny fraction of possible delegates. Or to put it another way, 98+% of the delegates have minimal influence on the eventual nominee.

Everybody knows it's crazy, I just wanted to put some numbers on exactly how crazy.

Horse Racing

I'm wary of starting into a political blog topic, because it can be both painful and addicting... kind of like habañero sauce. But I will anyway. I guess it's Monday and I'm surrendering to Cranky Old Guy inside my head. Though in reality, this is more of a media post than a political post.

It stems from my usual Monday morning frustration with Cokie Roberts. She is always annoying, but at least she has the flexibility to find different ways to annoy me. Today's approach was Standard Option B. It's less annoying than Option A but more pervasive across the media so more in need of calling out.

What is Option B? Replacing political commentary with horse race handicapping.

Now I'm not the first to complain about this, either with Roberts or in general. It's a theme that's gained a small bit of traction in the last eight years or so. I probably had my thoughts on the matter focused when I read Eric Alterman's book What Liberal Media?. There are worse examples of reporting and commentary in the media, but Roberts is particularly reprehensible because she has venue on the relatively serious NPR and because she's seen as a respectable commentator.

But like I said, this morning is mostly about the horse race handicapping. Consider this, the next time you're reading or listening to political commentary. How much of the coverage is about who the candidates are and how they would govern, and how much is about their electoral strategy, where the 'momentum' goes, and who's moving up or down in the polls? The vast majority of political coverage you hear will focus on either who leads in the horse race, or handicapping on what's going to happen next in the horse race. Hint: any time someone talks about a poll, this is what they're doing. But it goes beyond that too, into discussing what's the strategy of a particular candidate, or how are they changing their message.

We're so used to this that it's a bit tough to take a step back and think about why that's kinda crazy. Think of it this way: if you're making a decision on who to vote for, what do you need to know? Well, you certainly need to know what the candidates will do if they're elected. That includes things like their plans, their history, their stands on general issues, and so forth. It's probably also worthwhile to know about their personalities: how they function, who they associate with, what's their past history. (I'm used to being annoyed by coverage of this personal stuff, and it does get overblown, but it is relevant to how the person would govern.)

You can start collecting that kind of information straight from the candidate. They will tell you all about themselves and what they want to do. And that's good, but of course it's also highly biased. To some extent you'll get contrasting stories from their opponents, but that's going to be just as biased. Debates and live exchanges can help with this but it's still not a complete picture. So it's critically important to have somewhat evenhanded media coverage of these things, so that you can get an independent story on what the candidate is about.

And yet the bulk of the coverage is not about the candidates and their ability to govern; it's about who's going to win. That's a lot easier because it has a clear storyline. (It' an incredibly lazy way to report, because you can just keep writing the same basic story while changing the names and numbers.) It also defends a bit against accusations of bias, by focusing on metrics like poll numbers and dollars raised or spent. Most importantly, I think it drives a sense of excitement about elections, because races are exciting, right? And that excitement drives TV ratings or newspaper sales or what have you. But all of this is completely unhelpful in deciding who you should vote for.

It's not that this horse race coverage is inherently wrong. It's good to know what is happening with polls or money, and it's an interesting behind-the-scenes look at how political machinery functions. But that should be way down the list of topics to cover, instead of #1 through #3 in every news story.

It's really tough to avoid this stuff. I mean, as a reporter you need to talk about these candidates every day for 18 months or so. How many times can you write about their health care ideas? If the candidate is even mildly competent this stuff isn't going to change very much. But it can be done. Matthew Yglesias is making a pretty good effort at it. He's not religious about it but his balance is pretty close to right. I'm sure there are others out there too who are doing a good job, but after 2004 I had to cut back my political blog reading quite a bit.