archive for November, 2004

Netflix Price War

Ok, enough with the election posts.

Netflix recently lowered their prices for their standard plan from $21.99 to $17.99. This may sound like a big drop, but just a few months ago the plan ran $19.99 per month. It turns out the competition is heating up, and they’re worried. Cnet has an interesting piece on the price war.

Price is important, but it’s not the only reason someone will choose one rental service over another. Netflix needs to do more work to create a community out of their customers. Members should be able to create a friends list, and recommend movies to each other. There should also be discussion forums for people to talk about movies and more. It’s as simple as adding a few more tables to their database and updating their website. The community would be a cheap and effective way to attract new customers and retain existing ones.

If they really want to beef up their service, they should add music CD and video game rental to their existing plan. The music industry would hate a CD rental service, but I’m pretty sure it’s completely legal and could attract a lot of music fans. Video game rental would attract new customers and could be even more profitable than movies because people would hold on to the media a lot longer which would reduce shipping costs. Netflix is great, but as I’ve said before, they have plenty of room to improve.

Voting Methods

As I mentioned earlier this week, there are several methods for placing a vote, and counting votes. I said that I was leaning towards instant-runoff, but since then I’ve changed my mind.

Our current system (plurality I think), each person gets one vote and whoever gets the most votes wins. It’s easy to understand, and easy to execute. Unfortunately, it almost always forces people to vote strategically instead of voting their wishes. It also puts third parties at a significant disadvantage. Unless you think that there are always two sides (no more, no less) to every issue, then it’s obvious that a duopoly (two party system) is bad. Not to mention that sometimes the two major parties will be on the same side of an issue.

The solution: change our voting methods. First of all, instead of just having to pick one candidate it would help a lot to be able to pick multiple candidates. Beyond that, it would be nice to be able to rank your preferences. We should also use the most effective for counting votes that will most respect the voters wishes.

The Condorcet method allows voters to specify their preferences by ranking the candidates, and then provides a mathematically sound method for taking those votes and selecting the candidate that most people prefer. Unfortunately, it’s not the easiest method for the layman to understand, and would require computers to do the counting (not always a bad thing). It will be a tough fight for supporters of the Condorcet method to get it implemented.

Approval voting works by allowing voters to select all of the candidates that they approve of. It doesn’t allow them to rank their choices, but it’s easy to understand and easy to tabulate. It’s also better at allowing voters to speak their mind, and it gives third parties a real opportunity to win by avoiding strategic voting caused by the spoiler effect.

If you’d like to learn more about different voting methods click on over to

Electronic Voting

Electronic voting is a great idea, if executed properly. For example, it might be a bad idea to put your elections of a company whose CEO is “committed to helping Ohio deliver its electoral votes to the president next year.” (see 1, 2) Really? Ohio? That’s funny… it’s not like it’s a swing state or anything.

To do it right, all of the software involved in electronic voting should be open source. The most important reason for doing this is because the code would be open to public scrutiny. Any interested party in the world would be able to look for flaws in the software, and offer fixes. We’d probably want some sort of paid “gatekeeper” making sure all of the changes were appropriate, but a gatekeeper would be a lot less expensive than a whole commercial development project. This leads to the other benefit of using open source code: cost. Election software really shouldn’t be that different from one place to another, so there are huge savings to be made by having everyone in the country (or world even) use the same code base and only making small changes as they see fit.

Electoral College Followup

I got an email from someone who read my blog today, so I thought I’d reply to it here. I was planning on doing a follow up like this anyway, so this lets me do both at once. I don’t know if the reader wants to be identified, so I’ll remain anonymous. The points made in the email were that the Electoral College is beneficial because it produces stability and benefits third party voters by making candidates take them into account in swing states.

Here’s my reply:
I’d agree with you on some of the points. I’d prefer EC reform over totally throwing it out. If we changed to a system where electoral votes were distributed based purely on population instead of population plus statehood, then we could still get the beneficial stability without sacrificing equal say (the stability comes from having to count far fewer votes at the national level). I can’t say that the US Constitution is unconstitutional, but I believe that the distribution of electoral votes is in contradiction with the the spirit of the 1st and 14th amendments. I don’t know what the populations of the different states were back when the constitution was written, but I bet they were a lot closer than they are today.

As far as winner-take-all systems helping third parities, I don’t think it helps them as much as a proportional split of electoral votes would help. I think going into the election, Bush knew that he didn’t have a chance at winning California or New York, the two states with the largest urban population. Since he didn’t have a chance, he didn’t have any incentive to appeal to third party voters like the Greens. The same is true of Kerry. He knew he had California in the bag, so he didn’t need to listen to third party voters either. Winner-take-all really only helps voters in swing states. Proportional voting would make every state a battleground state. The other benefit is that we might actually end up with third party candidates winning some electoral votes. It would probably not enough to win an election, but it would bring them into the spotlight and put their issues on the table.

It might seem like proportional voting would lead to more instability. This is probably true, but I don’t think it would be enough to cause problems. It’s no where near as bad as the instability from a pure popular vote. I think the benefits of the above two changes far outweigh any draw backs. If we have a lot of problems with “local voting errors”, then we need to fix the system, not hide it.

UPDATE: By the way, it was Gordon who sent the email. Check out his blog sometime.

Election Results

Ohio hasn’t finished counting votes yet, and all of the media is saying that it’s “too close to call” but if you ask me, it looks like Bush won. I didn’t vote for Kerry or Bush, and I really don’t like either of them. In my opinion, it’s probably better if Bush wins because at least I know he’ll be gone in four years while Kerry could be around for another eight. I hope we have better options next time.

I’ve tallied up my votes and compared them with what actually won. Overall, out of the 21 things that I voted on, I voted the same way as the majority 11 times. That’s about a 52% success rate. I did much better with the California Propositions (everybody I voted for lost). Of the 16 props on the ballot, I voted the same way as the majority 11 times or about 69% successful.

Election Day

I was planning this big multi-part blog series on elections in the US. As you can see, that didn’t happen. I did find time to research all of the candidates and propositions, so I went ahead and voted this morning.

I’m really tired of the duopoly that we face each election. People are always saying “I’m voting for the lesser of two evils” or “if I vote for X, I’ll just be throwing my vote away.” Well, that’s because the voting system we have only allows us to specify our favorite candidate, and people feel like they have to vote strategically, possibly voting for someone who’s not the favorite to prevent someone that they really don’t like from winning. It doesn’t have to be that way though. There are other voting systems, Instant-runoff, Borda count, and the Condorcet method for example, that allow you to specify your preferences for as many of the options that you see fit. Instant-runoff voting is probably my favorite. It’s not perfect, but it’s pretty easy to understand and would be a huge improvement over the current system.

Another issue in American elections that I wanted to talk about is the Electoral College. There’s really two problems with how it’s used. The first is that it doesn’t provide equal representation for all citizens, and winner take all puts excess pressure on voters in swing states and makes voters in non-swing states feel like their vote doesn’t count.

The non-equal representation problem is caused by the way electoral votes are distributed. Wyoming only has about 500,000 people in it, but it has three electoral votes. California, on the other hand, has a population of about 34 million and gets 55 votes. That means a voter from Wyoming has almost four times the voting power as a voter from California. If California were to split into 68 mini-Californians, then a Californian’s voting power per person would be equal to that of someone from Wyoming. It just doesn’t make sense that changing some boundaries would have such a huge impact on the voting system. Unfortunately, to fix this unfair system we’d need to amend the constitution, and do do that we’d need support from many of the states that are benefiting from this inequity, so it’s not very likely to ever happen.

What we can do is move away from the winner take all notion to a proportional representation system. Two states currently split up their electoral votes to better match the will of the people. Switching to a proportional system can occur at the state level, so it’s much easier to get done. Of course, it doesn’t always benefit the two major parties, so there is going to be some push back before it can happen. California for example, is very likely to vote for a Democrat for president in a winner take all situation, so you’re not going to see many democrats in California pushing for proportional voting. Florida last election went the other way. We wouldn’t have needed all of the recounting anyway because out of their 27 electoral votes, only a few would have been contested, and it wouldn’t be enough to swing the election on its own.

Oh yeah… there’s one other major problem with the Electoral College that should be fixed, but probably won’t until it’s exposed to the general public. There are currently 538 electoral votes, and you need at least 270 to win. If we had a strong third (or fourth) party, it would be easy to imagine no candidate reaching the 270 vote mark. That would mean the election would be entirely in the hands of the House of Representatives.

That’s all for today. I’ll try to come back and flesh out some of these ideas later, but I can’t make any promises.


Last Sunday, Jesse and I bought a cat named Sergio (not much of a website yet). He’s about a year and a half old, and is a black and white short hair. So far, everyone is getting along really well, but he is still getting use to the place. We were planning on getting a kitten, but Valley Cats, the adoption agency, doesn’t like to adopt out single kittens, and we’re not ready for two cats. I’ll try to keep everyone up to date on how Sergio is doing, but you can remind me if I’m not.

Next Entries »