Wednesday, October 27, 2004
The 2004 Election Sweepstakes
Someone has to do it, and since our blogger-in-chief has bigger things on his mind for the rest of the week, I'll take a crack at this.
Following up on the "success" of our Democrat Primary Pick'em (which I won, incidentally - link to follow), it's now time for the main event in our World Series of Politics.
The Bipartisan Commision on Political Games hasn't had much time to meet, but I'd like to offer the following "game questions" as elements of the contest. Game scoring will be finalized as quickly as possible, but will likely involve a fixed point value per questions, with deductions made for the absolute value of the difference in your prediction and the actual result. The questions that are higher on the list will be worth more than the ones farther down (values will be settled soon) and the question list itself may undergo some changes.
Here's the first cut:
1. The electoral vote count. (The big question.)
2. The number of states won by each candidate. (Washington, D.C. counts as a state for the purpose of this question.)
3. The national popular vote tally, by percent, for the top two candidates. (One example of an acceptable answer would be 49/Kerry - 48/Bush. You need not include a number for Nader, since only insane people are voting for him anyway.)
For the moment (until I am prevailed upon that it doesn't matter), associate a candidate with each number in the previous questions. That's in case we have some strange and totally outrageous situation where the candidate who wins the national popular vote wins fewer states and fewer electoral votes. Don't ask me how that could happen...um...again. So really, there are six questions being asked above, and for the purposes of scoring, they will probably be scored separately.
4. The winner of the election. (That's all or nothing, of course. It would have been closer to question 1, but it's not going to be worth as much as the questions that precede it.)
5. Kerry's popular vote percentage in Texas.
6. Bush's popular vote percentage in Massachusetts.
7. The number of states in which Nader polls more than 2 percent.
Once the questions and scoring have been made official (in this blog post, or in another), entries may be made as comments to the official posting.
Happy predicting! Remember to vote!
Update:
Blogger ate my instant update, so I will post both at once. I do much of my best thinking in the shower, so I've settled a few scoring elements, I think.
It occurred to me that weighted ordinal scoring might do the trick. For each question, you get a certain number of points for each person you beat on that question. The more important questions will carry a higher-than-unity weighting, while the less important questions will be worth one point per place (or less)
Upon additional consideration, I've decided that the first question really is only one question, as the electoral-vote count is going to be a two-way split. This is barring the unforeseen, such as Ralph Nader getting the last details of his contract with Satan finalized or completing the construction of his mind-control machine.
I'll have more to say as I think about it.
|
Following up on the "success" of our Democrat Primary Pick'em (which I won, incidentally - link to follow), it's now time for the main event in our World Series of Politics.
The Bipartisan Commision on Political Games hasn't had much time to meet, but I'd like to offer the following "game questions" as elements of the contest. Game scoring will be finalized as quickly as possible, but will likely involve a fixed point value per questions, with deductions made for the absolute value of the difference in your prediction and the actual result. The questions that are higher on the list will be worth more than the ones farther down (values will be settled soon) and the question list itself may undergo some changes.
Here's the first cut:
1. The electoral vote count. (The big question.)
2. The number of states won by each candidate. (Washington, D.C. counts as a state for the purpose of this question.)
3. The national popular vote tally, by percent, for the top two candidates. (One example of an acceptable answer would be 49/Kerry - 48/Bush. You need not include a number for Nader, since only insane people are voting for him anyway.)
For the moment (until I am prevailed upon that it doesn't matter), associate a candidate with each number in the previous questions. That's in case we have some strange and totally outrageous situation where the candidate who wins the national popular vote wins fewer states and fewer electoral votes. Don't ask me how that could happen...um...again. So really, there are six questions being asked above, and for the purposes of scoring, they will probably be scored separately.
4. The winner of the election. (That's all or nothing, of course. It would have been closer to question 1, but it's not going to be worth as much as the questions that precede it.)
5. Kerry's popular vote percentage in Texas.
6. Bush's popular vote percentage in Massachusetts.
7. The number of states in which Nader polls more than 2 percent.
Once the questions and scoring have been made official (in this blog post, or in another), entries may be made as comments to the official posting.
Happy predicting! Remember to vote!
Update:
Blogger ate my instant update, so I will post both at once. I do much of my best thinking in the shower, so I've settled a few scoring elements, I think.
It occurred to me that weighted ordinal scoring might do the trick. For each question, you get a certain number of points for each person you beat on that question. The more important questions will carry a higher-than-unity weighting, while the less important questions will be worth one point per place (or less)
Upon additional consideration, I've decided that the first question really is only one question, as the electoral-vote count is going to be a two-way split. This is barring the unforeseen, such as Ralph Nader getting the last details of his contract with Satan finalized or completing the construction of his mind-control machine.
Aside:I'm already starting to vacilate, but for the moment, given the two way split, you're going to be off by the same amount on both candidates regardless of how you pick. So the first question might be changed to "The number of electoral votes won by the winner," and will be worth the most ordinally. Question 4 will force you to identify your winner all-or-nothing, so you pick an EV number that you think the winner is likely to gain (and this could itself depend on who you think will win) and then collect your "bonus" points for picking the winner.
I don't know much about contract-law, but if anyone could force a consumer-friendly compact on the Prince of Darkness, it'd be old Ralph. As for mind-control, I originally thought that if you happened to see a bunch of shambling, mottled, glassy-eyed zombies at your polling place, groaning "Greeeeeeeeeeens!" or something like that, then watch out. Alas, that applies far better to the last election. Maybe pod-people works better for this one...these pod people, and not those (Scroll down a bit.)
I'll have more to say as I think about it.
|
