A Night For Numbers

I’m in a very mathematic mood – if such a thing is possible. I got home late after Monday Night Football, sat down and played some poker online. I haven’t written about poker too much lately. Maybe that’s because of how poorly I’ve been doing for nearly two months.

Some of it is bad luck, but the majority is bad play. I see the trend, which is too much aggression on marginally winning hands. If you go all in four times and win only three, you’re gone. I have to be more conservative in that way. Keep my neck off the chopping block.

I have moved down in stakes and reined myself in. So far I’ve done OK against lesser competition. We’re still up since August 2003, but much of our winnings have been squandered by me.

I have to maintain discipline. I can’t play on tilt. Bluffing is a good plot technique in a novel, but a losing strategy in real life poker.

Like I said, I was really into numbers tonight. There was poker and before that my new found infatuation with the ridiculous traffic this site had on Monday. With more traditional, higher ranked sites now on the Ashlee Simpson story, I’ll soon be relegated to the third and fourth page in the Google results and my traffic will tail off.

My final numbers play was looking at the latest election polls. It is too late to look at the popular vote. Analyzing raw numbers is a fool’s game since it isn’t how we elect a president anyway.

I looked at state by state polls on the three sites I’ve grown to enjoy for this: The New York Times, Slate&#185, and my new discovery RealClearPolitics.com.

I love thumbing through the charts and maps on each of the sites and reading their analysis. This is definitely like predicting the weather… actually predicting a snowstorm. I say that because predicting snow is inherently difficult. There are parameters that interact with each other and the data is never as complete or as well initialized as you’d like.

The wild card in this election is voter turnout. Most of the major polling companies limit their surveys to likely voters, and they are qualified based on historical criteria. It seems to me, and this is gut not science, that the turnout for this election will be higher than historical norms. That would mean there will be more voters than the surveys take into account. Will those additional voters vote the same way as the likely voters surveyed?

If the election does draw a heavy turnout, will lines or delays at the polls send people home without casting a ballot? Will those people correspond proportionally to the survey results?

I don’t know. But, it stands to reason, the more unknown variables that are thrown in, the less likely it is that the election will be accurately called.

Just as each individual forecast has a separate degree of difficulty, so too do elections. This one is incredibly tough to call, but is fascinating to look at piece-by-piece. And, unfortunately, just because I have lots of pieces to look at doesn’t mean I will understand any more.

&#185 – Last week I wrote about Slate’s state-by-state polls showing Kerry ahead. Tonight that is reversed with President Bush leading 276 to 262.

Leave a Reply

Your email address will not be published. Required fields are marked *