Politico Plagued by Rookie Mistakes!
These people--Mike Allen, Jim VandeHei, and Jonathan Martin--are supposed to be trained professional political observers. Yet they write:
Republicans Fear 2008 Meltdown - Politico.com: Even in a neutral political environment, 2008 would be a very tough year for Republicans, especially in the Senate. Simply put, the Republicans are defending more seats than the Democrats -- 21 Republican seats are up this cycle, while only 12 Democrats plan to seek reelection. So the playing field automatically is in the Democrats' favor...
By my count, 8 of those 21 Republican incumbents are from states that went for Bush by more than 20% in 2004; 14 are from states that went for Bush by more than 10% in 2004. By contrast, only 2 of those 12 Democratic incumbents come from states that went for Kerry by more than 20% in 2004; only 3 from states that went to Kerry by more than 10%; while 4 of the Democrats come from states that went for Bush by more than 10%.
In a neutral political environment, those 14 would be hard for the Republicans to lose.
In as neutral political environment, those 3 would be hard for the Democrats to keep.
In a neutral political environment, you might think that the more partisan states would split 18-3 for the Republicans in the Senate, leaving 12 battleground states.
In a neutral political environment, you would think that the battlegound states would split 6-6, making a total of 24-9 for the Republicans--a shift of 3 in the Republicans' favor from the current situation.
In a neutral political environment, you would expect the Senate to swing toward the Republicans in 2008.
Now I am not a trained professional political observer. A trained professional political observer might say that the prediction for a neutral political environment would not be 24-9 but would be 21-12--that Max Baucus and Tim Johnon are skilled politicians who have won in their respective states in the past in spite of the large presidential vote edge for Bush, and that the right thing to expect in a neutral political environment is that they will hold their seats (and also that the Democrats would hold Mary Landrieu's seat in Louisiana). I don't know.
I do know that "neutral political environment... Republicans are defending more seats... playing field automatically is in the Democrats' favor" is a gross insult to the intelligence of the Politico's readers.
UPDATE: Yep. Not only can't they do a simple swing count, but they have gotten under the covers with Matt Drudge. They're toast. Here's Glenn Greenwald:
Dear Mr. Greenwald,
I'm writing as someone who appreciates your writing and viewpoints. Your previous discussions on web traffic in regards to the Victory Caucus I felt were quite good. Regarding your current post in which you send questions to Politico, I feel I might be of some service to you. In my current occupation, I am a web metrics analyst. Through my company I have access to many reporting tools. One of them is Hitwise (you can check them out at www.hitwise.com.) One of your questions [to Mike Allen] I can easily answer:
(3) Do you know what percentage of The Politico's overall traffic is accounted for by Drudge links?
Drudge provides generally about 65% of all of the politico's traffic. The next highest website providing traffic to Politico is Google at about 3%. (See the attached file.)
It's rankings in the Politics grouping in Hitwise is very spiky, varying between number 60 and the number 4 spot in a jagged saw-tooth pattern (indicative of no natural audience). The spikes to high rankings coincide with Drudge traffic (see the daily spreadsheet.) Additionally, while they rank highly on days when Drudge links (and lowly on days when they Drudge does not), their metrics for pageviews are quite low (again, indicating no natural audience.)
If I can be of any assistance in this or further explorations of web metrics, please feel free to contact me.
Would like to help in any way.
[name withheld]
