Huh????? Then how is his prediction a prediction?????
What he's saying is that the math allows him to predict a 55 seat gain for the GOP, but he believes that the validated methodology he's using understates what experience tells him will really happen. That's my position as well. There are several effects that by definition cannot be modeled mathematically, and they are going to be bigger factors this year than usual. Unmodeled effects include the coattail effect mentioned above, the time zone effect, where the results from polls that have closed with affect voters further west (the Dems tried this with the Florida Panhandle in the Bush v. Gore election), and the level of fraud. I won't go into the rest of the discussion, but the point is that honest pollsters know that this year is going to be worth watching IF FREEPERS AND OTHER REAL AMERICANS VOTE AND DON'T STAY HOME OUT OF FALSE CONFIDENCE!
“Then how is his prediction a prediction?????”
He’s like the weatherman hedging his bets: “My instruments and models tell me to expect a slight drizzle, but my gut tells me it will be sunshine and blue skies all day.” No matter what happens, you can never accuse him of being wrong. The marvel is that his students each pay thousands a semester in tuition just to hear this nonsense.
“55” seems to be the safe bet among prognosticators now. Anything beyond that is gilding the lily. Anything less would be cast as a losing day for the Republicans, even tho it’s hard to imagine they won’t take the House... now it’s all about the margin.