Pollster.com

Articles and Analysis

 

New Senators 'Outliers'

Topics: Outliers Feature

Joe Lenski notes that 2010 will bring more new senators (at least 15) than any election since 1980.

Nate Silver's new Senate model shows Democrats losing 6-7 seats; Megan McArdle thinks it could be too conservative.

Alexander Burns declares robo-polling the winner in Florida's primary.

Peter J. Woolley and Dan Cassino write up their experiment on polling on independent Chris Daggett in NJ.

Survey Practice publishes its August issue.

 

Comments
CompCon:

I have to agree with Megan MacArdle. I think with every passing day that shows the stimulus has failed, each unemployment report that shows there is no recovery, and as the economy stagnates more before the elections, it will push the democrats down further and they will lose IL, WA, WI, and CA. All 4 of those states are within the margin of error now - and only one of the democrats is above 49% (Feingold at 50%).

Nate says there is a 20% chance that the democrats lose the senate. I'd put it at about 50% right now.

Nate said, "And so far, the Democrats’ strategy of Bush-bashing does not seem to be resonating in these states. "

Good! Could we finally move on and start campaiging in 2010 instead of 2006?

If Nate says a 6 - 7 seat loss, the democrats are extremely close to losing it. The republicans would have to pick up 3 of the 4 toss-up seats but the momentum is definitely on the side of the GOP.

____________________

CompCon:

Another note on Nate Silver - here is a quote from him that shows he thinks he is now in control of all polls nationwide:

"Our pollster ratings are always becoming more sophisticated and we’re going to be looking at appropriate ways to punish pollsters who dodge putting their necks on the line."

So now Nate admits that he is going to "punish" pollsters based on whatever random criteria he feels like? That is a level of arrogance that is at a new high for the self-proclaimed "king of polls".

After his pollster "ranking" disaster, he doesn't have a tremendous amount of credibility left and now that he declares his intent to "punish" pollsters, there is even less now. He's turning into a thug.

____________________

boomer40:

The pollster ranking was not a "disaster" by any means. And I'm not sure why we should listen to Megan McArdle on ANYTHING, let alone polling, something in which she has even less expertise than her supposed specialty, economics.

____________________

CompCon:

I guess calling it a disaster or not is in the eyes of the beholder. Until his controversial decision to apply a heavy weighting factor to his rankings that can't be substianated with any actual data, he was viewed as someone who simply collected and reported data. After that, he is seen by many as manipulating data based on his personal gut feeling.

Now he is talking about "punishing pollsters" who don't comply with some arbitrary rules he just made up.

He is rewarding pollsters based on his opinion of whether they are more accurate or not and will punish those that don't bow to his wishes. That destroys what credibility he had left. He isn't interested in reporting polling - he is interested in controlling polling.

____________________

boomer40:

"Until his controversial decision to apply a heavy weighting factor to his rankings that can't be substianated with any actual data, he was viewed as someone who simply collected and reported data."

He has always done a bunch of weird and arbitrary stuff with his rankings and predictions. So does every organization tied to polling, whether it be the Pollster blog (what to report?), the Pollster aggregator (what to include?), and polling firms themselves (whom to call? How to weight?). If you think he has suddenly jumped the shark and fundamentally changed his way of doing business you haven't been paying attention.

____________________

CompCon:

The ranking controversy was a little different. Where his weird and unecessarily complex calculations and analysis usually could be defended to some extent by the data collected, this weighting factor was something he simply made up based on what he thinks might make a pollster more accurate. He couldn't defend it and quit trying.

I've been paying attention to fivethirtyeight for 4 years now and the ranking flap hurt his credibility a lot. Going to the New York Times only makes a few liberals comfortable and now his threat to punish pollsters leads me to believe that his head has gotten much too big for his shoulders. He isn't the ultimate authority on polls that he thinks he is.

____________________

sjt22:

Now he is talking about "punishing pollsters" who don't comply with some arbitrary rules he just made up.

Its his rankings, if you don't like them then don't pay attention. In any metric there are factors which reward or punish certain inputs and actions. The idea that this is some power play by Silver is asinine, he's just refining his model.

____________________



Post a comment




Please be patient while your comment posts - sometimes it takes a minute or two. To check your comment, please wait 60 seconds and click your browser's refresh button. Note that comments with three or more hyperlinks will be held for approval.

MAP - US, AL, AK, AZ, AR, CA, CO, CT, DE, FL, GA, HI, ID, IL, IN, IA, KS, KY, LA, ME, MD, MA, MI, MN, MS, MO, MT, NE, NV, NH, NJ, NM, NY, NC, ND, OH, OK, OR, PA, RI, SC, SD, TN, TX, UT, VT, VA, WA, WV, WI, WY, PR