Quinnipiac Has Scott (R) over Sink (D) for Florida Gov. on 9/23-28

[JCMach1]

Republican Rick Scott is leading Democrat Alex Sink in the race to become Florida's next governor, with Sink being hurt by President Barack Obama's unpopularity among Republican and independent voters, a poll released Friday suggests.


The Quinnipiac University poll shows Scott, a Naples businessman, favored by 49 percent of 1,151 voters to 43 percent who preferred Sink, the state's chief financial officer. The survey was taken between Sept. 23 and 28 and it claims a margin of error is plus or minus 2.9 percentage points. It was Quinnipiac's first screening of voters likely to cast ballots in the Nov. 2 election... http://www.gainesville.com/article/20101001/ARTICLES/101009970/1118?Title=Poll-GOP-s-Scott-leads-Sink-in-Fla-governor-race

[bluestateguy]

1. No more polls

nt

[Imajika]

2. You can't hide from polls....

...on a political discussion forum.

Reading about and discussing polls are one of the main things these forums are for.

[FBaggins]

4. Sure you can.

You just can't demand that everyone else facilitate that hiding.

[FBaggins]

3. No more posts demanding no more polls.

Just as likely to be obeyed (and just as deserving).

[stevenleser]

5. It's an outlier. Everyone else I have seen has Sink ahead n/t

[FBaggins]

7. It's not much of an outlier... but it's also not enough to panic about.

Actually four of the five most recent polls show Scott ahead. Two of them (rasmussen and this Quinippiac poll) tend to run a bit toward republicans this cycle, while the Mason Dixon poll (my favorite for state-level polling in this part of the country) retained the significant Sink lead.

Chances are that the two point Scott lead from Reuters and CNN/Time tell the current story. A very competitive race where either side could win.

I think that the deciding factor will be how many independents show up on election day (drawn by the 3-way Senate race).

[Freddie Stubbs]

12. Sink's internals have it at a dead heat

[FBaggins]

16. Which should be a concern, shouldn't it?

Just think of the types of scenarios that cause a candidate to release his own internal polling... and how infrequently that polling is anything but the most optomistic look at a race.

This race was close to locked up in Aug and early Sept... and now the best-case scenario is that we're tied?

I didn't expect Chiles' dropping out to be as positive a move as some here did... but not this bad.

[Freddie Stubbs]

17. It means that she is probably behind a bit

Fortunately, there is still a lot of time for her turn things around.

[FBaggins]

18. Yeah... but I can't figure out what went wrong.

I'd say that it was because Dick Morris essentially endorsed her today... but that wouldn't show up in these polls. :rofl:

[BluegrassDem]

Quinnipiac has a strong Republican tilt this cycle

Chuck Todd even talked about how Republicans all have huge leads in their polls. Either their likely voter model they use is too stringent, or they're picking up on something other pollsters aren't seeing.

[Jennicut]

10. Yeah, makes me want to march down to Quinnipiac U in Hamden and question them on it.

I live in Connecticut and have been to Quinnipiac a few times. They are usually not this biased.

[FBaggins]

14. It may not be bias... it may be a process error.

Polling firms have to make some assumptions about the shape of the electorate. Almost every poll weights their results by comparing the sample's proportion of republicans/democrats, men/women, etc. to demographic norms for the state. Some of those are comparatively fixed and/or easily identifiable from census data... but some of them are in flux from one election to the next.

When poll after poll shows a larger number of people self-identifying as republican, they may decide that this represents an actual shift in demographics and end up weighting to a different norm than the majority of polling firms that use prior year's turnout data and don't adjust what they weight to.

This is the very error that caused Rasmussen to go so far afield in their first national year (and they admitted it). The electorate was much closer to historical norms than the raw polling data had indicated and a high percentage of the rasmussen polls were overly optimistic for republicans.

I don't know if Quinnipiac is making the same error (or even if it is an error this year), but it's possible. It's also possible to do so without "bias" if it is what their raw survey data reflects.

That's why you can't pay attention to any one polling firm.

[BluegrassDem]

6. Quinnipiac has a strong Republican tilt this cycle

Chuck Todd even talked about how Republicans all have huge leads in their polls. Either their likely voter model they use is too stringent, or they're picking up on something other pollsters aren't seeing.

[book_worm]

8. Quinnipiac also had Scott losing the primary to McCollum...

[Kdillard]

11. Hmm that is good to know but I am concerned that 4 out of 5 polls show him leading when she was

doing so well earlier.

[Freddie Stubbs]

13. The problem is they didn't poll for McCallister

It's a good bet that at least 30,000 of his votes were people who thought they were voting for McCollum.

[Kdillard]

9. Hmm we'll see. I hope this is wrong.

[DCBob]

15. I hope this is wrong but Florida is no doubt tough state for Democrats..

Just ask Mr Meek.