As I mentioned in my commentary on earlier CRA poll, their first poll was not very reliable because it was conducted over a period of two weeks in the midst of an election campaign where voters opinions are very volatile.
Some commmenters disagreed with my view. They cried foul that I would criticize the CRA poll due to their impressive results in 2003 (they predicted Lord would win by 1 point and he did). Here are some of the comments:
- "Don Mills at CRA predicted the 2003 election almost to a tee. Their methodology, I think, is very sound." - David Campbell
- "CRA's methodology must be sound as they predicted the 2003 election successfully." - scott
Well now ladies and gentlemen we have a CRA poll with the same methodology as the 2003 election. It was conducted at a similar point during the campaign, over a similar time period, with a reasonable sample (665). It shows the Liberals will win by 2%.
Therefore if CRA's track record from 2003, the opinion of David Campbell and the opinion of scott are any indication, the Liberals will win on September 18 by a margin of 2%.
Congrats to Mr. Graham and the Liberal Team ;)
(Hat tip: Alvy Singer)
Posts
13 comments:
My prediction stands:
Liberal - 32 minimum
Cons - 23 maximum
NDP - 0
However, it is possible NDP may pick up one seat either in Fredericton-Silverwood or Tantarmar ridings which will put Cons at 22.
I think, ironically, if the Conservatives win, they will lose Fredericton-Silverwood but, if they lose, the Liberal vote will go up at the expense of the NDP and the Tories will hold on to Fredericton-Silverwood.
Right now I am seeing 35 Lib, PC 19, NDP 1... but it is way too early to be saying such things.
35 seats! Ha, I'd like to smoke whatever is in your pipe. Give your head a shake. It's one poll, and with the margin of error it's a dead heat. If you're so confident about the majority seats going Liberal, why don't you start naming them! It's easy to throw out numbers, but actual riding names you will have difficulty! 35 seats is crazy!
The 29 seats I currently have listed a safe or leaning Liberal + Campbellton, Miramichi Centre, Southwest Miramichi, Kent South, Moncton West, Kings East and York would make it 37.
I think that is conceivable.
My feeling right now is 35. I will make a final projection on Sunday which, depending on how things go, could have the Liberals as low as 20 or so.
Has Scott gone anonymous at 3:27 PM LOL.
NBPolitico, you are within reason.
As I said over at David Campbell's blog, though one may claim that CRA is more accurate than Omni when it comes methodology, the average New Brunswick can not make hide nor hair of samples, methodology and polling accuracy. All they may know is what they read in bold print in the MSM at the top of the page. (like the bus photo last saturday in the Times & Transcript)
And why the CRA poll was buried without detail in L'Acadie Nouvelle makes no sense to me.
As well, I don't put too much weight in polls because in a tight election like this one, things can swing quickly from one side to the other very quickly. Remember, the Omnifacts poll is coming out on friday, so if you believe these things sway public opinion and the undecided than probably the last one to come out will be the one on voters minds throughout the weekend and onto election night.
As the great Dief once said, "polls are for dogs." ---or Liberals. LOl I agree with him wholeheartedly as I try not to pay attention to them. To me, the only poll that counts is on election night.
"And why the CRA poll was buried without detail in L'Acadie Nouvelle makes no sense to me."?
Where does taking the entire front page equal buried? Maybe Scott didn't actually look at the paper today.
What is the total readership of L'Acadie Nouvelle? Because you know not many New Brunswickers are on the internet compared to other provinces, so many couldn't read it online. I guess the CRA crowd were hoping for a big radio truck audience to get their point across. We'll see if they blow it up tonight on the eveing news. Knowing the CBC, they'll be all over it. Yet, it's hard to do two things at once, so i guess they'll have no choice but to pull their nose out from under the Liberal ass it was in.
"CRA's methodology must be sound as they predicted the 2003 election successfully." - scott
Scott you have changed the tune now, sir.
Believe it or not Bernie is done. Burn Bernie burn, with envy that is.
scott - we shall see, I expect the other papers will cover the poll tomorrow.
Anon: We'll wake up my friend, can you not count. Hindsight is 20/20.
I still stand by my comments regarding their 2003 prediction, however, I do not put any weight in polling companies to predict an election everytime as many were saying the same thing as you and I were with regards to Strategic Counsel and Alan Greggs flawless methodology for the 2006 general election. We all know what happened to their numbers as Nik Nanos and SES's rolling poll won the day. Who knows, maybe the omnifacts methodology will prove us wrong.
Again, boys, the bottom line is I don't put weight in polls. Sometimes they are right and sometimes they are wrong. There is considerable margin of error. That is why there was considerable error with Dithers and the Liberals as they were governing by them.
nbpolitico: your dead on buddy. They will release the CRA poll tomorrow in the english dailies, well at least the TJ. (embargoed release) However, again I stress, Joe Public on the street could care less about polls and most of the time can't make sense of them, other than they know 45 is bigger than 43. LOL Let's see what more polling stuff we can waste are time with on friday. Cheers.
scott - I agree, polls are not that relevant to the average voter. It is interesting to see though that from Friday to Sunday a scientifically valid sample showed that the Liberals were leading.
Post a Comment