Revisiting my mistakes

As covered in my election post mortem post, I miscalled 7 ridings on election night.

Some were honest mistakes:

• Saint John Portland was listed as too close to call so I don't feel too bad about it. It stayed with the incumbent by 277 votes while I thought it would go to the challenger by a small margin. This was a case of collapsing NDP vote not going to the Liberals but instead to the Tories which we saw in a few places around the province.


• Dieppe Centre-Lewisville I said I was confident would go Liberal. Despite this, it stayed Tory by the smallest margin of the night: 59 votes. This seems to have been due to the Tory areas of Lewisville joining the riding but we'll have to wait for poll-by-poll results to know for sure.


• Charlotte-Campobello was another I was confident with but maybe I shouldn't have been, I always expected this to be a close race but I was sure it would be close in favour of the Liberal. It stayed PC by 282 votes.


• Southwest Miramichi I had listed as somewhat uneasy. This was a seat the Tories had to pick up to win and when it started to go Liberal on election night, I knew my prediction of a Liberal victory had to be right. As I mentioned both in the post mortem and in early comments, it is a "small c" conservative riding, but Rick Brewer, a former minister, plays well here and has been a great constituency MLA.

The other three ridings however is where I really missed the boat.

Centre-Péninsule-Saint-Saveur

This riding was listed as "too close to call" with my call being a PC gain. This was because I thought Louis-Philippe McGraw was a strong candidate. My own advice to myself was that McGraw had been out of the riding since he lost it so that would hurt him. Also Peninsula hospital downgrades had outraged people and sent them fleeing from the Tories but, I thought, the Liberal failure to promise to reverse the cuts, contradicting comments by Denis Landry and Hédard Albert, could give McGraw a boost.

I couldn't have been more wrong.

The final result? Landry 56%, McGraw 44%.

A win by over 1000 votes and 12% is not close. Not even a little bit. I really misread this riding.

Woodstock

Here was a real shocker. I can't wait until the poll-by-poll results become available here so I can better understand what happened.

While I thought I knew what I was talking about in Centre-Péninsule, I don't really know the area well, don't know any people from the riding and was doing a lot of guessing. Woodstock was a different case.

I said I was "99% sure of the outcome" in Woodstock. Quite frankly, I almost put it under the heading "I am more sure of the fate of this riding than of my age, name and sexual orientation". Fortunately I did not!! A number of other bloggers were calling Woodstock PC so, despite my certainty, I didn't quite put it in my top category. Thank God.

I know Woodstock. My father was born there. I have family there. I have friends there. I have been there hundreds of times.

Now, this riding is a lot more than just the Town of Woodstock but I thought I had taken that into account. The Town is the very north end (pretty much) of the riding while there are lots of rural areas to the south. The town and the few points north were outraged by the Valley hospital restructuring and were sure to vote Liberal. I guessed they would vote 80% Liberal, we'll have to see the poll-by-polls, but obviously this was wrong. Not only did David Alward win, he won huge. He won by 13% of the vote.

I did not see this coming. Quite frankly, I think if I had the choice between choosing 55 seats for the Tories and the Tories winning Woodstock by 13%, I would have chosen the former.

Clearly three things have happened here:

a) people in Woodstock aren't as angry as I thought
b) the population of Woodstock (town) must be a fair smaller fraction of that of the whole riding than I though
c) David Alward must be a far better constituency man than I gave him credit for

Fredericton-Silverwood

This, ladies and gentlemen, was the upset of the night.

I shifted this back and forth and back and forth between the Tories and Too Close To Call and finally unded up with it in the NDP column. It was only once marked as Liberal and that was in my preliminary projection which bore no resemblance to my final call.

I made a lot of mistakes here. First, let's revisit the facts as I understood them.

Brad Green was vunerable. He won in 2003 by a margin of 40% over 38% over one of the weakest candidates any party has ever put forward by the Liberals. Rick Miles was an unknown, but was certainly a better candidate than the Liberals put forward in 2003 and should have had the edge.

After that is where my thinking started to falter.

It was my belief that the NDP were a big challenge to a Liberal victory here. They had 23% in 2003, and though redistribution hurt them, they still had over 20% of the vote under the new boundaries. They had a better candidate in 2006 than they had in 2003. They had their leader running next door. These things told me that the NDP would at least hold their vote if not grow from there.

Therefore, if the NDP vote was going to grow, a Liberal victory seemed impossible. It was Dennis Atchison's NDP candidacy which first caused me to move this into the leans PC column.

The Atchison campaign ran radio ads, something impressive for an NDP campaign, and bloggers seemed to be saying that this was a race between Green and Atchison so, I jumped on the bandwagon.

At the end, though I wasn't really to confident to see this go NDP (I put it in the "ridings to watch"/"too close to call" category), I did call it that way. In my mind I thought that the NDP was going to pick up a riding somewhere, somehow, for sure. The best chance I could see on the map was Fredericton-Silverwood, so I made the call.

The final result was Rick Miles 47%, Brad Green 41%, Dennis Atchison 12%.

As you can see Atchison was not even in this race. If I had known that, I would probably have called it safe Liberal as we have established that Brad Green was weak and in trouble.

Where did I go wrong?

1.) I thought Miles was a weak candidate. He was an unknown who I'd never met. The only thing I saw of him was the CBC TV profile on Fredericton-Silverwood in which he and Dennis Atchison had an argument about the merits of voting NDP. Atchison seemed relatively smooth while Miles performance was not impressive. I thought that this was an encounter CBC had staged, but it turns out it was a matter of two candidates being in the same place in the same time, cameras appearing and a reporter goading them for sparks. Miles performance here was not great, but I thought it was indicative of his overall performance. I was wrong.

2.) I thought Atchison had become "the" opponent for Brad Green. As I've said over and over, Brad Green was very vunerable and I thought all along he was likely to lose his seat. A strong NDP candidate was a saving grace for Green. An NDP candidate who gained the momentum could have done the unseating. However, it is clear that the consensus around Atchison did not exist, it in fact collased around Miles who, as the Liberal candidate, was natural.

3.) I thought Rick Miles was without experience. He, as I've discussed, was a newcomer to politics. What I didn't know was that his campaign was being managed by none other than Jim Mockler. Mockler was the campaign manager for the Liberals in Fredericton South in the 1998 by-election and the 1999 election. He knows the riding and he knows how to run a campaign. From 2000-2003, Mockler was on the provincial executive of the party, from 2002-2003 as president. Though the candidate lacked experience, his team did not and this can, as we saw, make all of the difference in the world.

I thought the Liberals had 30 seats without this one in their tally or even on their radar. They won 29. The differnce between 28 seats and 29 seats is the difference between a weak and a stable majority. The Liberals are very lucky I was wrong with this riding or else they would be presiding over the same sort of weak government that Lord had from 2003-2006 and might have been back to the polls long before 2010. Thanks to Rick Miles and his campaign team, they've probably got four years to govern.