Wednesday, February 23, 2011

Current 1st Count Estimate and how I got there

Kevin Falcon - 2975
Christy Clark - 2600
George Abbott - 2100
Mike de Jong - 825

1) There were 35,000 existing members of which I am assuming that 27,000 will vote.   Of that number, I am assuming that any riding with a sitting MLA that is backing an candidate will have that candidate win the riding, I am modifying their result based on how large the membership was in November, the higher the number of members, the higher percentage of the vote they take.   George Abbott and Kevin Falcon both have 20 ridings and Christy Clark and Mike de Jong have one riding each.

I further modify the results using the past MLAs that are backing people, I given each former MLA a mild impact on their former riding unless they are well known.

I assign a regional bonus for each candidate - ie Mike de Jong gets a bonus in Abbottsford and Chilliwack etc.  George Abbott gets general bonus in all of rural BC.

The candidates then have a notoriety bonus - Christy Clark has the highest, but it falls outside of the range of CKNW.  Kevin Falcon and George Abbott have a mild effect and Mike de Jong none.

Existing members are the ones most likely to change their mind in the campaign and therefore how well the candidates do the on the ground campaign matters.

With all that in consideration, I get to George Abbott - 44.7%, Kevin Falcon 33.3%, Christy Clark 18.4% , Mike de Jong 3.6%

2) New members - there are 55,000 of them.   I am assuming 5000 are lapsed members like myself and another 3000 are spontaneousness new members.   I am treating this 8000 in the same way as the existing members.   I am also assuming they are concentrated more in Liberal held ridings than NDP held ridings.   I also assume that 90% of new members will vote.  Adding them to the existing members means that pool is now 34,000

The other 47,000 are divided as follows

Clark - 18,000
Falcon - 15,000
de Jong - 11,000
Abbott - 3000

I know that the top three campaigns have some clumping of votes with a couple Surrey ridings sucking in a lot of new members.   I am not certain about the Abbott numbers, but based on some anecdotes I think I am in the ballpark.   I am assuming the Abbott sign ups are more rural than urban and more likely to be in ridings held by MLAs supporting him.

The 47,000 mass sign ups will have about 42,000 vote, or be about 55% of the total vote.

Once I crunch all these numbers, I get the result of: (note the error range is larger now because I have made more assumptions that I do not have clear data on)
  1. Falcon - 2975 (2750 - 3200)
  2. Clark - 2600 (2400 - 2800)
  3. Abbott - 2100 (1900 - 2300)
  4. de Jong - 825 (700 - 950)
I will keep refining the data as I get more information.

At the moment I am 80% confident that Kevin Falcon will win, 15% George Abbott and 5% Christy Clark.   I see very little space for Christy Clark to gain many votes from Abbott supports if he drops off and that means she can not catch Falcon.   My new calculations give Abbott an outside chance to move ahead of Clark on the second count and then win on the final count.

4 comments:

Anonymous said...

You should take that list of former MLAs who support different candidates and give a bonus for the ones who were the last elected person from the formerly held ridings.

Everyone agrees, those current NDP ridings will be just as important and they used to be BC Liberal.

Bernard said...

I did something more or less like that with the former MLAs

Anonymous said...

The face that you are the only person in the province with Falcon in front makes this suspect. Even Clark Abbott and Falcon all agree Clark is ahead.

Bernard said...

I keep doing the math over and over again and with no caucus support and her new signs ups being less than half the new members, I can not make the spreadsheet math work to put her in the lead.

I have yet to see anyone explain how they think she is in the lead, not one has laid out an argument that makes sense