Due to my inability to keep to the old schedule, I'm writing a new one.
First off, I should explain a bit more about my cat (Boo) and what has happened to him.
In March, my Roomate (whom I shall call Kris) noticed that Boo was peeing a lot, much more than usual. She decided to test his urine and it came back positive for sugar - something not normal. We called a vet and the vet tested Boo's blood, and turns out he has Diabetes. We gave him insulin, tested multiple times a day, and switched his diet. He went into remission in June
.
In late July, Kris and I noticed that Boo was throwing up at about the same time every morning. We could not figure out why. A week or so later, Kris noticed Boo not acting normal, and we again called the vet who ran a number of tests. We were referred to another vet, and from there to another. We eventually found out that Boo has a combo of Irritable Bowel Disease, Fatty Liver Disease, Pancreatitis, and inflamed intestines. This is a very bad combination, as the treatment for any one of them (pain killers, for the Pancreatitis, for example) makes another worse (constipation) and other drugs can't be processed due to the Liver problems.
We had a feeding tube put in, and since then, have been feeding Boo though the tube using a syringe. The process of feeding takes about 15 minutes and Boo is sometimes resistant. In addition, he throws up once or twice a day, and can only eat so much at one time. Due to possible feeding tube damage when he vomits and the need to get a minimum of food in to him so he can physically survive, this means he needs to be monitored 24 hours a day.
Kris and I have been doing so. More her than I actually, as she knows much more about medicine, and I have a bad habit of passing off work to others when I can. We've agreed that we would take "shifts" of 10 to 10. From 10pm to 10am, I am "on duty". The idea is for me to be awake during these hours, and sleeping during the remainder. Since I only need 8 hours sleep, the exact timings are flexible. We may change these timings (for example, to 6 to 6, with me still on the "night shift") but in general, I'm awake at night.
At least, that is how things are supposed to go.
I've been feeling ill recently. I finally got a family doctor, the first in my adult life. He told me that my Asthma was flaring up. The difficulty in breathing was making me sleep much more than 8 hours (due to poor quality sleep) and this put stress on Kris, which in turn, but stress on me (I was not choosing to sleep all day long) The new medication that the doctor gave me however seems to be having the opposite effect, in that now I can not get to sleep. Due to these things, my sleeping has been all over the place, and when awake, I've been having difficulty concentrating. Now that the election is over, and Kris begins at university, our schedules are getting more set, and I will hopefully be able to get into a normal routine.
The schedule for the new few days is therefore "get into a routine". This is not very good for posting on a blog however. Therefore I will commit to the following.
My first goal is to finish gathering and sorting the data from the election. Once that is done, I will examine where I went wrong, and compare my results to that of other projectors. None of us did very well, but I may have done worse than average. I will examine the reasons why.
I'll take the real data and I will examine it in comparison to my projection. My goal here will be to find ways to adjust the model to better match the results. I suspect that I can do this somewhat easily by taking a third of the ADQ vote and applying it to the PLQ.
Sometime before the Ontario By-Elections I will make a post about it. I will also live-tweet the results.
Once all that is done, I will have to make a serious change to the ElectoMatic, the most serious change ever made. In the post explaining it, I will describe the "Ratio" system I use, and how it is much better than the "Flat" system used by others. I will also explain how I will add a new "Swing" method to the mix.
Once that is complete, I will take the new model out for a few test runs, putting in past election results (IE 1979 federal) and past result popular vote numbers (IE 1980 federal) and seeing how well it works. I will continue to tweak it until I get it to work as close to reality as possible. I will also focus on trying to estimate huge swings (1980 to 1984 federally in Quebec, or 1988 to 1993 in Western Canada) and anything else I think could be useful.
After that, we'll have to see what is going on in politics!
Wednesday, September 5, 2012
Brian Pallister elected as Manitoba Leader of the Opposition
Pallister, Manitoba PC leader, has won a by-election, becoming the new Leader of the Opposition.
http://electionsmanitoba.ca/2012Fortwhyte/results/division.asp
http://electionsmanitoba.ca/2012Fortwhyte/results/division.asp
Analysis to continue
I've woken up feeling refreshed and will continue my analysis shortly. I will also focus on why the model failed (in short, ADQ voters voting Liberal) and why I failed (I saw the polls indicating this would happen but did not want to believe them)
Yet another delay
I can't get my mind off the assassination attempt, and am having trouble focusing on the numbers. I've also been very sleep deprived recently, and that is not helping. Regardless, the numbers I have right now are as follows.
Northern Quebec
PQ - 107,261 - 48.88%
PLQ - 51,188 - 23.33%
CAQ - 45,696 - 20.82%
QS - 9,025 - 4.11%
ON - 3,637 - 1.66%
PVQ - 276 - 0.13%
Aut - 2,371 - 1.08%
I also have a map showing which regions I am using.
Northern Quebec
PQ - 107,261 - 48.88%
PLQ - 51,188 - 23.33%
CAQ - 45,696 - 20.82%
QS - 9,025 - 4.11%
ON - 3,637 - 1.66%
PVQ - 276 - 0.13%
Aut - 2,371 - 1.08%
I also have a map showing which regions I am using.
Marois Assassination attempt?
As the new premier-elect was finishing her speech, a security person rushed her off the stage suddenly.
2 minutes later, some PQ operatives came on stage, saying that they had to protect her as something happened. He then asked everyone to leave.
1 minute or so later, Marois came back and said the same thing. She mentioned a sound grenade.
3 minutes later there is news footage of a man lying on the pavement and a rifle.
What
The
Hell
I've never seen anything like this before and probably never will again.
If this was an assassination attempt I have one thing to say
I hate the PQ, and Marois. I think they are potentially dangerous.
But they won.
They won the election, fair and square.
No matter how much you hate or fear the other guy, you can't try to stop them by force.
Democracy is the cornerstone of our freedom, and we must respect it, even when it tells us the things we don't want to hear.
2 minutes later, some PQ operatives came on stage, saying that they had to protect her as something happened. He then asked everyone to leave.
1 minute or so later, Marois came back and said the same thing. She mentioned a sound grenade.
3 minutes later there is news footage of a man lying on the pavement and a rifle.
What
The
Hell
I've never seen anything like this before and probably never will again.
If this was an assassination attempt I have one thing to say
I hate the PQ, and Marois. I think they are potentially dangerous.
But they won.
They won the election, fair and square.
No matter how much you hate or fear the other guy, you can't try to stop them by force.
Democracy is the cornerstone of our freedom, and we must respect it, even when it tells us the things we don't want to hear.
Tuesday, September 4, 2012
So what happened
Quoting from a previous post about the CAQ heartland, south of Quebec City.
"If the PLQ starts winning big here and the CAQ is still managing to find seats somewhere, it means the old ADQ voter has split between the CAQ and PLQ. ."
More analysis later on, of course, as well as full results (eventually)
PROJECTION CHANGE
As the counts come in it is starting to become more and more clear that the PLQ will take near it's projected maximum.
Our projection is now for a PQ minority based on counts that have come in.
Our projection is now for a PQ minority based on counts that have come in.
We will examine where we went wrong overnight.
Timings
Live Blogging will begin when the polls close. I've chosen to make a new post every few minutes.
Things to keep an eye on.
What should you keep an eye out for as the results roll in?
Sherbrooke
Jean Charest should lose by a convincing margin. The first 10 or 20 polls in any riding can show just about any result, just ask the leader of the Quebec wing of the Marxist-Leninists (who was shown leading in his federal riding in 1997, based on 1 poll) Beyond 20 polls however things get a bit more stable. At that point here is what you should see: Charest trailing. Maybe Charest jumping into the lead for a few minutes, but going back to trailing. Here is what you should not see: Charest ahead, and staying ahead. If you do see the latter it is a sign that the PLQ Campaign has done far better than expected. A Charest win in his riding should spell a PLQ minority government.
The PLQ should sweep the ridings in West Island Montreal. The PQ is certainly not going to sweep them and there is no evidence that either the QS or ON has managed to appeal to them in large numbers. If the CAQ starts winning seats here all bets may be off. The CAQ is not expected to take the Anglo vote, and if it does, it could decimate the PLQ, and, possibly thrust itself into government as a result.
Areas of Francophone Montreal are expected to split between the PQ and PLQ. The QS is expected to take Mercier and Gouin. The QS could have a strong shower in Laurier-Dorion, Sainte-Marie--Saint-Jacques, and Outremont. They may also do well in Rosemont, and Hochelaga--Maisonneuve. ON is expected to do well around these areas too. Should the QS start leading in any of these ridings (past 20 polls) and hold on to the lead, they may well be able to win the riding, and the same goes for the ON. This would be on the strength of the Student vote, and those whom the Students have convinced to follow them.
Parts south of Quebec City (Beauce, Bellechasse, Arthabaska, Levis, etc) are expected to be the heartland of CAQ support. If the PLQ starts winning big here and the CAQ is still managing to find seats somewhere, it means the old ADQ voter has split between the CAQ and PLQ. If, however, the PLQ sweeps this area and the CAQ is doing awful, it means the CAQ has failed to find new voters and has lost old ADQ voters.
The PQ is expected to sweep all the ridings in this area (Chicoutimi, Dubuc, Jonquiere, Lac-Saint-Jean, Roberval, even Rene-Levesaue, and Duplessis) a failure to do so would indicate all the existing projection models have a flaw and other areas of supposed PQ strength could be equally misleading.
The PVQ has had an awful campaign, only finding candidates in half the ridings. This is their best riding in the province, and, the leader has chosen to run here. The lower the vote share for the PVQ in this riding, the longer it will take to rebuild the party.
Perhaps the most interesting "smaller party" this election is the new provincial Conservative Party lead by former MP Luc Harvey. He is running in Levis and is not expected to do well. You can gawk at him in this riding as the results come in.
Sherbrooke
Jean Charest should lose by a convincing margin. The first 10 or 20 polls in any riding can show just about any result, just ask the leader of the Quebec wing of the Marxist-Leninists (who was shown leading in his federal riding in 1997, based on 1 poll) Beyond 20 polls however things get a bit more stable. At that point here is what you should see: Charest trailing. Maybe Charest jumping into the lead for a few minutes, but going back to trailing. Here is what you should not see: Charest ahead, and staying ahead. If you do see the latter it is a sign that the PLQ Campaign has done far better than expected. A Charest win in his riding should spell a PLQ minority government.
Anglo-Montreal
Franco-Montreal
South of Quebec City
Saguenay
Notre-Dame-de-Grace
Levis
Another - 70 #PQ | 30 #PLQ | 22 #CAQ | 2 #QS | 1 #ON
.
- 70 #PQ | 30 #PLQ | 22 #CAQ | 2 #QS | 1 #ON
.
.
Final Proj - 70 #PQ | 30 #PLQ | 22 #CAQ | 2 #QS | 1 #ON
.
- 70 #PQ | 30 #PLQ | 22 #CAQ | 2 #QS | 1 #ON
.
.
Turnout?
At this point in the day, the media usually talks about the turnout they've seen so far. Much of this discussion happens just after lunch, when the journalists have had a bite to eat and gone to cast their ballots and asked around... at least, that is what I suspect.
How is the Weather!?
It always amuses me that at about this time of day, the media wants to talk about how the weather will impact the election. They list the areas where it is raining and say that low turnout will help this or that party. Alternatively, they list areas with great weather and imply people would rather spend the day walking in the park rather than voting. When it comes down to it, outside of truly extreme weather, none of this really ever effects the results.
.gif)
Subscribe to:
Posts (Atom)