Oddsmaking

[bunnyfeet]

A question for my actuary friends (i'm looking at you, satogaeruIn an interview with CNN, Fareed Zakaria says

Comments

[satogaeru]

The thing about actuarial odds is that they work when you have a LOT of people, but they don't necessarily work for just one guy. If you're willing to ignore that mathematical fact I can get you some numbers.

When I calculated it it was more like 1 in 10 odds, given that he's 72. But probably what you want is, what are his odds given that he's 72 with second-stage melanoma. My coworker says someone has actually worked that out and she's going to send it to me. I'll post it when I get it.

[bozini]

My concerns about the Vice Pres are not just about stepping in as President. It is more about the actual responsibilities of the VP. It seems to me choosing someone who doesn't believe in science would be a bad fit as chairmen of NASA or on the board at the Smithsonian (Smithsonian Environmental Research).

[tonyvila]

I really have to go with Matt Damon on this one: "I need to know if she really think that dinosaurs were here 4000 years ago. I want to know that, I really do. Because she's gonna have the nuclear codes."

Reference Video:

[tonyvila]

Looks like the embed code didn't work - The video can be seen here: http://www.huffingtonpost.com/2008/09/10/matt-damon-on-palin-like_n_125334.html

[tonyvila]

Here's an excerpt from a 2004 actuarial life table (McCain is 72):

Exact age - 72
Death probability - 0.032978
Number of lives - 67,016
Life expectancy - 12.01

This is for the average person. Not having McCain's 1500 page medical records, I can't comment as to his specific situation, nor am I qualified to. I know he's had skin cancer, which was reportedly excised before spreading.

He's gonna have a really stressful job, which pulls him out of the average a fair bit, but other than that, it looks like he has a 3.3% chance of dying within a year with a life expectancy of 12 more years. if he's statistically normal.

[tonyvila]

And that's my wholly-unqualified, non-partisan comment. (except for the icon, of course)

[bunnyfeet]

not being up on my statistics, using that same table, in order to find the likelihood that today's average 72-year-old will die in the next 4 years, do we add the death probabilities for a 72, 73, 74, and 75 year old? which gives us about 15.2%.

Not exactly "something like 1 in 5", but not far off.

[tonyvila]

Seems like the easy way to do it, but you get a bad answer. The way the table works is that if you make it to 72, this is your chance of living another year. if you make it to 73, here's your chance for that year. Still sloppy and wrong (but righter) is to do it multiplicatively. Say you go to the store and something is 60% off and you have a 40% off coupon. It's not free - it's 76% off. By the same token, you multiply it out. It works out to 14.3%, which as I say, is voodoo math. satogaeru will have much better data than I, merely an interested amateur. I could always call my dad and ask - he's an actuary....

[elemess]

The AJC had some local actuarial firm run the numbers. They figured McCain would be fine for both terms (assuming election and reelection), while Obama would be fine for far longer.

They took into account McCain's cancer and Obama's smoking to individualize it a bit more; otherwise, actuarial tables only cover a broad swath of people and can't accurately predict any one person.

[baudot]

Can we take into consideration the considerable odds of an assassination attempt on the president? Though only a handful have succeeded, the odds are pretty high that the attempt will be made. Reagan showed that just because you have Alzheimers doesn't mean you can't shug off a bullet wound, but McCain might not manage the same stunt.

[satogaeru]

I <3 Jodie Foster!!

[tonyvila]

Not as much as Hinckley did, apparently :)