**** OFFICIAL comprehensive Chiefs playoff scenario thread ****

[alnorth]

Originally Posted by ChiefsFan5840

What am I not able to figure out a use for? I have a bachelors degree in actuarial science, I'm not exactly an idiot when it comes to numbers. I just think most of these projection system sites like AccuScore and what not are way off because of things they don't include. With more weeks left I would say they are more accurate, but still not as much as a knowledgeable fan could use without looking so much at stats and just using their general feel. With 2 games left I'd say using your own judgement on a probability we win 2 games, or win 1 game is a lot better than reading a percentage off one of these sites that are based strictly off statistics.

Cool. If you are trying to pull math nerd-rank, I guess I could mention that I have passed the 1st 4 exams and am sitting for 5 this spring. (CAS side) Not that it really matters.

You do not need to know every variable of a complex system to attempt to make accurate forecasts. We certainly dont have every aspect of the weather figured out, but the meteorologist still does a halfway-decent job telling me what will happen the next few days. Earthquake and hurricane catastrophe models probably dont have those probabilities down anywhere close to 100% either, but we still rely on them. How? Why? We identify a few variables, try to make predictions based on our imperfect knowledge, and see how well that model would have predicted the past. (For example, we have dozens of football seasons. How well would accuscore have done in predicting Week 13 20 years ago?) When a new event makes it apparent that there is a flaw in the model (accuscore begins to lose predictive power, an earthquake with a specific magnitude, epicenter, and regional damage occurs that the model does not show as reasonably possible) you revise the model.

You'll learn more about constructing actuarial models when you sit for exam C, if you havent already.

[WebGem]

Originally Posted by alnorth

Cool. If you are trying to pull math nerd-rank, I guess I could mention that I have passed the 1st 4 exams and am sitting for 5 this fall. (CAS side) Not that it really matters.

You do not need to know every variable of a complex system to attempt to make accurate forecasts. We certainly dont have every aspect of the weather figured out, but the meteorologist still does a halfway-decent job telling me what will happen the next few days. Earthquake and hurricane catastrophe models probably dont have those probabilities down anywhere close to 100% either, but we still rely on them. How? Why? We identify a few variables, try to make predictions based on our imperfect knowledge, and see how well that model would have predicted the past. (For example, we have dozens of football seasons. How well would accuscore have done in predicting Week 13 20 years ago?) When a new event makes it apparent that there is a flaw in the model (accuscore begins to lose predictive power, an earthquake with a specific magnitude, epicenter, and regional damage occurs that the model does not show as reasonably possible) you revise the model.

You'll learn more about constructing actuarial models when you sit for exam C, if you havent already.

I just finished the preliminary's this year but I had to do all 5. CAS doesn't have to take MLC, right? If you don't take MLC what is the 5th preliminary? Or is it just a module type thing? And I wasn't trying to pull math-nerd ranks, I was showing that I can apply it if I need to I just don't trust the numbers. Like when cdcox was telling us the Chargers had an 86% chance to win in each of their remaining 3 games. I just don't believe that and think it's way the hell too high.

[alnorth]

Originally Posted by ChiefsFan5840

I just finished the preliminary's this year but I had to do all 5. CAS doesn't have to take MLC, right? If you don't take MLC what is the 5th preliminary? Or is it just a module type thing? And I wasn't trying to pull math-nerd ranks, I was showing that I can apply it if I need to I just don't trust the numbers. Like when cdcox was telling us the Chargers had an 86% chance to win in each of their remaining 3 games. I just don't believe that and think it's way the hell too high.

CAS accepts either MLC or 3L (3L is their own version of that half-exam, less life contingencies which are of limited use anyway for casualty actuaries, much heavier emphasis on statistics). CAS students can basically pick one or the other, depending on if they feel they are stronger on statistics or life con. Exam 5 is the first CAS upper-level exam (there are 5 cas upper-level exams, exams 5 through 9). No idea what the SOA is doing anymore, they seem to be going heavy with the modules.

[WebGem]

Originally Posted by alnorth

CAS accepts either MLC or 3L (3L is their own version of that half-exam, less life contingencies which are of limited use anyway for casualty actuaries, much heavier emphasis on statistics). CAS students can basically pick one or the other, depending on if they feel they are stronger on statistics or life con. Exam 5 is the first CAS upper-level exam (there are 5 cas upper-level exams, exams 5 through 9). Not idea what the SOA is doing anymore, they seem to be going heavy with the modules.

Oh okay, the only person I knew doing the CAS route moved to Chicago to work for Zurich NA a couple years ago and I lost contact with him so I don't know much about it. Good luck.

Just out of curiosity, put a percentage on the Chiefs making the playoffs. I say 65%. It's with a heavier weight on the possibility of SD losing a game than most people would have though.

[alnorth]

Originally Posted by ChiefsFan5840

Just out of curiosity, put a percentage on the Chiefs making the playoffs. I say 65%.

same, about 2/3. Little over 50/50 that they win out, a little under 10% that chiefs win the division as a 10-6 team. Chargers about 1/3 chance of winning the division (ignoring their unlikely wild card scenarios), and the raiders chances are a rounding error.

[cdcox]

Originally Posted by ChiefsFan5840

Oh okay, the only person I knew doing the CAS route moved to Chicago to work for Zurich NA a couple years ago and I lost contact with him so I don't know much about it. Good luck.

Just out of curiosity, put a percentage on the Chiefs making the playoffs. I say 65%.

My software says 52.5%
Football Outsiders has it at 54.9%
Accuscore has it at 86%, but that is ridiculously high. Plus their tiebreakers aren't programmed correctly. I don't consider them to be serious about their craft.

A couple of years ago, Don Bank, a professional journalist with SI gave his "gut feel" about the chances of each team making the playoffs at somewhere between the half-way and 3/4 mark. The odds per playoff spot didn't even add to 100%. So, having a rigorous system definitely improves your odds. Having a computer to evaluate scenarios definitely increases prediction skill.

My best moment was calling the Chiefs draft order 24-hours before the NFL did. Close follow up was correctly describing the exact scenario under which the Chiefs could make the playoffs in 2006.

Worst moment was calling the Broncos the division champs in week 6 last year. But overall, the software leads me in the right direction far more often than it leads me astray. Actuarial science is all about managing risk, and you are far better off with a quantitative model than without one.

[WebGem]

Originally Posted by cdcox

Worst moment was calling the Broncos the division champs in week 6 last year.

You mean you didn't predict them to finish with the same record as the 0-6 Titans?

[cdcox]

Originally Posted by ChiefsFan5840

You mean you didn't predict them to finish with the same record as the 0-6 Titans?

No, but nobody else did either.