Replies

"Am I the only one who has done the math? As an actuary, let me tell you that there is between a 91% and 93% chance that Republicans WILL regain the Senate in 2002 elections, assuming no one switches parties and Cheney remains VP."

Actually, I have done the math, and your assessment seems wildly optmistic. How did you arrive at such a figure?

I have set up a simple computer program to simulate the election results, based on assumed probabilities for what I deemed the 7 truly competitive races (SD, MN, NJ, MO, NH, CO, AR). I assumed all other races were won by the incumbent party. I gave the GOP a 50/50 chance to achieve a pickup in each individual race in SD, MN, NJ and MO. I gave the Democrats a 50/50 chance to pickup AR, and a 40/60 chance at picking up each CO and NH.

I then ran the program 10,000 times. The results:

GOP gains 4 seats: 0.82%

GOP gains 3 seats: 5.58%

GOP gains 2 seats: 16.11%

GOP gains 1 seat: 27.33%

Stalemate (neither party gains): 28.08%

Dems gain 1 seat: 16.04%

Dems gain 2 seats: 5.16%

Dems gain 3 seats: 0.88%

This puts the odds of the GOP regaining the Senate at approximately 50%. The odds of the Democrats increasing their majority is only about 22%.

If anyone wildly disagrees with my assumptions, let me know. It's an easy program to run, and I can post results under different scenarios.

Well, of course, in any calculation, garbage in, garbage out...one's assumptions about the races will determine the outcome; what you have assumed is very, very pessimistic and may well be based largely upon Democrat propaganda.

Your first assumption is that there are seven competitive seats, and that they are the ones you mention. You ignore any chance that any other seat (27 more up) will change hands.

You then assume that the four Democratic seats are just coin-tosses with no trend. You also assume that Arkansas is such a coin-toss, but seem then to put Republicans ahead 60-40 in Colo. and NH.

Colorado is a big difference between me and you: I would call this a near-certain keeper for R: in fact the chance of losing OREGON is greater than that of losing Colo. On the possibility of Republican losses, then, I would give us a 40% chance of losing Ark, 40% col NH; but only 30% ch of los OR and 20% ch of los CO. Of the four, we expect then to hold (.6+.6+.7+.8)= 2.7, a loss of 1.3.

Even more importantly, the four Democratic poss-losses you cite are really PROB-losses, goners. I would say only a 10% chance Dem can hold SDak; 20% MO; 30% MN; and 40% NJ. From the four, then, we expect to pick up (.9+.8+.7+.6)=3.0.

That sums up your seven and adds one possible other State where R could lose, Oregon. Eight states so far in play; net gain 1.7.

I would add Iowa as a tossup, 50-50; and then say that Georgia and Montana are only 60-40 ahead for D; and Louisiana and Illinois only some 70-30. From those five expected gain is (.5+.4+.4+.3+.3)=1.9.

Thus I have a total of 13 states in play, not your seven.

I would then say that there is a 50% chance some other imponderable place will switch D to R; 25% that two such will; one-eighth chance that three will, etc. This means an expected R gain from this term alone of (.5+.25+.125+.0625...)=1.

Overall, then, gain is (1.7+1.9+1.0)=4.6.

The entire calculation of the chance that R fails to gain 0.5 or more (remember roundoff) yields only some 3% as compared to 8% the last time I had done it-- Missouri and Texas had not looked so good for R as they now do.

Thus the chance we DO gain one or more is some 97%.

Now, someone out there may say that just as you are too pessimistic, I am too optimistic. While I feel I am right-on, if the truth lay halfway between us on each of the individual states we would still get to around an 89-90% of the gain of one or more.