Reply #58: Both of your scenarios are implausible

[TruthIsAll]

YOU:
Do you think 11% Gore voters voting for Bush and 5% Bush voters voting for Kerry is more plausible than 16% Gore voters voting for Bush?

ME:
YES, BUT IT IS STILL IMPLAUSIBLE, ONLY LESS SO.
11% OF GORE VOTERS FOR BUSH IS HIGHLY UNLIKELY, BUT PLAUSIBLE.
16% OF GORE VOTERS FOR BUSH IS A PRACTICAL IMPOSSIBILITY.

KERRY WILL NEVER GET JUST 5% OF REPUBS. THAT IS IMPLAUSIBLE.
BUT I WILL RUN THIS IMPLAUSIBLE SCENARIO IF YOU WISH.

YOU:
How about 10% Gore, 4% Bush?

ME:
YES, BUT IT IS ALSO IMPLAUSIBLE.
10% OF GORE VOTERS FOR BUSH IS PLAUSIBLE, BUT NOT 4% BUSH FOR KERRY.

KERRY WILL NEVER GET JUST 4% OF REPUBS. THAT IS IMPLAUSIBLE.
BUT I WILL RUN THIS IMPLAUSIBLE SCENARIO IF YOU WISH.

LOOK AT PARTY CROSSOVERS IN THE FINAL NATIONAL EXIT POLLS FOR 1996-2000 AND THE 2004 PRELIMINARY 13047 POLL:

90-92% OF DEMOCRATS VOTED FOR THE DEM.
8-10% OF REPUBS VOTED FOR THE DEM.

LET'S BE REALISTIC.
DON'T ASSUME OUTLANDISH HYPOTHETICALS.
USE HISTORICAL PARTY CROSSOVER RATES, WITH "PLAUSIBLE" ADJUSTMENTS.

EVEN WITH IMPLAUSIBLE ASSUMPTIONS FAVORING BUSH, MY SENSITIVITY
ANALYSIS SHOWS THAT HE IS A LOSER.
YOU:
This really doesn't matter as long as you are not trying to "prove" that Kerry won the popular vote. And I notice that you did not say "proof" this time.

ME:
DOES 99.99% PROBABILITY CONSTITUTE PROOF BEYOND A REASONABLE DOUBT IN A COURT OF LAW?

YOU:
But if the point of the exercise is to "prove" that Kerry won the popular vote, we would need to be sure that there is no plausible mix where Bush could have won it. That is why I am being so picky.

ME:
THERE IS NO "PLAUSIBLE" MIX IN WHICH BUSH WINS.
THE TURNOUT/NEW VOTER SENSITIVITY ANALYSIS PROVES IT.

YOU HAVE SUGGESTED TWO "IMPLAUSIBLE" SCENARIOS.
BUT I WILL RUN THESE SCENARIOS IF YOU WISH.