[Bigh4u2]

"So according to Gallup, if you polled 48-47 before and 48-40 after, that means there has been no bounce, or if you polled 47-42 afterwards, you got a -1 bounce.

Does that make sense to you?'

Actually, it does.

Consider that the 'bounce' in candidates pols would have to show an 'increase' in his. Not a decrease in his opponents.

If you use the 48-47 and 48-40 figures, it just shows that some that were in the 'Kerry' column are not questioning their decision.

[Bigh4u2]

"If you use the 48-47 and 48-40 figures, it just shows that some that were in the 'Kerry' column are not questioning their decision. "

Sorry,.

Should have read 'are NOW questioning their decision.'

[Truthsearcher]

The ultimate goal is have more people vote for you than your opponent.

Decreasing the number of people who votes for your opponent is as good a means to achieve that objective as getting more people to vote for you. Therefore any analysis of polling should take both strategies into account, not just the latter.

[Conservative til I die]

That is an odd situation, and one I would assume doesn't happen very often, where the candidate who mighte get a bounce gets no increase, but his opponent's completely bottoms out.

No, it's not a bounce, but it's jsut as good as one. It's a "swing" which might be a more correct term.

Which would you rather have, Bush going from 48-48 to 51-46 (a 5 point swing/bounce) or Bush going from 48-48 to 46-31 (by your definition, a negative two point bounce)????