How much cheating would be necessary to turn Ohio?

[Jersey Devil]

All the impressive graphs and charts I've seen here showing huge Bush votes in heavily Democratic districts, more than 100% turnout in districts, etc., made me start thinking just how much cheating would be necessary to steal Ohio (though you could apply this to any state).

Kerry "lost" Ohio by about 137,000 votes and there are 88 counties in Ohio, so roughly 1,500 votes per county would do it. So how many machines would there be (assuming equal population in the counties which I know is not a fact but for these purposes I think it is OK to assume that). How many machines per county? I frankly have no idea but I will assume 1,500 machines per county (there are 22 in my little town in NJ for 18,000 population on election day). That would be 1 vote per machine in order to take Ohio.

No one would notice. It would be easy. But, if for some reason you cannot cheat in some of those counties (66 of 88 Ohio counties have punch card voting), you have to get your numbers from the other 22 counties to make up the difference (you can basically say the same for Florida with optical scan vs touchscreen).

It is election night, you are Karl Rove and have the exit numbers. you are desperate so you go for it - there are obvious clues that will be left of a fix - thus the strange numbers we have been seeing in Ohio and Florida.

Just one man's speculation but so far I think it is as good as anyone's.

[glitch]

1. Votes still being counted aren't they? So we don't know the final margins

yet. It's quite possible to cheat with punchcards as long as no one manually recounts the ballots. Riskier there and with the optical scans because they can be manually counted and compared, if someone demands a recount.
Important to appear overwhelming so no one would demand the recount.
Touchscreens can be verfified after the fact too, in a manner. Convass the precincts, get sworn affadavits (I think dems are mad enough to forgo voter privacy in this case, don't you?) and there you go!
Easy enough and not expensive to do just a few precincts at first for testing purposes.
All that is required is movitivation. Apparently that is where we will fail.

[seabeyond]

2. not if prepunched

then it is over voting and they are "thrown away"

i heard this is what was done in florida in 2000?

[glitch]

4. True. I actually prefer canvassing selected precincts, that way you

don't have to worry about what the voters used or trust that votes haven't been manipulated or destroyed.

[hertopos]

3. Actually, it all depends on the nature of fraud

For example, if the fraud mainly turned Kerry vote into Bush, you only need a half of 137k plus one.

Hertopos.

[idiosyncratic]

5. Well . . . according to these numbers, they screwed up

BIG TIME!!

How Do You Get Percentages Like These?