Do these Cleveland precinct results look right to you?

[○∙◄☻¥tπ[╪AV┼cVê└]

Election night canvass:
Precinct                              Badnarik   Bush   Kerry  Blank  Peroutka
1801 CLEVELAND 4A                       0    10   307     0     0
1802 CLEVELAND 4B                       0    21   283     0     1
1803 CLEVELAND 4C                       0    11   304     0     0
1804 CLEVELAND 4D                       1    23   511     0     0
1805 CLEVELAND 4E                       0    14   338     0     3
1806 CLEVELAND 4F                       0    21   290     0   215
1807 CLEVELAND 4G                       2    57   258     0     3
1808 CLEVELAND 4H                       1    12   454     0     3
1809 CLEVELAND 4I                       0     6   468     0     0
1810 CLEVELAND 4J                       0    20   424     0     1
1811 CLEVELAND 4K                       0    10   387     0     1
1812 CLEVELAND 4L                       0    15   319     0     0
1813 CLEVELAND 4M                       1    17   476     0     2
1814 CLEVELAND 4N                     163    11   318     0    7
1815 CLEVELAND 4O                       0    10   419     0     7
1816 CLEVELAND 4P                       1    31   223     0     0
1817 CLEVELAND 4Q                       1    17   477     0     2
1818 CLEVELAND 4R                       4    12   455     0     4
1819 CLEVELAND 4S                       2    30   374     0     0
http://boe.cuyahogacounty.us/boe/results/history/2004/110204_GE_Canvass.txt


Not fixed in latest results:
Precinct    Badnarik   Bush   Kerry   Peroutka
CLEVELAND 4A     0      10     320       0
CLEVELAND 4B     0      21     289       1
CLEVELAND 4C     0      12     313       0
CLEVELAND 4D     1      24     518       0
CLEVELAND 4E     0      14     356       4
CLEVELAND 4F     0      20     299     215
CLEVELAND 4G     3      58     261       3
CLEVELAND 4H     2      12     478       3
CLEVELAND 4I     0       7     494       0
CLEVELAND 4J     0      20     436       1
CLEVELAND 4K     0      10     413       1
CLEVELAND 4L     0      15     338       0
CLEVELAND 4M     1      17     478       2
CLEVELAND 4N  164     12     334      10
CLEVELAND 4O     0      13     439       6
http://boe.cuyahogacounty.us/boe/results/history/2004/EL52.TXT

[minionofmidas]

4F and 4N are the fishiest ones in the list. But there's several more very odd clusters of third party votes: 3B, 3I, 5B, 8G, 8I, 13X, possibly 10Q, 11M, and 15H although these might be genuine.
What I'd really like is a precinct map, o/c.

Kerry won the city of Cleveland 83.3 - 15.8, btw.
He won every single precinct within the city.
Of the 21 wards, the closest was the 21st, which Kerry won 61.5 - 38.0. In ten wards, he topped 90%. In four wards, 95%.

[Whacker77]

Send in Jesse.  That's the only way we can get to the bottom of this!  Remember, voter errror is voter error.  If some one can't correctly fill out a ballot when hundreds of thousands did, too damn bad.

[jimrtex]

4F and 4N are the fishiest ones in the list. But there's several more very odd clusters of third party votes: 3B, 3I, 5B, 8G, 8I, 13X, possibly 10Q, 11M, and 15H although these might be genuine.
What I'd really like is a precinct map, o/c.

Cleveland precinct map

Or just shorten up to  /boe/ and you will get to the Board of Elections site.

The SBOE results for 4F and 4N look somewhat odd as well.  There might be problems with East Cleveland 4-F; Euclid 3-C; and North Royalton 3-A.

You must be logged in to read this quote.

It was surprising how far Voinovich ran ahead of Bush.  Where Bush was in the teens, Voinovich was around 100.

[Andrew]

This isn't voter error.

It's machine error, programming error, or fraud.

[MODU]

This isn't voter error.

It's machine error, programming error, or fraud.

hahaha . . . yes, how dare those figures be accurate!  There MUST be fraud here somewhere.

Or maybe those communities actually like the third party candidates better?  Nah . . . impossible.  How dare someone vote a third party candidate!

HAHAHA

[stry_cat]

Is there an organized LP or CP affiliate in the questionable precincts?  Was there an organized GOTV effort or some other activity there?  How many registered voters are in each of the precincts? 

I agree something looks fishy.  While I like to see Badnarik beating Bush,  there still needs to be a reason for such an anonmolous result.

[A18]

Even if all the third party votes were supposed to go to Kerry, it wouldn't change the outcome of the election. Bush got over 50% in the state.

[jimrtex]

4F and 4N are the fishiest ones in the list. But there's several more very odd clusters of third party votes: 3B, 3I, 5B, 8G, 8I, 13X, possibly 10Q, 11M, and 15H although these might be genuine.
What I'd really like is a precinct map, o/c.

3B and 3I used the same polling place.
4F and 4N used the same polling place.
8G and 8I used the same polling place.

3B/3I is adjacent to 4F/4N with MLK Jr Dr running through them (towards the SE corner of Cleveland).  Given the Kerry majority, it is probably a black area.

If they used punched ballots and rotated the order between precincts, then it could be that the ballots have been shuffled.  This might not be recoverable, other than by statistical methods.

5B is an odd precinct alongside a freeway interchange that forms part of a CD boundary.   This explains the low vote total.  It may or may not provide a reason for the Peroutka vote.

10Q voted with 5 other precincts (10D, 10G, 10N, 10R, 10S).  It has a somewhat different street pattern than the other precincts.  This might explain the Badnarik vote.  Or it could be a very mild shuffling among the other 3 precincts.

11M shared a polling place with 11O and 11S.

13X shared a polling place with 13Q and 13Z.

14L shared a polling place with 14B and 14M.

15H shared a polling place with 15I.

E Cleveland 4-F shared a polling place with 4-A and 4-H.

Euclid 3-C shared a polling place with 3-A.

North Royalton 3-A shared a polling place with 3-B, 5-A, 5-C, 5-D, and 5-E.

[Andrew]

hahaha . . . yes, how dare those figures be accurate!  There MUST be fraud here somewhere.

Or maybe those communities actually like the third party candidates better?  Nah . . . impossible.  How dare someone vote a third party candidate!

HAHAHA

In 2000, precinct 4F had four votes for the Constitution Party candidate.  Precinct 4N had one vote for the Libertarian.  Those communities do not actually like third-party candidates:

http://boe.cuyahogacounty.us/boe/results/history/2000/110700_GE_Canvass.txt

For the record, I did not say there must be fraud here.  Fraud was one of three options I suggested.

[stry_cat]

4F and 4N are the fishiest ones in the list. But there's several more very odd clusters of third party votes: 3B, 3I, 5B, 8G, 8I, 13X, possibly 10Q, 11M, and 15H although these might be genuine.
What I'd really like is a precinct map, o/c.

3B and 3I used the same polling place.
4F and 4N used the same polling place.
8G and 8I used the same polling place.

3B/3I is adjacent to 4F/4N with MLK Jr Dr running through them (towards the SE corner of Cleveland).  Given the Kerry majority, it is probably a black area.

If they used punched ballots and rotated the order between precincts, then it could be that the ballots have been shuffled.  This might not be recoverable, other than by statistical methods.

5B is an odd precinct alongside a freeway interchange that forms part of a CD boundary.   This explains the low vote total.  It may or may not provide a reason for the Peroutka vote.

10Q voted with 5 other precincts (10D, 10G, 10N, 10R, 10S).  It has a somewhat different street pattern than the other precincts.  This might explain the Badnarik vote.  Or it could be a very mild shuffling among the other 3 precincts.

11M shared a polling place with 11O and 11S.

13X shared a polling place with 13Q and 13Z.

14L shared a polling place with 14B and 14M.

15H shared a polling place with 15I.

E Cleveland 4-F shared a polling place with 4-A and 4-H.

Euclid 3-C shared a polling place with 3-A.

North Royalton 3-A shared a polling place with 3-B, 5-A, 5-C, 5-D, and 5-E.

So it basically boils down to the precints vote in the same place and some idiot mixed up the votes so that you can't tell which one is which.  Nothing extremely wrong, but clearly is something that proper proceedure would have prevented.

[○∙◄☻¥tπ[╪AV┼cVê└]

Send in Jesse.  That's the only way we can get to the bottom of this!  Remember, voter errror is voter error.  If some one can't correctly fill out a ballot when hundreds of thousands did, too damn bad.

So everytime there's something weird, it's voter error? Explain to me why people screwed up so much in those 2 precincts, but not the rest of the ward?

And I suppose the precinct in another Ohio county where Bush got close to 4000 more votes than people who voted was also voter error. Sorry, that doesn't work.

[Jake]

Send in Jesse.  That's the only way we can get to the bottom of this!  Remember, voter errror is voter error.  If some one can't correctly fill out a ballot when hundreds of thousands did, too damn bad.

So everytime there's something weird, it's voter error? Explain to me why people screwed up so much in those 2 precincts, but not the rest of the ward?

And I suppose the precinct in another Ohio county where Bush got close to 4000 more votes than people who voted was also voter error. Sorry, that doesn't work.

Keep grasping at straws, but like Philip said, if all these "fradulent" third party votes went to Kerry, he'd still lose. 

About the 4000 votes.  Maybe the voter registration paid off for the democrats.  Only the dead people picked Bush.

[zorkpolitics]

Clearly looks like official incometence.  I think this is a punch card area, so a machine must have been set up so that sometimes people punching Kerry punched the third party chad.

Fraud is a highly unlikely, why would anyone switch so many votes in just 2 precints?  Especailly since it shows up as a strange result.

[Alcon]

Odd that 77% of people don't find these results odd and basically 100% who have commented do.

Voter error?

[○∙◄☻¥tπ[╪AV┼cVê└]

Ahh, this is fun. Let's look at the 3rd party votes just in those 2 precincts


Precinct Badnarik Peroutka
4N 164 10
4F 0 215

Assuming that the 3rd party voter were randomly with probability p for Badnarik and probability 1-p for Peroutka, then
Probability of 4F is (1-p)^215
Probability of 4N is p^174+174*p^173*(1-p)+15051*p^172*(1-p)^2+862924*p^171*(1-p)^3+36890001*p^170*(1-p)^4+1254260034*p^169*(1-p)^5+35328324291*p^168*(1-p)^6+847879782984*p^167*(1-p)^7+17699490469791*p^166*(1-p)^8+326457268665034*p^165*(1-p)^9+5386544932973061*p^164*(1-p)^10

You decrease one of these at the expense of the other. Solving for these to be equal, p=0.4124, which gives us a probability of 2.25*10^-50.

The odds are 1 in 4.4 *10^49.

[jimrtex]

Precinct and election district maps and ballots for Cuyahoga County (Cleveland) are available at:

Cuyahoga County Board of Elections.

Maps/Polling Places shows not only the precincts and other boundaries, but also polling places.  Many precincts share a polling place.  These are street-level maps, and Cleveland is in a single large PDF file with each ward in a single page.

View Ballots shows the ballots for each precinct.  The ballot order is scrambled at least for President.  With 4 candidates and 1 disqualified candidate, there are 120 possible pemutations.  I didn't check the down ballot races, but the results for the 11th SBOE where there were 3 candidates seemed to be somewhat anomalous in the precincts with the extra Badnarik and Peroutka voes.

In a punch card system such as used in Cuyahoga County, the ballot card itself does not show the candidates.  It does have perforated holes that can be punched out that are numbered, so that can make a hand count, as well as tell someone to "punch the Hole 3".

I have heard of one case (in Houston) where someone helping out with ballot counting was visually checking the ballots, and punching out a hole in a primary race if the voter had not voted for a candidate in that race.

When you vote, you slip the ballot card into a guide, that has a ballot booklet fitted to it.  You find the candidate you wish to vote for, and punch the hole next to the name.  There is typically a hole number and an arrow pointing to the hole next to the name:

George Bush and Dick Cheney (Republican)    2 ==>  |  O
John F Kerry and John Edwards (Democrat)    3 ==>  |  O
etc.

(see the initial page of the ballots for a picture of a ballot).

It was my experience that I totally ignored the numbers when voting, and simply followed the arrows.

Typically there will be several voting devices/voting booths in each precinct.  They are inexpensive, have no electrical parts, and typically the voting booth will fold up into an easily luggable package.  If we assume five minutes per voter (there were 46 separate ballot items on the ballot that I looked at, including 25 where there were more than one choice), as well as reducing lines in busy periods, one device could handle almost 150 voters in a 12-hour period.  Typical turnout was around 500 per precinct meaning that 4 or 5 devices/voting booths might be used.

Because the ballot order changed from precinct to precinct, the voting devices used for each precinct are different, including where the same polling place was used for more than one precinct.  If several precincts voted in a single room, then you have to be sure that a voter is directed to the correct device/booth for precinct.  If the voting officials set up one of the devices wrong (4 booths for 4F, and 1 for 4N in one row; and the reversed grouping for the other row), then directing a voter to choose one booth from the row will mean that some are directed to a booth for the wrong precinct.  Most voters would never notice.  How are they to know that the ballot order is supposed to be Kerry, DQ, Peroutka, Badnarik, and Bush; rather than Peroutka, Badnarik, Bush, Kerry, and DQ?

After he has voted, you have to be sure that he deposits his ballot into the correct ballot box for his precinct.  Using the wrongly setup device for a precinct would result in the wrong candidates being voted for.  Placing the ballot in the wrong ballot box would result in the ballot being counted for the wrong candidate (according to voter intent).

Other possible sources of error.  Misdirected fraud.  It is easier to mask voting fraud in in a 95% Democrat district.  If 450/500 is 90%. 
(450+50)/(500+50) is only 90.9%.  Who is going to notice the extra Kerry votes?  It is only when, 0/500 is 0%, and (0+50)/(500+50) results in a 9.1% Badnarik vote that something is noted as going amiss.  So if the fraudster punches an extra 50 ballots with hole 3 punched, when he should have punched hole 5 for _that_ precinct, they will be cast for the wrong candidate (according to frauster intent).

Bad advice from partisans outside the polling place.  A Democrat campaign worker could exhort voters to not let Bush-Ashcroft-Diebold-Harris steal their votes.  Be sure to punch Hole 3 and be sure to check for hanging chads.  If he were an ACT outsider he might not even realize that there were multiple precincts at a polling place.  He might believe that telling someone to vote for "3" was legal, whereas voting for "Kerry" was not.   Or voters might not know their voting precinct.

Campaign Worker: 3B or 3I?
Voter: I don't know.
Campaign Worker: Where do you live?
Voter: Over there, pointing vaguely in the wrong direction.
Campaign Worker: That would be 3B vote for "3" and check for chads.

A confused voter might simply skip using the punching device, and punch hole 3, check his ballot that he has voted "right" and deposit the vote.

Absentee voters.  This problem was noted before the election.  The voter uses a punch card, put doesn't use the voting device.  Instead he is given a booklet with the candidates and their hole numbers.  He looks at his booklet, finds the number he wants to punch out and punches the chad out.  In Cuyahoga County, the absentee booklets were printed with a single candidate order.  They were then customized for a given precinct by including the correct number:

Bush        4
Kerry       2
Badnarik  3
DQ           1
Peroutka  5

If you lay the booklet next to the punch card (similar to how you would vote at a regular polling place), then scan across, you may punch the wrong hole.

So let's look at the ballot order:

3B Kerry is 1st, 3I Peroutka is 1st.
3B Badnarik is 4th, 3I Kerry is 45h.

3I Peroutka 70 votes.
3B Badnarik 41 votes.
4F Kerry is 1st, 4N Badnarik is 1st.
4F Peroutka is 3rd, 4N Kerry is 3rd.

4N Bandarik 163 votes.
4F Peroutka 215 votes.
5B voted alone, but it is an odd precinct alongside freeway.
8G Kerry is 2nd, 8I Peroutka is 2nd.
8G Badnarik is 5th, 8I Kerry is 5th.

8I Peroutka 27 votes.
8G Badnarik 51 votes.

8H also voted at this polling place.  Both Peroutka and Badnarik received 0 votes.  If there were misdirected Kerry votes from 8H, they would benefit Bush (to 8G) or DQ (to 8I).  If there were misdirected votes into 8H they would benefit Bush(from 8I) or DQ(from 8G).

8H Bush 45 (8G:19 and 8I: so it is possible that Bush received some Kerry votes from 8I.  There might be other precincts where Bush benefited, that were not so obvious as the Badnarik/Peroutka cases.  It would be expected that Bush received 5% of the vote, so if he got 10% it might not stand out except to someone particularly familiar with the precinct.
10Q I didn't check.  6 precincts voted at this polling place.  The Badnarik vote was 11.
11M Kerry is 2nd, 11O Peroutka is 2nd.
11M Badnarik is 5th, 11O Kerry is 5th.

11O Peroutka 10 votes.
11M Badnarik 30 votes, Peroutka 10 votes.

11S also voted at this polling place (Badnarik 1, Peroutka 0).  There is not an obvious source for the 11M Peroutka votes, unless they were Bush votes from 11O.  Misdirected Kerry votes from 11S would benefit Bush (11M) or DQ (11O).  Misdirected Kerry votes into 11S would benefit DQ(11M) or Bush (11O).  Bush support was similar in all 3 precincts, and was up around 10%.
13X Peroutka is 1st, 13Q Kerry is 1st.
13X Kerry is 4th, 13Q Bandarik is 4th.

13X Peroutka 37 votes (of 358 total)
13Q Badnarik 9 votes (of 105 total)

The above suggests that voters were using the wrong voting device, but depositing their ballots in the right ballot box.  In this case, the number of cards pulled out of the box would match the number of signatures in the register.

If they used the right device, but the wrong box, there would be a difference between the number of cards pulled out of the box, but within the register.

13Z also voted at this polling place (but had no Badnarik/Peroutka votes (only 96 total). 

This combination is somewhat atypical, in that 13X had many more voters than 13Q and 13Z, and was also much more heavily for Kerry than the other 13Q and 13Z.  In 13X Bush received about 2% of the vote.  In 13Q and 13Z Bush was around 1/3.  This area appears to be near downtown, just east of the river, and slightly inland from the lake.
I had mistakenly list 14L.  It does not have a large Peroutka/Badnarik vote.
15H Bush is 3rd, 15I Kerry is 3rd.
15H Kerry is 4th, 15I DQ is 4th.

15H Badnarik 15 votes, Peroutka 7 votes.
15I Badnarik 4 votes, Peroutka 2 votes.

This is just north of the zoo (separated by a parkway).   The residential areas are south of Denison, with the voting location the YMCA on W 25th St.  The whole area appears to be cut off from the city as a whole.
East Cleveland 4-F Badnarik is 1st, 4-H Kerry is 1st.
4-F Kerry is 3rd, 4-H is Peroutka.

4-F Badnarik 35 votes.
4-H Peroutka 0 votes.

4-A also voted at this polling place.
Euclid 3-C Peroutka is 1st, 3-A Kerry is 1st.
3-C Kerry is 3rd, Badnarik is 3rd.

3-C Peroutka 88 vote (of 391)
3-A Badnarik 8 votes (of 70)

This is another instance of the precincts being of disiimilar size.
North Royalton 3-A Peroutka is 1st.

3-A Peroutka receives 37 votes.

There is not an obvious source for the Peroutka votes from any of the 5 other precincta at this polling place.  This is the only precinct of those listed that was carried by Bush.  N.Royalton is in the southern end of the county.  3-A is north of the Ohio Turnpike, though most of the precincts at the same polling place are to its south.

[○∙◄☻¥tπ[╪AV┼cVê└]

Whoa, that explains a lot.
It looks like some pollworker really screwed up.

[jimrtex]

So it basically boils down to the precints vote in the same place and some idiot mixed up the votes so that you can't tell which one is which.  Nothing extremely wrong, but clearly is something that proper proceedure would have prevented.

I think so.  If voters deposited their ballot in the wrong ballot box, it should have been caught since the count of ballots would not match the number of persons who had signed the register.  The election workers probably were told to make sure that voters deposited their ballot in the box for the precinct.  This would be easy enough to teach.  "We've got 2 precincts voting here, it's very important that the voter puts their ballot in the correct box.  The ballots are all the same, so if they get it wrong we're screwed."  Placing the ballot in the ballot box is the most public part of the process.  If there are poll watchers, they would focus on that, even maintaining their own count.

But even if you are being careful that voters place their ballot in the correct box  you might not realize the main reason for keeping the ballots separate - the meaning of a particular punch on the ballot in one precinct, does not mean the same thing as the same punch on the ballot in another precinct.

The ballot cards are generic.  It is the punching devices that are set for a specific ballot order.  Go to a voting booth with the wrong punching device, and you could end up voting for Peroutka or Badnarik instead of Kerry or Bush.  

Let's say that there are 5 voting booths for each precinct.  The central election facility sets them up, and delivers all 10 to one polling place.  What if the central election facility swaps the labels on two of them?  At the polling place, the election workers set the voting booths in two rows of 5, but because of the labels there are pair that are swapped between the two rows.  Send a voter to the correct row for his precinct, and he still votes on the wrong device.

Or maybe they ignore the labels, and mix them up.  Or maybe there is a space problem, and they improvise, putting 6 in one row and 4 in the other, and the message that one is different is lost during the day.

When the voter enters the polling place, they go up to the clerk with the register for their precinct, which they sign.  They get a blank ballot.  They are told to vote at one of the voting booths along the far wall.  They go stand in another loose line (the main line will be out the door, they won't let voters in faster than they can vote, but they don't want to have unused voting booths either), so during busy periods there may be a dozen or so people in this intermediate line.  Since it is so short, it isn't well formed, and people might get in the line for the wrong precinct.  Or it might not be a line, but more of a clump, with most voters recognizing a loose order.  As you chat while you wait, you hear a election worker telling people to go to the next available voting booth, "there's one free over their".  It is busy, you see that a booth is empty and you take it.  It is empty because it is for the other precinct.
Maybe the election worker who could have made sure that you went to the correct set of booths was distracted, or maybe they just were careless.

So my guess is that the problem was careless election workers, with perhaps some help from voters who didn't follow instructions.

[jimrtex]

Odd that 77% of people don't find these results odd and basically 100% who have commented do.

Voter error?

In your case, it appears that it was.  The question was not whether the results were odd, but whether they "looked right".  75% (15 of 20) answered NO.  But you may simply have been a victim of Jfern's wording of the question.

[○∙◄☻¥tπ[╪AV┼cVê└]

It looks like in precincts 4F and 4N, Kerry lost 379 votes to Badnarik and Peroutka, and Bush lost 10 votes to Peroutka, for a net loss to Kerry of 369 votes.  The fact that this happened in some precincts and not others suggests errors on the part of pollworkers.

[minionofmidas]

Excellent work Jim! Whao!
This is so obviously what happened here.

 "We've got 2 precincts voting here, it's very important that the voter puts their ballot in the correct box.  The ballots are all the same, so if they get it wrong we're screwed."  

More importantly, they are screwed.

You must be logged in to read this quote.

Which is obviously a VERY bad idea. I guess it was "justified" by keeping all the candidates' chances alike, but really all it does is confuse voters and falsify the election results - systematically if you assume that parties' voters are unequally informed due to class etc.

You must be logged in to read this quote.

But that person should bloody well know well enough he can't do that, shouldn't he? Any poll worker not remembering a thing like that (of course he's hypothetical) basically belongs prosecuted.

Once more, great work, Jim.

[jimrtex]

Excellent work Jim! Whao!
This is so obviously what happened here.

It makes assumptions about how precincts with a shared polling places work.  A quick scan of the Ohio election law doesn't show a specific provision for sharing polling locations.  There is a provision that allows precincts to be combined, rearranged, or enlarged, if voting machines or marking devices are used.  So perhaps this is interpreted as meaning "combined for the purpose of conducting elections."

There is another section of the law that limits precincts to 1400 electors This was increased in 2003 from 1000 electors, probably in response to federal law restricting purging of voting rolls.  The only exception to that part appears to apply more to apartment complexes and the like which have large numbers of voters in a small area.

You must be logged in to read this quote.

More importantly, they are screwed.[/quote]
But that isn't as obvious.  The election workers understand if the register count and the ballot in the box count don't match, they are going to have to account for the difference.  If you're in an area that is going to vote 95% Kerry, why else does it matter?

You must be logged in to read this quote.

Which is obviously a VERY bad idea. I guess it was "justified" by keeping all the candidates' chances alike, but really all it does is confuse voters and falsify the election results - systematically if you assume that parties' voters are unequally informed due to class etc.[/quote]
I found the Ohio election law dealing with ballot rotation.  The names are placed in alphabetical order, but then in each successive precinct, the names are rotated by one.  This is indeed what was done, as the order is Badnarik, Bush, Kerry, (Nader), Peroutka.  A side effect of this is that for the particular pattern we're seeing (Kerry votes counting for Badnarik or Peroutka), the precincts have to be two or three apart, or a number different modulo 5.  For example, 4F and 4N are 8 apart (8 modulo 5 is 3).

If the precincts were one apart, there would be some Kerry and Bush swaps (in the case of the Kerry for Peroutka and Badnarik situations, there is also the possiblity of Badnarik and Peroutkas votes going for Kerry.  It doesn't happen of course, since Kerry gets 90%+ of the votes, and Bandarik and Peroutka got close to zero).  I would expect that it might be slightly likelier that consecutively numbered precincts would be sharing polling places.  But perhaps not, since it is difficult to map a linear sequence to a 2-D space.

Since the election law doesn't really anticipate holding elections from multiple precincts at the same place, the only real risk from rotation is printing error, (a candidate is left off some ballots), if we assume that voters can read names.  (Note that it was suggested that in 2000 in Florida, that some voters were confused by the ballot order.  Florida law specifies that the party of the governor appears first, in all precincts.  Jeb Bush was the first Republican governor in over 100 years, and so for the first time in membory of any Florida voter, the Democrats were not first on the ballot).

Ballot rotation may be a good idea that only fails when several coincident things occur (multiple precincts at one polling place, use of punch cards, and failure to get the voter to use the correct polling booth).

You must be logged in to read this quote.

But that person should bloody well know well enough he can't do that, shouldn't he? Any poll worker not remembering a thing like that (of course he's hypothetical) basically belongs prosecuted.[/quote]
In this hypothetical case, the voter wasn't paying attention.  So he wasn't really aware that the poll worker was pointing at an empty  booth among a particular set of voting booths.  The poll worker was then distracted by a first time voter who needed assistance in placing the ballot in the punching device.   As he did, he repeated his instructions.  Since _he_ knew about the difference in the voting booths, he assumed that the voter did as well.   Or the poll worker may not have been given instruction in the matter - or perhaps he was, but it wasn't a point of emphasis.  The point of emphasis this year may have been to get the correct number of ballots in the right box, and to handle provisional votes correctly (they go in a yellow privacy envelope before being deposited in the ballot box, the regular ballots go in gray).

[Will F.D. People]

(Note that it was suggested that in 2000 in Florida, that some voters were confused by the ballot order.  Florida law specifies that the party of the governor appears first, in all precincts.  Jeb Bush was the first Republican governor in over 100 years, and so for the first time in membory of any Florida voter, the Democrats were not first on the ballot).

Republican Bob Martinez was governor 1987-1991. Republican Claude Kirk was governor 1967-1971. Small flaw in an otherwise brilliant analysis.

[Andrew]

That's some great research, jimrtex.

It points out another of many problems in our voting system.

It's true that you need multiple things to happen in order to run into this problem, but it's also true that none of these things are unusual.