Question on EVoting for MSAccess folks

Not being an Access kind of person (SQL and Oracle being my "thing"), I have a question with regards to the ballot setup on electronic voting systems.....

Let's say there's a "typo" in the database and the file says candidate=KerrEy instead of Kerry....

So when the software sends the instructions to store the vote and it captures it in the "Kerrey" record instead of the "Kerry" record, one may never see this, correct? The system appears to be functioning correctly, yes?

Let's assume the "Kerry" record was predefined with x votes and only the "Kerry" records are printed on the results reports.

MSAccess would never return an error, would it? Not if the database were designed the same for Touchscreens AND Optical Scan ballots, correct?

Do you see where I am going? Is it possible?

but there are simpler schemes.

To return ="Kerry" instead of like "Ker"% (Access is built on MySQL, very similar to regular SQL) then yes, you would see a number returned for Kerry votes, not Kerrey votes. Are you basically suggesting that a # of Kerry records are pre-populated, and the actual voting is recorded as Kerrey votes? Because I'd imagine that it's all numbers-based (faster to query, etc) than text-based, with reference tables to link the numbers to their text equivalents. You could still do the same thing... in fact, there are any number of nasty things you could to to an unsecure, unverified database manipulation program...

... I believe that the Diebold MS Access schema stores it not by label, but by candidateID.

BTW: the early Cobb County 2000 configuration had Mickey Mouse and George Bush running (suspect that it was a training exercise).

Shows a separate database showing candidateID=candidateName.

Is that not the case?

So if Kerrey=candidateID=6
and Kerry=candidateID=5

and I tell the database to store in one place, print from the second, haven't I successfully manipulated the votes without anyone ever being the wiser to my having done so?

... and also could well be done by precincts.

One with any sense would set up a table with a column for each candidate. Then you would either insert a record with a value of 1 in that column, or if you are aggregating just increment it. If by some chance you spelled the name of the column wrong it would generate an SQL error. Access uses the same SQL as MS SQL Server.

If on the other hand you just had a table with a president column and the code had insert a record with the name of the candidate you could get into lots of Matching problems, even Kerry, KERRY, and, kerry would be treated differently if not explicitly coded to ignore case. This would be an extremely stupid way to code the system (so it probably is the one they chose).