This is what I was talking about in the OP. You can certainly get much more accurate in the counting of ballots than .01%
However, you cannot design an election system that will allow a voter to accurately record his desired vote with 99.99% accuracy or better. Frankly, I don’t think you can get 99%.
It doesn’t matter how well the votes are tabulated beyond that accuracy level. You’re measuring something that doesn’t exist. Once you go beyond that level it’s a matter of luck who wins.
Just throw some facts into the debate, in New York (and many other places), we vote by using electro-mechanical voting machines. These machines have switches which clearly indicate the candidate selected and prevent voters from voting for two opposing candidates. They total the votes internally, and do not provide for a physical ballot. Each machine is independent of each other. When the election is completed, the machine is unsealed, and totals on the counters in the machines are manually recorded by election workers. The election totals from each precent are forwarded to the county election headquarters.
In a recount, the totals on the machines are rechecked, and the sums are recomputed to make sure there is no human error in the recording and totalling of the votes. Also absentee and other paper ballots (submitted in the case of disputed voter registration or other similar cases) are totalled.
These machines have been used for at least 50 years, though they have stopped being manufactured. I believe that the reason that manufacture was stopped was because of the expense of these machines compared to cheaper methods like punch or fill in the blank type ballots.
Scylla
Apparently I misunderstood your OP. I thought you were making reference ot the issues driving the current requests for manual recounts in 4 Florida counties. Those efforts are driven by concerns that the machine counts have not accurately or reliably recorded valid ballots.
I agree that there is no way to ensure that a percentage of people will mistakenly mark their ballot on election day.
That’s right. At last report, on CNN.com, Gore was leading NM by 173, which is closer than Bush’s lead in Florida (961, pending all the re-counts and the absentee ballots, which do not have to arrive until November 17).
We are an impatient culture, used to immediate gratification.
Latest word is that Bush’s request to stop the hand counts asked for by Gore will not be heard until Monday. (But won’t the PBC re-count be over by then?)
TRIVIA: The largest user of punch ballots in the USA is Los Angeles County, which is also the most populous county. (NYC is spread over five counties or boroughs.)
Do you realize how computer illiterate most people still are? I hate to use an example of one, but my mother, bless her, is still befuddled by ATMs, VCRs and CD players. It was a major task for her to learn how to use the TV remote! It’s true that about half of all American homes are now on the WWW, but I’d bet that only one or two persons in each of those homes is proficient at using it. I’m the ONLY person in my immediate family who has ever used a personal computer. Most of my relatives have never used one either. We may have to wait another generation before enough people are computer savvy to use computer voting.
There were a lot of odd goings on during the election, and it was close. Not counting absentee ballots, and let’s be strictly nonpartisan here, but Florida was very close.
There is voter error, machine error, the potential for impropriety the ealry call by the media, etc. etc.
Without holding a revote, is it possible to ascertain what the results of this election in Florida actually were?
It’s so close that we are outside our margin for error. I don’t think we can get an accurate result at this point.
Exactly. That’s why the only fair solution is to split Florida’s electoral votes between the two.
I would like to see this done in every state. Statisticians and other assorted experts can determine just how accurate the election is in measuring the people’s will, and anything inside that variance is called a tie.
To do otherwise leaves the result up to pure serendipity, which is what we’re likely to see in Florida.
Seriously, you could award them alternately–as long as you have an even number of odd-numbered states splitting, it isn’t a problem. If there’s still an odd one, and it doesn’t make a difference (as, at the moment, it wouldn’t in Florida), don’t worry about it.
I don’t think you could mathematically end up with a tie and a single vote left over, since there are 538 total; at worst, you’d end up 269-268. So the easiest thing to do in that case might be to throw out the last odd vote. Alternately, you could award it to the winner of the popular vote.
Los Angeles County has ATM style voting machines. For the last ffew weeks you could go to certain areas and vote at a kiosk. (I did not do this so I am uncertain about details.) One of the advantages is that the ballot is in about 5 languages.
I do not know what prevents you from voting at the kiosk in the morning and then running down to your polling place and voting there, too.