The ‘Hot or Not’ Solution

Odd that it didn't post the end of my comment. Here's what my last sentence should have been: And the more likely for the vote to be distorted in weird ways by the presence of candidates that can't win but take away votes from the people who otherwise should have won.

Unfortunately the spoiler effect happens under Range voting as well.

Look at 2000. Most people agree (well, most people not blinded by party loyalty) that more people preferred Gore to Bush, and that Gore would have won in a race against Bush alone.

In a Range election, if only Bush and Gore were running, most people would give Bush zero and Gore 100, or Bush 100 and Gore zero. Even "honest" voters would do that: since the ratings are relative to other candidates, it is expected that voters will "stretch" their ratings to cover the full zero to 100 range.

Now, enter Nader. Honest voters who like Nader best and Gore second best would now lower their ratings for Gore from 100 to some value representing how much they like Nader more than Gore. Bush, however, stays at either zero or 100 on most ballots (except on the ballots of those rare people who like Nader best and Bush second best). It is possible that having Nader on the ballot would raise Gore a bit on some ballots of Bush voters, but it seems likely that most Bush voters would leave Gore at zero even with Nader in the race. This gives a significant advantage to Bush against Gore, because the 3rd candidate's platform is perceived as closer to Gore's than to his own. Vote splitting, plain and simple.

If Nader actually got popular enough so that it was a close 3-way race, the situation would get worse. Even those voting purely strategically would end up punishing Gore for Nader's entry into the race. If a Nader fan thought that Gore was in first place, Nader in second, and Bush in third, they would be tempted to "bullet vote" for Nader, not approving Gore even though they prefer him to Bush. Enough people doing this would hand the election to Bush, when he clearly is not preferred by most.

Unfortunately the spoiler effect happens under Range voting as well.

Look at 2000. Most people agree (well, most people not blinded by party loyalty) that more people preferred Gore to Bush, and that Gore would have won in a race against Bush alone.

In a Range election, if only Bush and Gore were running, most people would give Bush zero and Gore 100, or Bush 100 and Gore zero. Even "honest" voters would do that: since the ratings are relative to other candidates, it is expected that voters will "stretch" their ratings to cover the full zero to 100 range.

Now, enter Nader. Honest voters who like Nader best and Gore second best would now lower their ratings for Gore from 100 to some value representing how much they like Nader more than Gore. Bush, however, stays at either zero or 100 on most ballots (except on the ballots of those rare people who like Nader best and Bush second best). It is possible that having Nader on the ballot would raise Gore a bit on some ballots of Bush voters, but it seems likely that most Bush voters would leave Gore at zero even with Nader in the race. This gives a significant advantage to Bush against Gore, because the 3rd candidate's platform is perceived as closer to Gore's than to his own. Vote splitting, plain and simple.

If Nader actually got popular enough so that it was a close 3-way race, the situation would get worse. Even those voting purely strategically would end up punishing Gore for Nader's entry into the race. If a Nader fan thought that Gore was in first place, Nader in second, and Bush in third, they would be tempted to "bullet vote" for Nader, not approving Gore even though they prefer him to Bush. Enough people doing this would hand the election to Bush, when he clearly is not preferred by most.

misterharban:
There is strong evidence that a switch to Range Voting would lead to a much larger improvement in our democracy than the elimination of election fraud. See: http://rangevoting.org/RelImport.html

And I've done exit polling in Beaumont, Texas. Average middle Americans had no problem with it. In fact it experimentally leads to fewer spoiled ballots than our current voting system. So in some sense, voters find it easier to use, not harder.

tomruen:
You have it completely backward. Strategic voters are also rewarded in IRV and other systems, and in fact the result is even _worse_. This issue is given rigorous consideration by Poundstone and proponents of Range Voting. See http://rangevoting.org/StratHonMix.html

The simple fact is that IRV is one of the worst voting methods, according to extensive Bayesian regret calculations by Princeton math Ph.D. Warren D. Smith (and those same metrics show Range Voting to be effectively the best).

Range Voting is also cheaper/easier to administer and count, and does not incentivize the adoption of fraud-prone electronic voting machines like IRV does. And it can be counted in precincts, whereas IRV requires central counting, which is another open invitation to a central fraud conspiracy, and a loss of transparency.

Range Voting is just a hands down radically better system than other methods that have been proposed. It's high time this simple and brilliant reform came to light.

Range voting/Approval voting are GREAT REFORMS for questions like inclusion in political debates, where voters want to hear from more than one candidate, but seems a non-reform for picking a single winner were voters are penalized by honest ratings - merely helping a second choice beat a first.

IRV and STV (Single Transferable vote for multiseat elections) are the way to go - to maximize the number of voters influencing the winners.

Range voting/Approval voting are GREAT REFORMS for questions like inclusion in political debates, where voters want to hear from more than one candidate, but seems a non-reform for picking a single winner were voters are penalized by honest ratings - merely helping a second choice beat a first.

IRV and STV (Single Transferable vote for multiseat elections) are the way to go - to maximize the number of voters influencing the winners.

All equipment, and soft ware should be standardized by the federal Government. And should be fully open source. So that the public, and global computer community can inspect the system for security weaknesses, and back doors on an ongoing basis. No closed proprietary systems should be used. Your just asking for vote fraud that way.

Lastly whichever paper ballot system is used. Every paper ballot should be stamped by the voter with a unique digital signature before they cast that ballot. The machine that stamps the paper ballot should print three paper receipts. One for the voter. One that goes into a lock box under the custody of the federal government. One receipt goes in a lock box under the custody of each party. With a different lock on the box for each party.

A representative sampling of the digital signatures should be checked against the digital signatures stamped on the paper ballots in the presence of all parties. If they don't match. You have fraudulent paper ballots. And a big problem. This will assure each voter that their vote was counted. And that each vote counted was cast by a individual voter. In other words democracy.

Then all voting systems should be opened to the public with the directive to tear the system apart, and expose any weaknesses they can find. And suggest ways to close the weakness.

Sadly there will not be enough time to fix this disgrace before the elections. So congress should authorize a massive strike force against vote fraud with extensive monitoring of the coming election. And with extensive and open, transparent exit polling to help catch discrepancy's, and fraud. 2 of 2

Technically speaking our disgraceful voting system is very easy, and relatively inexpensive to fix. Though no expense should be considered to much to assure our democracy. And the integrity of each of our votes.

One way to fix it is for everyone to cast a paper ballot like they do in Germany that is hand counted by civil servants over an approximately two week period. Preliminary election results are called the night of the election based on exit polling. The Germans have shown this method to be highly accurate, and relatively secure.

Of course in the US exit polling does not work well. And is used to assist in the commission of massive voting fraud. As was done in our past two national elections. In the US we use the highly accurate exit polling to determine how many pre-punched fraudulent votes need to be substituted for legitimate votes for the loosing party to win.

The other way is to use electronic means to produce, and mark a ballot. But regardless of what electronic system is used. It is critical that the end product be the creation of a paper ballot that can be easily reviewed by the voter before that paper ballot is cast by the voter. So that the voter can change, or discard that ballot if it was not marked as they wanted to vote. Ideally the paper ballot would be hand counted as in the German example above.

If the paper ballot is machine counted, then random hand counting of a representative number of each batch of votes must be done to assure the accuracy of the electronic totals. Plus very careful exit polling. If there is any discrepancy, all paper ballots should be counted by hand. And electronic equipment, and software checked for failures. 1 of 2