The "potential for voter error" index used here seems like it has some flaws -- it's partially useful, but also seems to have some problems.

Currently, the accepted way to measure this is by the residual vote rate (a rough, approximate measure of the number of votes lost due to voter or machine error). The data I've seen suggests that the residual vote rate for precinct-count optical scan machines is about the same as the residual vote rate for modern DREs ("electronic"). These two systems are the least error-prone of any widely used voting technology (they're both better than level machines, punch cards, older systems).

Therefore, I believe there is no good reason to list "electronic" machines as less error prone than "opscan". They are comparable.

There is an important distinction to be drawn between precinct-count opscan and central-count opscan. The former is what is normally used for paper ballots cast at a polling place; the latter is what is normally used for absentee ballots (e.g., postal votes). On average, precinct-count opscan systems have measurably lower residual vote rates than central-count opscan systems. The advantage of precinct-count opscan is that it gives immediate feedback to the voter that allows the voter to correct some kinds of mistakes in marking their ballot. (There is wide variation in the residual vote rate for central-count opscan systems, depending upon the procedures in use at the central counting facility: the best election officials set up their central-count opscan machines to flag any ballot that contains overvotes or ambiguous marks and have their staff inspect, by hand, every ballot that is flagged, to ensure that voter intent is properly captured -- but even then, it's not clear if you can get the residual vote rates of central-count opscan down as low as the residual vote rate for precinct-count opscan.) Absentee ballots are more prone to error.

For DREs ("electronic"), there is also an important distinction between the modern DREs used in most of the country, vs full-faced DREs used in a few east-coast states (e.g., NJ, NY) and the older-generation DREs (which aren't very widely used). The data I've seen suggests that the full-faced DREs tend to have significantly higher residual vote rates than ordinary DREs; also the older generation of DREs had worse residual vote rates than the kind of DREs that tend to be used in most jurisdictions today.

In short, I think this map is a bit of an oversimplification, and may in some cases be a misleading analysis of the relative degree of "error-proneness".

I voted in the Dallas/Irving TX area and used an evote machine...so I find your map inaccurate

Electronic voting may be seen as less likely for a voter to cause an error, but they are far from reliable. So far in West Virginia, voters participating in the early voting option have found that when they touch the screen to cast a ballot for Obama, the indicator immediately moves to show McCain as the selection - and the same thing has been reported as happening with votes for state senators, the Governor and other elections. In each case, the voter has intended to vote for a Democratic candidate, and a Republican candidate was indicated as the recipient of the vote instead. Fortunately, those who've noticed the problem were able to correct their votes before the final submission, but there's no way to know how many voters may not notice, or may not be familiar enough with touch-screen voting to understand that their vote was incorrectly recorded.

And before anyone thinks I'm trying to say that this proves Republicans are conspiring to steal yet another election, I'm well aware that the problem may be one of calibration, not malfeasance. If the touch-sensitive areas of the screen are not properly aligned with the candidates they represent on the ballot, that, too, could explain the problem.

Whether it's intentional sabotage or a mechanical issue, however, it means that votes on those machines are inherently unreliable, and that's the LAST thing we need.

Personally, I actually feel lucky that my disability means I have to use an absentee ballot via mail. I feel like there's a better chance that my vote will at least be correctly counted.

Electronic voting machines may have great potential to improve accuracy in the future, but right now, they're more likely to cause problems then solve them.

This all gets me wondering how prevalent the hacking/ stealing of votes is in our current times compared to the stealing/political party machinations of previous years (I grew up in the NE and we were pretty good at keeping parties in office). You might have just given me a topic for my research paper. Awesome map and stats.

Your data for Washington D.C., the District of Columbia, is inaccurate. We have both electronic and optical scan voting.

Residents of the nations capital are also disenfranchised in that we are not represented by Senators or a Representative with full voting rights in the US Congress. I'd ask that others support the District of Columbia receiving full representation in the US Congress.

Like others above, I'm reporting that your map is inaccurate. I live in Brazoria County, TX, and I will be voting on an electronic machine here just as I did in the primary and in last year's local elections. Your map shows optical scan.

To CMCGRAW:
What you ask for is only solved by a constitutional amendment. DC was never meant in the eyes of the founders to be more than a seat of power for the country where all of the seated congress people could live while they worked in office in the Capitol, thus removing the chance of the congresspeople getting a SECOND vote on-top of the one they already cast in their respective house. You were purposely disenfranchised because you are the seat of government. The moment DC carries representation, the rest of the country becomes disenfranchised by the seat of government getting an extra 3 votes total.

Regarding the voting in California: Depending on county, both opscan and emachines will be available for voters. Paper ballots may be requested only at the county's registrar beyond the opscan paper.

ABSENTEE VOTING
Check with the Registrar of Voters in the county in which you reside. Many will allow you to register permanently as a Permanent Mail Voter. You do NOT have to be disabled to become a PMV. It is quick, simple and easy to vote this way and, yes, your vote is ALWAYS counted. I live in Alameda County, California where is no longer called "Absentee Ballot". It is simply "Vote by Mail".