And then what happens? Another online poll that might go horribly wrong

It appears some robovoting hit this poll. Robert Phelan’s letter pretty well sums up my thinking on this issue.

Sirs:

By now you must be aware that your on-line Prove It poll was seriously compromised. I voted “count-me-out” once under my own name, but after the individual who corrupted your poll revealed himself, I tested your polling system with two consecutive “count-me-in” votes, which were both apparently accepted.

Leaving aside my distaste for your support of politicized, Lysenko-style “science”, as both a social scientist and computer systems consultant I respect data and am appalled by the shoddy manner in which your organization collected it. A few suggestions:

1. State clearly the purpose of your poll and exactly which data will be used for that purpose.

2. You stated that you would pass the results to the government:

a. if the results had fairly resulted in a “count-me-out” majority, would those results have been passed on?

b. it would be helpful top explain what you would do with the comments you requested from the “count-me-outs”;

c. since the results were to be passed, presumably, to the UK government, foreigners such as myself should have been excluded from the voting. Checking the IP location of voters should be easy.

3. No one, either inside the UK or outside received the follow up e-mail. The explanation provided about ensuring one vote per person, frankly, makes no sense.

4. Maintaining a confidential list of voter names, e-mail addresses and IP’s to verify non-duplication would be easy. Making the voting a two-step process, where the voter had to respond to a follow-on e-mail would be even more secure.

5. Maintaining a list of non-acceptable names for screening: Joseph Stalin, Lenin, Mao Tse-tung and Mickey Mouse all claimed to have voted no, as did Keith Briffa, Michael Mann, Gavin Schmidt and James Hansen.

7. Create a display page where interested persons can view the names who have voted. Given the politicized nature of the topic, a unified alphabetical list would be appropriate.

8. Test the security of your poll before putting it on-line. Find a good hacker and pay him only if he succeeds in breaking into your system.

If you people can’t even run an on-line poll, why should anyone consider your opinions on climate? If this poll was so important that you needed two ministers of HMG to introduce it, why didn’t you get it done right?

I intend my suggestions to be helpful; if you find them so then I would be glad to be of further assistance. I am bitterly opposed to the position you have taken on “AGW” but I would not allow that to interfere with my professionalism.

Oh, one last suggestion. Don’t even try to salvage the results of this poll. Wipe them, make the changes I’ve suggested and start again.

Robert E. Phelan

Adjunct Instructor of Sociology

Business Systems and Automation Consultant

A commenter on our site, “lihard” has seemingly confessed to adding a thousand votes via a script. There was a period of about 15 minutes where the count jumped about 1000 votes. It appears “lihard” was at fault as he pre-announced it here in comments. Of course there was little anyone could do about it. I speak for myself and the moderation staff in saying we strongly object and are offended by his ballot stuffing and want to make clear that it is not condoned in any way. Whether or not the poll was put together with apparently no security in place does not justify any kind of dishonest activity.

However, since that burst (if indeed he, lihard, did one) the vote count has steadily risen, I believe those to be valid. If the Science Museum has any logs, they should be able to filter those ~1000 in question out. I hope they do.

I don’t condone ballot stuffing in any form. Unfortunately it can happen when polls like this one don’t appear to have the most basic simplistic security. The interesting thing here is that if anybody wanting to stuff the poll, no matter what side of the argument they are on, could easily have done so. No special skills are needed to boost the counter…just keep clicking the submit button. Any kid can do it.

Perhaps the Science Museum didn’t think of security for cyberspace like they do for their exhibits. The internet is a harsh place and prone to such things. The lack of due diligence for security is as troubling as the language they used which originally caught my attention.

The polls we do here at WUWT don’t suffer from these problems, as they have anti-ballot stuffing security built in courtesy of WordPress. I hope that the Science Museum will upgrade their poll security if they choose to continue with it. Also for the record, you’ll find me logged once in poll, shortly after posting this story on 11/23 approximately 9:30-10AM PST, with my full name and email address given. If anyone from the Science Museum (or the UK government) wishes to contact me, they can use that email address. – Anthony