New Poll on TiltMeters

[Muhammad]

Hmm, yes, I see how that could happen if your ISP assigns a new IP.

I'll see if there is any way to double-up the authentication on the poll to track by another method...

 

[Bird]

Well, Muhammad, you know I've just got to try it. When I went back to the home page before, I had not logged off after voting, so just now I tried it again, and I think I just got to vote a second time./w3tcompact/icons/smile.gif I hadn't thought about these dialup connections getting a different IP address (of course, as you know I don't understand these contraptions very well anyway).

Bird

 

[Muhammad]

Well I'll look into soemthing that might not let y'all with dial-ups vote multiple times, for sure now...

 

[scruffy]

Muhammad, I too suffered the 'Just Had to Try' syndrome, and there is one more vote to 'erase'! Suggest one per email address?
Specifically an ISP email address, not HOTMAIL or one of the freebie email addresses.
<P ID="edit"><FONT SIZE=-1>Edited by scruffy on 11/2/00 07:09 PM.</FONT></P>

 

[Muhammad]

Yeah, we're going to make some changes that should eliminate 99% of the double-voting issue.

Did you see what I mentioned in the feedback forum about in-forum polls that would only be opened to registered users. With those, the only votes you'd see would be from registered users.

 

[scruffy]

Muhammad,
yes I did, and the only suggestion would be for a loop to check for a previous vote by the user, if it finds a previous vote, then kick them back to the login page w/o the opportunity to vote again. At least that would insure that the current login would not have opportunity for a double vote the next time the user logs in.
Have fun!!

 

[Muhammad]

Right, it would be 1 vote for 1 registered user as a maximum and easy to limit them if only registered users could vote.

 

[PaulB]

Muhammed - i don't know what the hell an ISP or a dialup or any of that stuff is, but I too had to try it, and I think I got to vote twice also. Maybe you will need to negate the current vote, and iron out the kinks to avoid the potential for tampering. I guess pinky swearing not to vote more than once might not cut with the oem gurus, would it?

 

[Muhammad]

Yeah, if you notice, the poll kind of doesn't work right now... /w3tcompact/icons/crazy.gif

We're making some changes so the poll right now... well, actually now fixing it I think. /w3tcompact/icons/wink.gif But once it is stable we'll remove the double votes and then it should be set. I'll let you guys know when to test it out.

 

[Ibrahim]

The end of Ballot Stuffing

A couple things:

1) In the past, we've kept a log of every single IP address that's voted for a poll. And while I noticed more than one user voting more than one time, it hardly swayed the vote one way or the other (Sure, when it first starts out, 1 vote will make a big difference). Like others have pointed out, even with the "Ballot Stuffing" the poll is probably more accurate than the polls for presidential races (where likely every vote is decided by green ink). It's not meant to be a scientific system of voting, but rather a taking of the "pulse" if you will of the site visitors.

2) The double votes you "hooligans" (just kidding) put in to "test" have been removed. But don't worry, it was easy, no harm done.

3) I added more security. Namely a barb wire fence and junkyard dog. (Actually, I rewrote the checking routines so that now, even if your IP changes, you can't vote again.)

4) Even tho I re-wrote the "security" it's still not perfect. People basically get one vote per computer. And I know a couple MORE ways around it (but I'm not telling). One or two (or ten or twenty) extra votes really WON'T make much a difference. TBN gets about 25,000 unique visitors a month (though not all of them vote, and not all of them post). Just vote to your hearts content and know that I've got it pretty much tightened up. (If it should break, it's definately NOT my fault. )

That is all.

Ibrahim