Electoral-Vote.com - Is it a Pro-Kerry website?

Electoral-Vote.com is projecting that the final electoral vote will be 538 (Kerry) to 0 (Bush). Are they secretly (or openly) a pro-Kerry website? The data they have doesn’t seem to support their conclusion that Kerry will win a lot of the states they have his winning.
http://www.electoral-vote.com/fin/oct07p.html

I think that site has either been hacked or its experiencing a malfunction because it actually showed Bush having the edge in electoral votes last night.

cj

What? it currently says:

http://electoral-vote.com/

Click on the link for projected final results.
http://www.electoral-vote.com/fin/oct07p.html

I am sorry, that should read projected final map.

:eek: That’s bizarre!

The webmaster has explicitly said he’s a Democrat, but his coverage has striven to be balanced, IMO. I’m not sure what to make of this.

Daniel

To get to the map Kel Varnsen - Latex Division linked to, click on “Projected Final Map.” I’m not sure what that means, but I noticed that the projected final shows Texas as 98% Kerry, 0% Bush, and 1% Nader. There must be some malfunction. There’s no way Nader would beat Bush in Texas! (Although I would probably hurt myself laughing if he did.)

Looks like that is just a default setting they use as a template before putting in the results of the actual polls as they come in. It could easily have been defaulted to 100% Bush. I hardly see it as in any way as indicative of partisanship. YMMV.

I’d bet it’s simply a coding error.

Its commentary is as pro-Kerry as that at electionprojection.com is pro-Bush, but both sites report poll results accurately. They do use different polls, and electionprojection is not updated daily, but it’s worth checking both places for a clearer picture.

Apparently Kerry has one hell of an October suprise. Maybe pics from Bush’s and Osama secret love nest. Anyways, I’m impressed that the electoral-vote algoritims are able to predict such a factor.

Seriously though, the site claims that it draws the least squares regression through the two candidates poll numbers. While I suppose that if Bush’s numbers have been trending downward in every state, even in states like Texas where he’s ahead a considerable margin, then projecting out might cause the computer to think he’ll be at zero in every state by Nov. 2

If you click on say Alaska, though, it looks obvious from the graph that the opposite is true but Bush is still given 0%. So I’d say that its broke, which is a pity since it would be interesting to know what the least square regression would be.

If you click and hold the “Depress for current map” button – you may have to click several times or hold the button down, it’s a bit fiddly – you get today’s results instead of projected, which shows Kerry 253, Bush 264.

Something is obviously broken, I’ve seen that site working correctly in the past.

From their Welcome Page:

Well, yesterday’s projected EC votes show a pretty substantial win for Bush. It should be noted, though, that the predictions here appear (based on te graphs that come up when you click on a state) to use data from all the way back in the June/July timeframe. What with the post-convention and post-debate poll volatility, I’m skeptical of the meaningfulness of these projections.

Yeah, I’ve never taken electoral-vote.com seriously. The “votemaster” uses a dodgy methodology to issue his predictions. His map carries a mixed bag of numbers from different polls, so it’s like apples and oranges. One day he’ll report Survey USA saying Bush is up 6 in Wisconsin, then a couple days later he’ll report Zogby saying that Kerry is up 3 in Wisconsin. That’s not a nine point leap for Kerry! Those are two different polls! His whole map is like that.

This projected final map showing Kerry winning 538 electoral votes is based on that sort of counting. His “final projections” are based on what would happen if his numbers kept changing at the same rate they were since the last change. So let’s use that Wisconsin example I just made: it looks like an increase for Kerry, but it really isn’t. Sure, Kerry may actually have gone up in most polls in Wisconsin, but that doesn’t guarantee he’ll keep going up by nine points or even two or one point every couple of days. Most polls agree that Kerry leapt up in the polls as a result of the debates, but will he keep leaping up by the same amount constantly until Election Day? Well, as a Kerry supporter, I hope so, but I’m sure that won’t happen.

If you want accurate numbers, pick one poll and stick with it! It’s important to follow as many polls as possible, but to remember that the data from these polls mix like pickles and ice cream. It’s noteworthy that most polls show Kerry trending upward, even though these polls will differ as to how Kerry is doing in any given state.

Looks like he’s working on it as we speak. The number is still 538, but he’s working on getting the colors on the states.

I’ll give the guy a break for sleeping in.

He admits that he’s a Democrat.

It’s clearly just a coding error. They’ll get it fixed.

Well yeah, but those different polls are supposedly measuring the same thing, so at least in theory this technique should be valid. Of course polling methodology is sketchy and supposedly getting worse, and the guy that runs the site has been quite candid about that. Nonetheless if he just used one polling firms numbers, its dubious that they would generate enough numbers frequently enough, especially in non-battleground states, to get an up to date feel for how the election is going.

Basically, its one technique to get a feel for how the election is going. No technique will be full proof except the poll that happens on Nov 2 (and even that one is far from flawless).

I agree that this technique is sketchy, but its not that bad. The site is obviously broken. Theres no way that Kerry has been trending upward at such a rate as to
overtake Bush in every state in less then a month.

There is no debate here. The site is run by a Kerry supporter and the aspect pointed out by the OP is clearly a mistake in the coding. Perhaps the OP might want to ask the mods to close this thread or moce it to GQ.