Electoral-vote.com misrepresents the truth, and here is why.

[abrock]

Edit: This post is not meant to deterime the real truth, only that it is not being represented by electoral-vote.com

This is something I've got to get off my chest, because the site administrator won't listen to me, and he is doing a horrible job trying to display who is in front of the presidential race. My opinion has nothing to do with who is in front at any given time, in fact it is just the opposite: He doesn't show who is really in front. Let me explain.

I've written directly to the administrator of the site, requesting that he use a little common sense in his addition of new polls to the map. We all know that some polling organizations are extremely partisan (Mason Dixon, Strategic Vision, for example) and others are just downright incompetent (Gallup). For those that do not know the methodology that Electoral-Vote.com uses to display results on the main map, let me explain:

He takes all new polls, from ANY polling organization, no matter how partisan, and adds them as 'the latest word' whenever they are released. So, purely as an example, the Zogby poll which came out yesterday (which is usually the best look at what is really going on) shows Michigan as going to Kerry by 6 points. Got that? If a Strategic Vision poll (Republican poller) comes out today shows for Michigan, it goes on the map, no matter if it shows a similar result or a completely inaccurate score like Bush being up 26 points.

In his attempts to remain non-partisan, the webmaster of Electoral-Vote.com has ceased to use the brain he was given at birth. Common sense is the best method to determine which polls should be used and which should not. I normally would not care, but this guy has been buying tons of ad space on college webzines and such, with the tagline "Kerry? Bush? Who's ahead? Click here to find out!" it is clear that one will not really "find out" at Electoral-vote.com because of a clear lack of common sense. In fact, misleading people in this election is even more dangerous than ever.

My conclusion: Completely ignore the site. I hate to say that, because it has a nice graphical map and the daily editorial is interesting to read, but the substance of the site will cause you undue grief because the true state of the election is not represented, and you can quote me on that.

[Bleacher Creature]

1. I couldn't have said it better.

It's so frustrating to see people getting so worked up over the daily "changes" on this thing.

[9119495]

2. His stuff on the senate races is good so I like it.

He has been doing a good enough job and he has pointed out that Strategic Vision is a republican firm. If there were Dem firms releasing their numbers, those would be used as well. I would like some suggestions about other sites that show the Electoral map and how it is arrived at. (E-V.com has the poll and percentages pop up over the state).

We need to use these sites, not the national polls. It will be all about the EV.

[abrock]

4. Look, I don't care if he would use (d) polling organizations as well.

That was exactly the point: Why use partisan polling organizations when you are trying to paint a proper picture of the race? It does NOT make sense.

[pmbryant]

3. You are absolutely right

There are many possible interesting and possibly valid choices of methodology for determining who is likely to be ahead in a certain state, given the recent polls taken there. The one used by that site is neither interesting nor valid.

--Peter

[Jacek-t]

5. Why? Because it doesn't show John Kerry in the lead?

And by the way, how you determine that a particular poll is partisan? And which polls are Democratic partisans? This guy simply reports the most up to date information, clearly stating the sources of his information.

[NinetySix]

11. It is useful.

Although he does use partisan polls, you can go by those results if you take them with a grain of salt.

See that NJ is close with Bush 1 point ahead. Now notice that the latest poll is Mason-Dixon. Now look at the graph showing the trend, and perhaps more importantly, the most recent Zogby poll for comparison. Often times you'll find a zig-zag up and down pattern, with Zogby's numbers above and other, "independent" pollsters below.

When a polling firm like Strategic Vision is used, he puts an (R) beside the name to show that it is partisan, so as not to misrepresent it. And if you look and see that Strategic Vision has Kerry up by 4 in Michigan, for example, it is instructive in that you can count on the lead being greater than that.

I look at it first thing in the morning every day, but I don't pretend that it's the final result, or even terribly accurate with regard to the actual state of the campaign. Remember, polling firms have their own standards as to the weight they assign to voters, and their results vary because of the differences in their standards, but additionally, few or none take into account newly registered or overseas voters.

Read it like me, and take heart -- it underestimates what we're gonna see in November. 2004 ain't 2000.

[pmbryant]

14. Yeah, that's it!

No.

The methodology of just picking the last poll, regardless of source, and using that as an indicator of who is going to take the state, even if the margin is only 1-2%, is inherently flawed. (Far better would be to take some kind of average of recent polls, and only award states if the margin is outside the statistical error.)

That makes it invalid, and anything invalid is not interesting. The best that can be stated is that it is misleading.

But Bush supporters can go ahead and get overconfident if they like.

Peter

[LuminousX]

16. Reputation is key

some polling entities have a better reputation for being more accurate than others.

[pmbryant]

19. Even reputable firms are subject to statistical error

Merely the statistical error on the margin between the candidates is 5-7%. Individual polls by reputable firms should bounce around by that amount. A single poll with a margin of 1%, 2%, 3% is essentially meaningless.

Of course, I'm not even mentioning sampling error, which can lead to all sorts of biases.

--Peter

[Jacek-t]

6. Re Gallup poll

If Gallup is so incompetent , why has John Kerry agreed to this organisation doing a polling to select voter participaiting in the second Presidential debate?

[DaveinMD]

8. Simple

Because we wanted to make sure we got that town hall debate.

[Walt Starr]

7. I LOVE electoral-votre.com, and here's why

He's using the bogus results and that motivates me to work harder. When it's all said and done and Kerry starts coming back in all of the polls, it will show up.

[DemocratSinceBirth]

9. I'm Not A Big Fan Of Gallup And Strategic Vision

but I see nothing wrong with Mason-Dixon's work....


If you can show me a final pre election poll by Mason Dixon which varied greatly from the actual results I'll revisit my opinion....

[HFishbine]

10. At the very least

He could use a trailing average.

[JerseygirlCT]

12. So they're not weighted at all?

[quaoar]

13. I like the site

Yes, it presents a distorted view of the state of the race.
But I don't look at it to see who is winning. I look at it to see what the latest polls are saying in each state.
I don't believe for a minute that Bush is leading in NJ or tied in Maryland. And the site admin doesn't either. He often points out that this or that poll is baloney.

Another thing I like about that site as opposed to others is that he "awards" states even if Bush or Kerry is only 1 point up. Other sites list so many states as toss-ups that there isn't much sense in checking it out.

It's fun to check the map from day to day and laugh at stuff like NJ and Maryland. But I wouldn't take that stuff too seriously.

[geek tragedy]

15. Start your own website then.

His approach is very simple, and he doesn't try to determine which polls have an agenda or bias. That beats the approach some here have of slamming any poll showing bad news.

[Jacek-t]

20. Yeas, exactly! Couldn't agree more.

[Name removed]

21. Deleted message

Message removed by moderator. Click here to review the message board rules.

[abrock]

23. Why can't I criticize an established and large website that is dead wrong?

"Start your own website" is not an acceptable answer. If I don't like what CNN is saying, I will complain about it. They are the outlet representing the issues taking place in society, and we are the people who make sure they don't feed us propaganda.

Unfortunately, most of us aren't doing our jobs, and you see the result.

[geek tragedy]

29. Virtually anyone can start their own website.

Very few people actually know a good polling firm from a bad polling firm.

No one was bitching about that site when it showed Kerry with 327 EV's.

[abrock]

32. I was. I just wasn't part of this forum at the time.

[Claire Beth]

17. The webmaster for that site....

is a KERRY supporter. Personally, I like the site. Everyone already knows the polls can't be trusted and the webmaster at electoral_vote.com has even stated this fact. I have emailed him a couple of times and he seems very reasonable to me.

[NewYorkerfromMass]

22. Well, I find that hard to believe

Look at this state graph for Ohio.
The trend lines make no sense.
Why did he pick the later part of June to begin them?
Why does he weight the outliers the same as the others?

http://www.electoral-vote.com/states/ohio.html

Looking at this graph shows me B/C are fighting very hard just to hold Ohio.

[Claire Beth]

26. What is it you find hard to believe.????

that he *IS* a Kerry supporter? I have emailed with him several times and know that he is. Also, check this link out and scroll down appx 1/2 and read for yourself. Just because the polls are all screwed up AND screwed up EVERYWHERE, not just this particular site....Just because things don't always look in our favor doesn't mean the messenger is on the opposing side. Be sure to read this link:

http://www.electoral-vote.com/info/welcome.html

[NewYorkerfromMass]

27. Well, No. 1, they could avoid SUSA and other pro-Bush polls

and No.2, as I tried to illustrate in my trend line critique, the weighting of most recent polls no matter where they come from is unbelievably misleading.
If they're trying to promote Kerry they could do MUCH better.

[Claire Beth]

28. I appreciate his efforts...

and think it's a fantastic site. Since none of the polls are completely accurate or dependable if they upset you, you might should avoid looking at them.

[Mayberry Machiavelli]

34. No I think their point is that the site administrator is pro-Kerry, NOT

that he is trying to use the site to promote Kerry. There is a difference.

[foo_bar]

18. the true state of the election is never represented

the true state of the election is not represented, and you can quote me on that

per Heisenberg. All polls are push polls, regardless of intentions.

[NewYorkerfromMass]

24. Look at this trend line for Maryland

http://www.electoral-vote.com/states/maryland.html

The site is a waste. With a capital W. Like the "prez".

[VirginiaDem]

25. I agree generally--why not try www.race2004.net?

www.race2004.net does a much better job IMO. They show all the state poll results, with and without Nader, and they keep battleground states undecided.

[abrock]

33. race2004.net sucks, too.

They've got nice poll listings, which I like, but they are clearly wrong on their summaries. Oregon is undecided? Yeah right! NJ is undecided? HAH! NH is fucking undecided!? Give me a break.

[liberalpragmatist]

30. He Should Also Average the Latest Polls

You're right - he should ignore the partisan polls and rely on the nonpartisan ones. To be fair, I don't think he can discount Gallup because they are nonpartisan and they have a good history - just really wild swings. He should use Gallup, but adjust the results based on party affiliation and use the RV samples, not the LV results.

And, as is the subject of my post, he should use a formula to average the latest nonpartisan polls when applicable if there's overlap between the dates.

[featherman]

31. Actually the map is pretty much correct today if you make the

intelligent and obvious adjustments. Add MD and NJ back to Kerry, both of which are slam dunks. (The two listed polls are bogus and please take notes of who they are because they are NEVER to be trusted with results like that. They are just insulting the intelligence).

These leaves Kerry with his base of 264 which is about where he is.

What's the problem? The site author states that FOR HIS SITE, the latest poll wins. That's his formula and I find it useful. That is the information he is providing not some prediction about how it will turn out. Get a grip, folks... they are just polls, not infallible Papal pronouncements...all are flawed and some way more flawed than others.

[Jane Austin]

35. In the "More Data" section, you can just use the polls

you want to look at.

In other words, he has a bunch of different maps with just one polling outfit per each.

[obiwan]

36. I asked the guy who runs this site what his qualifications were...

... he's an amateur. No professional polling experience, and a registered Republican, to boot.

Strictly an armchair hobbyist. His methodology is best described as "pretzel logic."

Kinda like the mechanic who sells you a new radiator when all you need is a little anti-freeze.

100% pure unadulterated bullshit.