PA-SurveyUSA: Obama remains 12 points ahead of McCain

[tokar]

No, 53% came close to losing the state for Kerry.  Bush also gained a bit over a percentage in Phila.

You've seen the post, that is all you will be getting.

He still won it.  Period.  If 53% in the burbs equates to  win, then Obama should have no problem winning PA because he will be besting that 53%.

Ugh...chicken...I get it.  Dont like to have your predictions on a record which can be ridiculed.  I get it...

No, my predictions are posted and were in September.  The link has been posted and I've referred to it repeatedly.  If you can't understand that, then you are just not that bright.

Now, 53% in Montco barely gave Kerry the state, with sizable wins in Delco and Phila.  Obama to carry PA, really needs to do better than 75% in Phila.  I think he will.

I know what you wrote, I've seen what you wrote.  You are predicting a 269-269 tie (Kerry states + IA + NM + CO - NH for Obama, all the rest for McCain).
Its one thing to say it on these forums for the posts to just disappear into the depths of the forum.  Its another to put it in color and make a USElectionAtlas prediction map so people can comment.

Why can't you just admit you comment was inaccurate?  "...just barely over 50% in the Phila 'burbs would probably lead to a loss in the state".  Kerry was just barely over 50% and he won PA, ditto for Gore.

[J. J.]

No, 53% came close to losing the state for Kerry.  Bush also gained a bit over a percentage in Phila.

You've seen the post, that is all you will be getting.

He still won it.  Period.  If 53% in the burbs equates to  win, then Obama should have no problem winning PA because he will be besting that 53%.

Ugh...chicken...I get it.  Dont like to have your predictions on a record which can be ridiculed.  I get it...

No, my predictions are posted and were in September.  The link has been posted and I've referred to it repeatedly.  If you can't understand that, then you are just not that bright.

Now, 53% in Montco barely gave Kerry the state, with sizable wins in Delco and Phila.  Obama to carry PA, really needs to do better than 75% in Phila.  I think he will.

I know what you wrote, I've seen what you wrote.  You are predicting a 269-269 tie (Kerry states + IA + NM + CO - NH for Obama, all the rest for McCain).
Its one thing to say it on these forums for the posts to just disappear into the depths of the forum.  Its another to put it in color and make a USElectionAtlas prediction map so people can comment.

Why can't you just admit you comment was inaccurate?  "...just barely over 50% in the Phila 'burbs would probably lead to a loss in the state".  Kerry was just barely over 50% and he won PA, ditto for Gore.

You didn't read it that well, because my prediction was for McCain to win CO, ultimately.

Because "barely 50%" isn't a number.  In some of the suburban counties Kerry had 57% Delco and won by 42,000 (rounding down), Montco by 46,000, Buck by 9,000, and those are just the inner suburbs.  If Bush lost them, barely, by a few hundred votes, Kerry's margin drops by 97,000 votes.

Kerry only gets 75% of the vote in Phila, he's lost another 38,000.  That is 135,000 of his total.  Kerry had a 145,000 margin in the state, so you are reaching the tipping point.

Than you have to do things like hope Kerry doesn't drop by 2 points in Westmoreland, Erie, Lancaster, Lackawanna, York and Dauphin.  None of those flip a single county.

Just over 50% of the Phila suburbs, coupled with about 75% in Phila would probably lead to a loss of PA for any Democrat.  Now, I think Obama carries PA and passes this benchmark.

[So the Heroes Fall]

Still predicting McCain to win Colorado and ignoring everything in the past month is not principled, it is idiotic. Then again, consider the source...

[J. J.]

Still predicting McCain to win Colorado and ignoring everything in the past month is not principled, it is idiotic. Then again, consider the source...

BRTD, I don't change my predictions.  I regard it as cheating.  Anybody can look at at polls one week and change them later.  I don't ignore current treamds, but I'm not like Zogby, or apparently you, either.

[Grad Students are the Worst]

Still predicting McCain to win Colorado and ignoring everything in the past month is not principled, it is idiotic. Then again, consider the source...

BRTD, I don't change my predictions.  I regard it as cheating.  Anybody can look at at polls one week and change them later.  I don't ignore current treamds, but I'm not like Zogby, or apparently you, either.

I don't understand why one prediction, made from looking at polls (in part, I assume) is superior just because it was made at an arbitrary time.

That just means it's more risky and subject to event changes (more fun!), and more likely to be wrong for self-limiting of data (more useless!)

[J. J.]

Still predicting McCain to win Colorado and ignoring everything in the past month is not principled, it is idiotic. Then again, consider the source...

BRTD, I don't change my predictions.  I regard it as cheating.  Anybody can look at at polls one week and change them later.  I don't ignore current treamds, but I'm not like Zogby, or apparently you, either.

I don't understand why one prediction, made from looking at polls (in part, I assume) is superior just because it was made at an arbitrary time.

That just means it's more risky and subject to event changes (more fun!), and more likely to be wrong for self-limiting of data (more useless!)

I didn't say superiour, just that once predicted, I don't change it.  I predicted NJ for Bush last time, which was wrong.  I could tell it was wrong two weeks after I made it, but the prediction was still made.

For me, it's more fun to make the prediction in September.

[Grad Students are the Worst]

ahh.

So, sometimes, your predictions are different from what you currently believe will happen?

That's confusing, but also explains a lot.

[So the Heroes Fall]

Still predicting McCain to win Colorado and ignoring everything in the past month is not principled, it is idiotic. Then again, consider the source...

BRTD, I don't change my predictions.  I regard it as cheating.  Anybody can look at at polls one week and change them later.  I don't ignore current treamds, but I'm not like Zogby, or apparently you, either.

How is it cheating? Obviously, the situation changes in the month before the election. Taking anything that happens then into account is perfectly logical. I don't see how changing one's prediction based on polls is anything like Zogby, who rather changes his POLLS based on what he thinks will happen, which is dishonest. But changing one's prediction based on what legitimate pollsters say is simply common sense.

[J. J.]

Still predicting McCain to win Colorado and ignoring everything in the past month is not principled, it is idiotic. Then again, consider the source...

BRTD, I don't change my predictions.  I regard it as cheating.  Anybody can look at at polls one week and change them later.  I don't ignore current treamds, but I'm not like Zogby, or apparently you, either.

How is it cheating? Obviously, the situation changes in the month before the election. Taking anything that happens then into account is perfectly logical. I don't see how changing one's prediction based on polls is anything like Zogby, who rather changes his POLLS based on what he thinks will happen, which is dishonest. But changing one's prediction based on what legitimate pollsters say is simply common sense.

I could probably predict the weather with 95% accuracy for 6:00 PM today.  That isn't fun.  I like having to predict things six weeks away where I really have to jump to some unwarranted conclusions.  I'd rather be 80% right 6 weeks out than 95% right a week before.

There is a difference between truly predicting and reporting.

[tokar]

Still predicting McCain to win Colorado and ignoring everything in the past month is not principled, it is idiotic. Then again, consider the source...

BRTD, I don't change my predictions.  I regard it as cheating.  Anybody can look at at polls one week and change them later.  I don't ignore current treamds, but I'm not like Zogby, or apparently you, either.

I don't understand why one prediction, made from looking at polls (in part, I assume) is superior just because it was made at an arbitrary time.

That just means it's more risky and subject to event changes (more fun!), and more likely to be wrong for self-limiting of data (more useless!)

I didn't say superiour, just that once predicted, I don't change it.  I predicted NJ for Bush last time, which was wrong.  I could tell it was wrong two weeks after I made it, but the prediction was still made.

For me, it's more fun to make the prediction in September.

Well your decision is rather dumb.

Just look at the Alaska Senate race for example.  6 months ago or whenever one decides to make a prediction, you would say its a strong Stevens hold, or even one month ago you might say a slight Stevens hold.  But now, Stevens was found guilty and it completely changes the dynamic of the race.  Maintaining a prediction of strong or lean Stevens wouldn't be wrong, it would be downright idiotic.

Of course this isn't a discussion about Senatorial races...but you get the point.
Maintaining that McCain would win New Hampshire after predicting such 1 month ago, in your case for example, is rather dumb.  McCain has since pulled back in ad spending and the polls have heavily tipped in Obama's direction.


Furthermore, if you consider changing predictions cheating, then you sir are a cheater:

On September 11 you said:
"McCain wins everything west of a state bordering the Pacific, except NM an CO."

On September 23 you said:
"I've since given CO to McCain."

[J. J.]

Well your decision is rather dumb.

No it was originally an academic assignment, way back in 1980.  It had to be in well before the election.

Furthermore, if you consider changing predictions cheating, then you sir are a cheater:

On September 11 you said:
"McCain wins everything west of a state bordering the Pacific, except NM an CO."

On September 23 you said:
"I've since given CO to McCain."

You'll note that on 9/11, I hadn't made any assignment of CO.  I did not say Obama will win CO.  It was still out, and that was actually one of my last states.

[tokar]

Well your decision is rather dumb.

No it was originally an academic assignment, way back in 1980.  It had to be in well before the election.

Furthermore, if you consider changing predictions cheating, then you sir are a cheater:

On September 11 you said:
"McCain wins everything west of a state bordering the Pacific, except NM an CO."

On September 23 you said:
"I've since given CO to McCain."

You'll note that on 9/11, I hadn't made any assignment of CO.  I did not say Obama will win CO.  It was still out, and that was actually one of my last states.

So if McCain doesn't win Colorado, and Obama doesnt win Colorado, are you predicting Bob Barr or something?


On another note, on September 11 you said Obama won't win PA, and now you are saying he will.