Romney on intrade

[RandFan]

WaPo-ABC tracking poll: High marks for President Obama on Hurricane Sandy response

Nearly eight in 10 likely voters say the president has done an “excellent” or “good” job dealing with what’s been labeled a “super storm.” Almost as many give positive reviews to the federal government’s response generally. Even two-thirds of those who support Republican Mitt Romney in next week’s presidential election say Obama is doing well in this area.
<...>
The hurricane situation provided the president with a “commander-in-chief moment,” and a challenge to both candidates on the campaign the trail. When it comes to his response to the storm, Romney gets more positive than negative reviews, but a sizable 35 percent of likely voters responded that they had “no opinion” on the matter.

It's tempting to say Obama learned from Katrina but there were major disasters prior to Katrina.

 

[AdMan]

In fairness, there's not a damn thing Romney can do by being there.

Not that it appears he'd be welcome anyway:

“I have no idea, nor am I the least bit concerned or interested,” [Gov. Christie] said when asked whether Romney would visit. “I’ve got a job to do here in New Jersey that’s much bigger than presidential politics, and I could care less about any of that stuff.”

http://www.politico.com/news/stories/1012/83104.html?hp=l1

 

[qayak]

Within the aftermath of Hurricane Sandy, people are seeing that President Obama is a real leader whereas Romney is just an empty suit.

The thing that boggles my mind is that people didn't know this before. Has there ever been an emptier suit than Romney?

 

[thaiboxerken]

Seems very hard for the GOP to spin this one against our awesome President. The most significant "criticism" is that Obama acted too fast. (which makes absolutely no sense)

 

[qayak]

In fairness, there's not a damn thing Romney can do by being there.

Actually, I disagree. Obama isn't "doing" anything either. Essentially he is showing people that he is involved, aware and concerned with their suffering. Romney could have done that.

These are the things that show the true character of people. Bush flew over New Orleans and couldn't be bothered to land. People lost a lot of respect for him when he did that. He wouldn't have done anything either, except to show people exactly what Obama showed them by touring the affected area of Sandy.

Romney is showing his true character by not going, although even if he did go, it would only be to privatize disaster relief and make sure people knew they weren't getting any handouts from him.

 

[boooeee]

this stundie level comment was post over at intrade the other day

I'm not sure what you think is wrong with that statement. I'm tempted to nominate your nomination for a stundie.

 

[Puppycow]

I'm not sure what you think is wrong with that statement. I'm tempted to nominate your nomination for a stundie.

Ten is probably too small a sample size to say whether the model is good or bad.

If you had 1000 samples, it should be right about 70% of the time +/- 3% or so.

If you flip 10 coins you're often not going to get 5 heads and 5 tails.

 

[boooeee]

this stundie level comment was post over at intrade the other day

Ten is probably too small a sample size to say whether the model is good or bad.

If you had 1000 samples, it should be right about 70% of the time +/- 3% or so.

If you flip 10 coins you're often not going to get 5 heads and 5 tails.

People often use "x times out of ten" to indicate a general statement about probability, not what they expect to happen in exactly 10 tries. As in "9 times out of 10, that works".

I see no reason to think that the intrade commenter wasn't speaking in the same way, but I suppose that it is open to interpretation. Curious if that was why Alfred thought the comment was stundie-worthy.

 

[JoeTheJuggler]

In fairness, there's not a damn thing Romney can do by being there. But yeah, the President's rapid response and Christie's continuing endorsement of same is hurting Romney bad in the polls. I don't think this was the October Surprise that anyone expected.

I wouldn't think the polls would have reflected this yet--especially since poll taking in the affected areas of the Northeast is still hampered.

 

[ProbeX]

66.4% but who is counting?

As we approach election day, Intrade is my OCD Heaven/Hell.

 

[Puppycow]

People often use "x times out of ten" to indicate a general statement about probability, not what they expect to happen in exactly 10 tries. As in "9 times out of 10, that works".

I see no reason to think that the intrade commenter wasn't speaking in the same way, but I suppose that it is open to interpretation. Curious if that was why Alfred thought the comment was stundie-worthy.

Well if the model got it right 10 out of 10 times, that would suggest that it's actually an even better model than originally thought, not that it's a "crappy" model. No?

ETA: I think I see the problem here. There is no "right" or "wrong" outcome as far as the model is concerned. If a model predicts a 70% chance of a certain outcome, and the outcome doesn't happen, that doesn't mean the model was "wrong" and if the outcome does happen, it doesn't mean that the model is "right". It would need to be tested many times before one could say how good the model is. No single outcome is significant evidence of anything.

 

[remirol]

Romney closed at $3.14 last night.

Between 6:40 and 6:50 Eastern time, his price rose to $3.54. Anyone wanting to short Romney, one of his supporters with a big wallet and no brains is splashing the pool again

ETA: And at 8:12a EDT, it's below $3.30 and falling. Opportunity passed.

 

[boooeee]

Well if the model got it right 10 out of 10 times, that would suggest that it's actually an even better model than originally thought, not that it's a "crappy" model. No?

ETA: I think I see the problem here. There is no "right" or "wrong" outcome as far as the model is concerned. If a model predicts a 70% chance of a certain outcome, and the outcome doesn't happen, that doesn't mean the model was "wrong" and if the outcome does happen, it doesn't mean that the model is "right". It would need to be tested many times before one could say how good the model is. No single outcome is significant evidence of anything.

Agreed on your second point. But if the model gives a 70% likelihood of something happening, and it happens 100% of the time over a large sample size, then I would say that the model is bad.

Look at it from the opposite way: If your model says a certain outcome has a 30% likelihood and it never happens over a large sample size, then something is wrong with your model.

When it comes to Nate Silver's presidential election model, you could argue that neither outcome would invalidate his model, since a 21% likelihood is not insignificant. But you could get more sophisticated than that. He is actually making quite a few probabilistic predictions, as he is predicting the outcome of each state as well as the Senate/House elections. One way to test his model would be to group all is ~90% predictions together. If his model is calibrated right, he should be correct about 90% of the time, assuming the sample size is big enough. Do the same for all his ~80% predictions, and so on.

I thought the intrade commenter raised an important point, even if they may have been a bit sloppy in their language.

 

[MattusMaximus]

Okay, let's just call it an "unknown" unknown.

You win

 

[remirol]

The markets over at InTrade are absolutely out of their tree right now. Someone is coming in, buying up Romney heavily in the space of a few minutes, watching the price self-correct, and mashing the "buy" button again.

It's bouncing between $3.20 and $3.55 relatively consistently at the moment. Almost certainly one of the campaigns or PACs piddling away some spare money, but if you've got a funded account, I would say "now's the time to play day-trader".

 

[JoeTheJuggler]

Romney closed at $3.14 last night.

Between 6:40 and 6:50 Eastern time, his price rose to $3.54. Anyone wanting to short Romney, one of his supporters with a big wallet and no brains is splashing the pool again

ETA: And at 8:12a EDT, it's below $3.30 and falling. Opportunity passed.

Well he's mostly staying above the price of a gallon of gas around here anyway.

(That occurred to me because gasoline was $3.149 here the other day.)

 

[The Don]

The markets over at InTrade are absolutely out of their tree right now. Someone is coming in, buying up Romney heavily in the space of a few minutes, watching the price self-correct, and mashing the "buy" button again.

It's bouncing between $3.20 and $3.55 relatively consistently at the moment. Almost certainly one of the campaigns or PACs piddling away some spare money, but if you've got a funded account, I would say "now's the time to play day-trader".

Unless someone is really confident of a Romney victory.

It's not impossible that the polls have been wrong and we get a dramatically different result than that expected.

 

[remirol]

Unless someone is really confident of a Romney victory.

It's not impossible that the polls have been wrong and we get a dramatically different result than that expected.

There's a lot of things that aren't impossible in this world, but worrying about a 1 in 6 chance (that isn't even properly equivalent to a die roll; it's a 1 in 6 chance that our knowledge is wrong, not a 1 in 6 chance that people will vote a certain way or not) would rapidly lead to madness.

Act according to what's likely, allow for what isn't appropriately, and dither not lest ye find thyself a quivering blob of fearful jelly.

 

[JoeTheJuggler]

It's not impossible that the polls have been wrong and we get a dramatically different result than that expected.

Isn't that just a restatement of what an probability estimate says? If the probability of an event is 80%, we expect that event would happen in 80 trials out of 100 (or that it won't happen in 20 trials out of 100).

It's not impossible to get struck by lightning, and therefore nobody really expects that to happen to him (including the people who are in fact struck by lightning--up until they are struck, that is).

 

[NotJesus]

There's a lot of things that aren't impossible in this world, but worrying about a 1 in 6 chance (that isn't even properly equivalent to a die roll; it's a 1 in 6 chance that our knowledge is wrong, not a 1 in 6 chance that people will vote a certain way or not) would rapidly lead to madness.

According to Nate Silver's latest calculations, it's now about a 1 in 7 chance.