|
|
|
|
Baseball Primer Newsblog— The Best News Links from the Baseball Newsstand
Wednesday, October 31, 2012
Come next Tuesday night, we’ll get a resolution (let’s hope) to a great ongoing battle of 2012: not just the Presidential election between Barack Obama and Mitt Romney, but the one between the pundits trying to analyze that race with their guts and a new breed of statistics gurus trying to forecast it with data.
In Election 2012 as seen by the pundits–political journalists on the trail, commentators in cable-news studios–the campaign is a jump ball. There’s a slight lead for Mitt Romney in national polls and slight leads for Barack Obama in swing-state polls, and no good way of predicting next Tuesday’s outcome beyond flipping a coin. ...
Bonus link: Esquire - The Enemies of Nate Silver
|
Bookmarks
You must be logged in to view your Bookmarks.
Hot Topics
Newsblog: Quiz: Do you know MLB rules? - SportsNation - ESPN (26 - 7:47am, Jun 19)Last: BDCNewsblog: [OTP-June] Economic Times: Hope politics, sports don’t get mixed up: Manmohan Singh (2118 - 7:38am, Jun 19)Last: BDCNewsblog: Stan "The Fan" Charles: After Biogenesis, Should MLB Players Still Have The Right To Arbitration? (28 - 7:11am, Jun 19)Last: Misirlou is bad, he's nationwideNewsblog: Murphy: Ruben Amaro Jr. doesn't "do" five-year plans, but the Phillies need a good one (25 - 7:08am, Jun 19)Last: Edmundo got dem ol' Kozma blues again mamaNewsblog: OT: The Soccer Thread June, 2013 (597 - 6:52am, Jun 19)Last: Shooty is in the Trust TreeNewsblog: Perry: Josh Hamilton and the terrible, horrible, no good, very bad night (1 - 6:50am, Jun 19)Last: Shooty is in the Trust TreeNewsblog: Primer Dugout (and link of the day) 6-19-2013 (1 - 6:23am, Jun 19)Last: Matt Chico's Bail Bonds (Dan Lee)Newsblog: Draft signings (117 - 5:48am, Jun 19)Last: Jack Carter, calling Beleaguered CastleNewsblog: Steinberg: St. Louis baseball writer makes hilarious jokes about the Nats (83 - 5:43am, Jun 19)Last: Jack Carter, calling Beleaguered CastleNewsblog: LATimes: Microsoft unveils new Xbox One console (211 - 4:22am, Jun 19)Last: vivaelpujolsNewsblog: Schoenfield: Scherzer has turned into another Detroit ace (20 - 4:13am, Jun 19)Last: vivaelpujolsNewsblog: ‘Old man’ Arroyo pitching better than ever (3 - 4:07am, Jun 19)Last: vivaelpujolsNewsblog: OMNICHATTER for June 18, 2013 (193 - 2:48am, Jun 19)Last: Sleepy supports unauthorized ramblingNewsblog: OT: NBA Finals and June thread (912 - 2:39am, Jun 19)Last: robinredNewsblog: WaPo - Sheinin | For Angels' Mike Trout, no ceiling applies (53 - 1:43am, Jun 19)Last: Walt Davis
|
|
Reader Comments and Retorts
Go to end of page
Statements posted here are those of our readers and do not represent the BaseballThinkFactory. Names are provided by the poster and are not verified. We ask that posters follow our submission policy. Please report any inappropriate comments.
All the available evidence suggests he's capable of both simultaneously.
You linked to a Gallup study and a New York Times study. Regardless, why would you base a model on what people say they will do rather than what you know they have done?
How do you know this? You're making an assumption based on a statistical anomaly. That's it. The idea that almost a quarter of Americans don't know they're registered to vote is absurd, especially with today's GOTV efforts.
***
It's funny how McCoy made a big show out of putting me on "ignore" but still obsessively replies to every comment of mine that's quoted by someone else. Very strange.
Ding, ding, ding!!
Which is what it is, I suppose.
No, I never linked to Gallup regarding this. I linked to the same Census report in both 643 and 1186. Please at least pretend that you're interested in the truth.
If you know from census data that 90% of people who say they're registered to vote actually vote, then it's not at all strange to have your model show that 96% of people who say they're registered to vote are likely voters.
My mistake. I thought it was Gallup. (But what a strange comment. Mistaking Gallup for the Census Bureau means one is uninterested in the truth? The author of the study was essentially irrelevant to the discussion at hand.)
But you don't know any such thing. All you know is what people said they did, not what they actually did. Do you seriously believe the Census Bureau went through 180,000,000 registered voter records to see who was lying and who was telling the truth?
If you want to keep pretending that almost a quarter of American registered voters don't know they're registered, that's fine, but it's an absurd assumption. Do you seriously believe that a quarter of the voters in Ohio or New Hampshire or Colorado haven't gotten a single call or mailer from a political party or candidate this year offering them an absentee ballot or reminding them to vote?
At first I thought "Cool, I didn't know WaPo had their own Nate Silver writing columns about the horse race." Then I realized the guy writes columns about actual horse races.
And the meteorologist should be disqualified for not knowing that Bernie Sanders is an independent.
Joek apparently believes that caller ID doesn't exist.
Is there a single sentient being with caller ID and voice mail in the year 2012 who actually answers the phone if he or she doesn't recognize the name or number?
---------------------------------------
Morris and Barone, meet your competition from the other side of the aisle: Jim Cramer, picking Obama with 440 EVs.
I see that on that site, the Post's horse racing writer Andy Beyer is one of two pundits picking Romney. The average prediction for the 12 pundits is Obama by 281-257. And Juan Williams is the only one with a split PV (Romney) vs EC (Obama) prediction.
It shows that you never actually clicked on the link, which explains why none of your arguments attempting to rebut it make any sense.
When did I claim non-participants were included in the topline poll results? I was referring to response rates and this idea that only likely voters answer the phone.
Beyer picks Romney, but the professional poker player picks Obama.
Think about this for a second, Andy. You seem to be predicting a 0 percent participation rate for pollsters.
Are you sticking to your claim that over 40,000,000 registered voters — nearly a quarter of the U.S. electorate — are unaware that they're registered to vote?
At the risk of enraging Shipman for questioning one of the government bureaucracies he likes so much, it seems illogical to base a model on the Census Bureau's survey data when we have the actual data.
That's down from a 35% response rate only a decade or two ago. Reason enough to be more cautious about the results than many here are. If the polls are off this year, looking at the partisan split of non-responders might be a good idea.
Better than Chris Cillizza predicting we'll have three independents. Maybe he thinks Chris Murphy is actually Joe Lieberman wearing one of those Mission Impossible masks?
Since you keep repeating this nonsense without responding to what I actually write, I'll just repeat what I wrote one more time:
A polling firm found (in one poll) that 96% of people who say they are registered to vote are likely to vote. While it's true that only ~72% of people who are actually registered to vote actually vote, the correct comparison is to the percent of people who say they are registered to vote who actually vote.
If Beyer says Romney is a "Mortal Lock", Obama partisans should be worried.
When did I dispute this? Never.
The dispute is over the proper regression of RV to LV, or claimed LV to actual voters. And history shows that 96 percent LV is more than a little on the high side.
Yes, and roughly 95 percent of former high school athletes exaggerate their stats when asked about them decades later. In other words, in a society where voting is seen as somewhat of a civic responsibility, a lot of people who didn't vote are going to be tempted to claim they did (or claim they weren't registered when they knew otherwise).
The idea that a model should be based on what people say they will do (or have done) rather than on what we know they have done is very strange, especially on a site like this one. Since when is survey data better than actual real-world data?
If a Mortal Lock is 284-254, then I guess that Romney's a veritable Secretariat.
--------------------------------------
That's down from a 35% response rate only a decade or two ago. Reason enough to be more cautious about the results than many here are. If the polls are off this year, looking at the partisan split of non-responders might be a good idea.
And yet nearly all of the big aggregate polling misses we've seen go back further than that.
But what does Smooth Jimmy Apollo say?
Think about this for a second, Andy. You seem to be predicting a 0 percent participation rate for pollsters.
Well, obviously I was making a rhetorical point, but in any event I know that there are at least two certain Obama voters out there whom pollsters would never be able to reach. You've got to be Charlie Brown the field goal kicker to think that anyone without a name on the caller ID is anything other than a junk call.
As much as I bag on Morris, I laugh just as hard at Cramer.
Maybe if you didn't have your wife slaving away while you're down at the pool hall ...
Speaking of which, it's easier to find a Romney voter in Berkeley than to find a regulation pool table. Maybe that "Bowling Alone" guy is right.
And yet nearly all of the big aggregate polling misses we've seen go back further than that.
When pollers had higher response rates they had higher error rates. As polls have gotten more and more sophisticated they have gotten better at measuring what is actually happening. Basically this whole argument about reponse rates and such is just a smokescreen to wishcast away "my guy" isn't winning facts.
Not basically. Wholly and entirely.
Swing and a miss.
If polling methodology was foolproof, Nate wouldn't have predicted Sen. Ken Buck, Sen. Joe Miller, or Sen. Sharron Angle back in 2010, or underestimated the GOP's House gains by over 20 percent.
The amount of faith people are putting in Obama's 2-point lead in Ohio, 1-point lead in Colorado, and 1.5-point lead in New Hampshire is kind of astonishing.
Politico Exclusive--Christie was Mitt's first choice for VP
Knives in the back sort of stuff, possibly? Just random leaking? Not really sure.
I'm not against the death penalty per se, but currently it's applied so haphazardly I can't get behind it.
Two NE guys on the ticket? Yeah right.
I'm putting them on this Google Doc for me to update when I get more.
2012 Election Predictions
Post a link to any published predictions mentioned before that I haven't listed yet, and I'll put them in there.
(Dick Morris doesn't have one for President, but does for Senate.)
It makes me really sad, but Prop 34 is very dicey at the moment. It's currently up by 7 points, but polling volume has been really weak. It's still just at 45%. The history of these kind of ballot measures is to sputter towards the end.
I would trade a Romney win for abolishing the death penalty.
No, I can believe he's not that bright. I remember him getting savaged on The Daily Show for his ridiculous financial/stock advice during the 2008 meltdown.
But I'm not sure about tickets and their effects on regional appeal in the first place. Ryan may not be able to get WI in the red column, for one. Romney is a man of three states yet of none at the same time - he seems just more national than regional.
I'm open to evidence.
I know this scenario happens. I don't have any estimate for how common it is, but I have no reason to think it's rare.
If true, I'm guessing this would be a much bigger story if Bush was still in office.
PA, WI, IL and now VA and DC.
But I'm not sure about tickets and their effects on regional appeal in the first place. Ryan may not be able to get WI in the red column, for one. Romney is a man of three states yet of none at the same time - he seems just more national than regional.
I'm open to evidence.
If Ryan can't turn Wisconsin red then there is no way Christie is going to turn NJ red. Having both Romney and Christie out of the NE does two things. It alienates power bases from other regions and increases red votes in a region that isn't going to go red which means it is largely useless. Romney won't turn Mass red, Christie isn't going to turn NJ red, and neither of them are going to turn NY or PA red so what is the point of them? Romney with his mormonism helps turn Utah and such but that region was going red anyway so again that is useless.
Every time I've registered to vote, there's been a question on the form asking for the last address at which I had been registered. I believe a notification is supposed to be sent to the election board in the prior location, but, government bureaucracy being what it is, I'm sure it doesn't always happen.
And he vowed to donate entire proceeds to the Todd Akin campaign, as it is well known and accepted science that women can reject implantation of ugly baby embryo.
Some morons are just dead-set against correcting their ignorance on voting matters.
Dan Quayle didn't hurt Bush that much (Dukakis self-inflicted his own wounds), but he didn't look that good of a pick in 1988.
The anti-death-penalty folks finally wised up and stopped trying to sell this as a "21st century barbarism" issue, and started selling it as a "state saves $" issue.
On balance I think it will save CA money. But it will also result in more dicey cases going to trial, instead of settling at LWOP / "slow death," and that's great.
Taking that death-penalty hammer away from the govt is an enormous positive, which will reveal itself over time at both the trial and appellate levels.
Another point against capital punishment - gotta save coin where you can.
Had there been the right Ohioan veep, though? Turmoil!
A prez candidate with a real shot at winning is a pretty rare bird. I don't think there's any way to create a breeding ground for them in a particular state (no snark, I just don't know how else to say it). A Portman-Rubio ticket in 2012 (probably the strongest candidate from each state) would have lost both Ohio and Florida by around 5% rather than 7% for a comparable Portman-Rubio from other states. For a variety of reasons they were never going to be strong, national candidates. Their giving the GOP ticket a bump in OH and FL wouldn't have compensated for their overall weaknesses.
Even if they would have swung OH and FL their weaknesses would have cost the ticket VA, CO, NH, NV, PA, NM, MI, WI, NC, IA...
FTFA:
He's two men's man!
It's possible but Tennessee went blue for Clinton twice despite Tennessee getting red and redder. Since the 1950's the only Dem Presidents to take Tennessee were LBJ, Carter once, and Clinton twice.
Of course, it voted Republican when Gore was at the top of the ticket, so that kinda weakens that argument.
I've never understood how people who "don't trust big government" would give said government the power of life & death.
Again, Silver estimated 54 House pickups. That is 14.3% short of 63.
Silver hedged on Joe Miller weeks before Election Day, writing a column specifically warning about Miller's falling prospects when he was still a sizeable favorite.
1996: Clinton 48, Dole 46, Perot 6
Wikipedia has it to 4 decimal places but I rounded it.
1992:
Clinton 933,521 (47.1%)
Bush 841,300 (42.4%)
Perot 199,968 (10.1%)
1996:
Clinton 909,146 (48.0%)
Dole 863,530 (45.6%)
Perot 105,918 (5.6%)
Assuming that every Perot vote was a vote siphoned from the two main parties-- no doubt with a giant sucking sound-- both GOP candidates would have needed roughly 75% of Perot's total to get past Clinton.
47-42-10 in 1992
49-46-6 in 1996
Don't know the breakdown, but Perot would have taken a pretty even share from each in a midSouth state.
Rana Foorahrar of Time is suggesting the House might go Dem. Scary.
coke to GB. And DA.
But 'over 20 percent' sounds so much worse than '14.3%'.
Silver also missed the 2010 election BY MORE THAN EIGHT SEATS! Which might not be as bad as HE WAS OFF BY ALMOST 15 PER CENT!!
Huge crowd for Romney in Colorado tonight.
It does and it doesn't. By the time Gore ran the state was even redder, he'd been away from the state for 8 years, running against another southerner, and had Lieberman as his running mate.
I'm not saying that Gore moved the needle by 10 points for Clinton but I think the presence of Gore on the ticket certainly helped gather a few points in Tennessee.
The day before Election Day 2010, Nate was still giving the Dems a 16 percent chance of holding the House. Today, Nate gives Romney a 16 percent chance of winning the White House. Does anyone seriously believe 2010 was more of a toss-up election than 2012, or that Dems were better positioned in 2010 than Romney is positioned in 2012? Anyone?
The Dems had to win dozens of elections in 2010 to hold the House. Romney has to gain 2 points in Ohio — if he's even behind in the first place.
I don't think it was 60-40 either. Someone had the link to the three predictions on the last page and I'd say it was something like 70-30 to 75-25. The Nevada one was the biggest miss where I think he had Angle as 86% likely to win the state.
The ticket they have now is going to cost them most of those states or they are safely red states.
How do you cut and paste an entire page from this thread? This was already asked and answered.
They are equally non-toss-ups. In both cases there would need to be systemic polling error for the predicted result not to occur.
You can't make this stuff up.
Perhaps the endorsement came wrapped in a copy of The Des Moines Register endorsing Romney in part because under President Romney Republicans are more likely to work with Democrats. It does have the same odor to it...
It's a little bit interesting, that this go-round no one's bothering to play the bipartisanship card. It's not even a fantasy any more.
Is this the political version of scouts vs stats?
As huge as this?
Or the crowd from this event? Politico: "The campaign announced that 30,000 people attended, but Secret Service sources said just 15,000 went through metal detectors."
Taunting aside, crowd size braggadocio (on either side) makes the 7-Eleven color-coded coffee cup system look scientific.
Yes? I mean, in 2010, it took really high R turnout to get those results. All the 50/50 races broke towards R. That's what happens in a wave.
There was roughly a 14% chance that all the polling was wrong in 2010 as well. Turns out it wasn't.
I'm willing to believe that 30k showed up but when they saw that Kid Rock was performing half of them left, making both counts accurate.
No, not in both cases; just in 2010. If Romney wins Ohio and Colorado, both of which are well within the margin of error, that will hardly constitute systemic polling error.
Latebreaking crowdgasms is a time-honored ploy. McCain was off the charts in the last days of 2008. Dole was wowing 'em in '96. It's not always complete fraud, though. Some of it is the funeral aspect that draws folks.
And 14,500 of that was a junior staffer named Bugs dashing back and forth through a detector.
edit: Oh. It was Buzzfeed. Click and learn.
There's a non-zero chance of this too.
Smaller than one point?
Possibly. But RCP shows Obama up 2.9 points in Ohio anyway. The margin of error from 11 combined polls which surveyed a total of 9,709 likely voters is less than 2.9 points.
Clickable, too! If Obama wins Iowa, those 431-76-5 branches reduce to 225-30-1.
Of the 9 states it lists, I make WI the likeliest Obama win, which reduces the branching to 230-24-2. Add NV (second likeliest) to Obama's column, and you get 119-9-0.
Just Obama nailing OH yields 244-11-1. Give Romney NC and you get 116-11-1. If Obama takes OH and VA, and Romney gets NC, Romney has all of 1 out.
The best I can see for Romney as of 11 pm or so is something like, he takes NC, FL, VA, and CO, while Obama takes NV and WI, leaving him and Obama with 4 paths each.
Cool toy.
[rimshot]
Right, but the margins in Virginia, Florida, Colorado, and New Hampshire are 0.3, 1.4, 0.6, and 1.5, respectively. (And if he's regressing those Ohio polls to historical party splits and voting trends, Obama has less than a 2.9-point lead there.)
Yes, but if you give Romney North Carolina & Florida, Obama has 66 and Romney 58, with 4 ties - that still seems like a lot of ties! If Virginia goes Romney, too, as I suspect it will, Romney has almost twice as many paths as Obama.
But 2.9 is what RCP says, too, and I thought that was a straight average.
Plus, failing to adjust historical results retroactively to give the incumbent the advantage of having one of the best state economies in the country (as is the case in 2012) is simply loopy wishful thinking. When your best shot is a tie, and that's according to Rasmussen, time's a running out.
You can't trust that liberal organ.
Yeah, only one new national poll came out today (Rasmussen), and that one was unchanged since yesterday (a tie). For whatever reason the WaPo tracking poll never showed up on either RCP or its own website.
-------------------------------------------
Yes, but if you give Romney North Carolina & Florida, Obama has 66 and Romney 58, with 4 ties - that still seems like a lot of ties! If Virginia goes Romney, too, as I suspect it will, Romney has almost twice as many paths as Obama.
If I had some ham, I'd make a ham sandwich, as soon as I could get me some bread.
You can't trust that liberal organ.
Now, now----I've heard that they're compassionate Christian conservatives. You know, like snapper.
Andy, it wasn't the lack of polls coming out. It's the endless analysis of the polls that are out. The vote will be soon enough everyone. Chill.
You must be Registered and Logged In to post comments.
<< Back to main