He wrote a post yesterday near and dear to my heart - the reliability of looking at polling averages to determine where a race stands.
This gets a little wonky at times, so if poll analysis bores you, move on.
In his explanation about what he calls the "uncanny accuracy of polling averages", Silver includes an important caveat:
...we’re only looking at races in which at least two different polling firms published a survey in the 30-to-60 day window. If you have just one company polling a race, you don’t really have much of an average, properly speaking. Our model addresses this by assuming much greater uncertainty in cases where the polling data is very sparse.Fortunately in Ohio, that isn't a problem. In September alone, the Ohio Governor's race was covered by eight different polling institutions.
According to Silver, his "538 snapshot" of the Ohio Governor's race with 32 days to go sits at a 6.4% Kasich lead. The snapshot is what Silver considers "the best estimate of the current standing of the candidates".
With that 6.4% lead in mind, Silver continues...
Gubernatorial candidates with a lead of between 6 and 9 points in the polling average, meanwhile, have a 9-to-2 record. If we combine their numbers with those for the Senate candidates, we find that candidates with a lead comparable to Mr. Toomey’s (between 6 and 9 points, with 30 days to go in the race) have 16 wins against 2 defeats, which corresponds to a 89 percent winning percentage.So even after the recent swarm of polling showing the race tightening, what chance does Silver give Kasich of winning the race?
[...]
But the bottom line is that, over the course of the past half-dozen election cycles, constructing a simple polling average has provided a very reliable indication of which candidates will win the general election in Senate and gubernatorial races. Polling deficits in the high single digits, with 30 days to go in the campaign, have only rarely been overcome. Candidates who trail by double digits in the polling average have almost never defied the odds. Even relatively small leads of 3 to 4 points can be surprisingly meaningful.
86.8%.
I like those odds.
Thanks, Nate.
UPDATE: More from Silver on Ohio...
Ohio: It’s Everybody Poll Ohio! week, with four organizations, including The New York Times, having released polls there in the last few days. All the new polls show the same thing: a widening margin for the Republican senatorial candidate, Rob Portman, but a narrowing one for their gubernatorial candidate, John Kasich. We now show Gov. Ted Strickland, a Democrat, with a 13 percent chance to save his seat, up from 8 percent last week.
I’m sure some people will ask why Mr. Strickland is still at 13 percent and not higher, when the two polls released Tuesday show him only 1 point behind Mr. Kasich among likely voters. There are several reasons. First, while the polls show Mr. Kasich’s lead narrowing, they still show him with the lead, and even a small lead can be surprisingly meaningful with barely more than a month left in the campaign. Second, the model is not in a rush to discard some slightly older polls, like those from SurveyUSA and Quinnipiac University, which had given Mr. Kasich a double-digit advantage; if these firms resurvey the race and find a different result, Mr. Strickland’s chances should improve in our model. Third, it is hard for any incumbent governor to be re-elected in a state where unemployment exceeds 10 percent, and other Midwestern Democrats are faring poorly in gubernatorial contests this year.
Yet again, you blatanly misrepresent what Nate Silver said about Ohio.
ReplyDeleteShameless.
Modern Esquire is right. You left out some important additional points from Nate Silver's string of articles:
ReplyDeletehttp://fivethirtyeight.blogs.nytimes.com/2010/09/29/the-uncanny-accuracy-of-polling-averages-part-i-why-you-cant-trust-your-gut/
"Politicians — the ones worth their salt, anyway — are exceptionally skilled at making believers out of people, and they’ll try to make a believer out of you. Some of the time, they’ll make a strong enough argument to persuade even the most seasoned observers. But a much smaller fraction of the time will they actually turn out to be right. That’s what the data says, and it says so pretty clearly."
Sounds like someone we know.
By the way, there was nothing misrepresented in the article posted, as anyone who has been reading that series of articles knows.
Here is the most recent content from Nate Silver on the Ohio Governor's race:
"Ohio: It’s Everybody Poll Ohio! week, with four organizations, including The New York Times, having released polls there in the last few days. All the new polls show the same thing: a widening margin for the Republican senatorial candidate, Rob Portman, but a narrowing one for their gubernatorial candidate, John Kasich. We now show Gov. Ted Strickland, a Democrat, with a 13 percent chance to save his seat, up from 8 percent last week.
I’m sure some people will ask why Mr. Strickland is still at 13 percent and not higher, when the two polls released Tuesday show him only 1 point behind Mr. Kasich among likely voters. There are several reasons. First, while the polls show Mr. Kasich’s lead narrowing, they still show him with the lead, and even a small lead can be surprisingly meaningful with barely more than a month left in the campaign. Second, the model is not in a rush to discard some slightly older polls, like those from SurveyUSA and Quinnipiac University, which had given Mr. Kasich a double-digit advantage; if these firms resurvey the race and find a different result, Mr. Strickland’s chances should improve in our model. Third, it is hard for any incumbent governor to be re-elected in a state where unemployment exceeds 10 percent, and other Midwestern Democrats are faring poorly in gubernatorial contests this year.
There’s also the odd fact that Mr. Kasich’s lead is dwindling as Mr. Portman’s is growing. On the one hand, it may be a sign that voters are beginning to scrutinize Mr. Kasich, whom strategists regard as having run a mediocre campaign. On the other hand, it’s possible that Mr. Kasich could be assisted by Mr. Portman’s coattails (although Mr. Portman’s lead appears to come more from persuading independent voters than from being able to motivate a huge turnout)."
Wow Anonymous, you apparently need adult reading comprehension classes.
ReplyDeleteIn the first comment, Silver was talking about campaign's where the polling has consistently shown the candidate seven-points down.
In the Governor's race, the polling this week from multiple polling outfits show the race as tied. In other words, Silver wasn't talking about races like the OH-Gov.
As evidenced by your second quotation by Silver, which, again, you apparently fail to comprehend.
Silver is admitting that his model is likely showing Kasich with stronger odds than he actually has and is admitting that his model requires him to continue to include a still fairly recent SurveyUSA and Quinnipiac polls that show leads well beyond the margin of error of more recent polling. But because they aren't sufficiently removed in time to exclude them, his model requires them to be used. Regardless, even with their inclusion, Silver's model does all him to increase Strickland's odds a by five percent in one-week.
Silver called the second most race to move to the Democrats' favor in a week--coming just behind California and the imploding Whitman campaign.
Silver also noted that "strategists regard [Kasich] as having run a mediocre campaign."
In other words, Silver has expressed anything BUT confidence in Kasich's victory.
But don't let you lack of reading comprehension stop you from parroting Keeling, who cites and article that doesn't even MENTION Ohio or Kasich as evidence of something Silver said.
So Silver calls it the second fastest race going to the Democrats, admits his model is giving Kasich better odds than is likely because it still includes the laughable SurveyUSA and Quinnipiac, and notes that Kasich's campaign is, by and large, viewed as being mediocre...
ReplyDeleteAnd this, in your mind, shows evidence of confidence for Kasich?!? LOL.