# PollyVote? **URL:** \ **Category:** Politics & Elections\ **Created:** [September 18, 2016, 10:23pm UTC](https://boards.straightdope.com/t/pollyvote/766311 "2016-09-18T22:23:24Z")\ **Posts on this page:** 9\ **Page:** 1
**Author:** ![DSeid](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/dseid/32/20194_2.png) [@DSeid](https://boards.straightdope.com/u/DSeid)\ **Post date:** [September 18, 2016, 10:23pm UTC](https://boards.straightdope.com/t/pollyvote/766311/1 "2016-09-18T22:23:24Z")
A mention in a comment on Wang’s site led me to [PollyVote](http://pollyvote.com/en/) and I thought some here might find it of interest - not for it’s reassuring take on the probable outcome (although that be nice) but out of interest in its methods and performance to date. The method: > [@](#): > > The PollyVote is based on the principle of combining forecasts. That is, PollyVote combines forecasts from different forecasting methods, the so-called component methods, each of which relies on different data. The PollyVote forecast is calculated by following a two-step approach: > > Averaging forecasts within each component method. > Averaging the resulting forecasts across component methods. … Think of it as 538’s PollsPlus (which includes econometric fundamentals) plus adding in other aggregators, and adding in four other approaches. The claimed results? [Damn good](http://pollyvote.com/en/about/faq-2/) (see 3.1). The current popular vote forecast? Clinton +6. Problem I see with it? The forecast is for a two-party race. Also there is bleed between methods, for example prediction markets are informed by read of the other methods. Anyone heard of this one before? Thoughts about it? ---
**Author:** ![Sherrerd](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/sherrerd/32/8_2.png) [@Sherrerd](https://boards.straightdope.com/u/Sherrerd)\ **Post date:** [September 18, 2016, 10:31pm UTC](https://boards.straightdope.com/t/pollyvote/766311/2 "2016-09-18T22:31:33Z")
> [@DSeid](#): > > The forecast is for a two-party race. Might not be all that unrealistic: today Carl Bernstein broke the news that Libertarian VP nominee Bill Weld is worried enough about the potential to help Trump become President that he’s considering stepping out of the race. Of course Johnson would keep going even if he received a heavenly messenger telling him his involvement will definitely give the White House to Trump, but Weld’s dropping out would weaken Johnson’s spoiler ambitions. (Hadn’t heard of the poll; thanks.) [http://www.newsmax.com/Newsfront/carl-bernstein-bill-weld-may-quit-despises-trump/2016/09/18/id/748868/](http://www.newsmax.com/Newsfront/carl-bernstein-bill-weld-may-quit-despises-trump/2016/09/18/id/748868/) ---
**Author:** ![DSeid](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/dseid/32/20194_2.png) [@DSeid](https://boards.straightdope.com/u/DSeid)\ **Post date:** [September 18, 2016, 10:35pm UTC](https://boards.straightdope.com/t/pollyvote/766311/3 "2016-09-18T22:35:50Z")
> [@Sherrerd](#): > > Might not be all that unrealistic: today Carl Bernstein broke the news that Libertarian VP nominee Bill Weld is worried enough about the potential to help Trump become President that he’s considering stepping out of the race. Of course Johnson would keep going even if he received a heavenly messenger telling him his involvement will definitely give the White House to Trump, but Weld’s dropping out would weaken Johnson’s spoiler ambitions. > > (Hadn’t heard of the poll; thanks.) > > [http://www.newsmax.com/Newsfront/carl-bernstein-bill-weld-may-quit-despises-trump/2016/09/18/id/748868/](http://www.newsmax.com/Newsfront/carl-bernstein-bill-weld-may-quit-despises-trump/2016/09/18/id/748868/) [Denied by Weld.](http://boards.straightdope.com/sdmb/showpost.php?p=19637069&postcount=26) ---
**Author:** ![Chronos](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/chronos/32/134_2.png) [@Chronos](https://boards.straightdope.com/u/Chronos)\ **Post date:** [September 18, 2016, 10:54pm UTC](https://boards.straightdope.com/t/pollyvote/766311/4 "2016-09-18T22:54:44Z")
That would only make sense if one assumes that the Libertarians will pull more votes from the Democrats than from the Republicans, which at the very least isn’t something one ought to take for granted. And meta-studies like this have a tendency to just amplify noise. You’d be much better off figuring out which predictor is most reliable, and then just following that predictor. ---
**Author:** ![Sherrerd](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/sherrerd/32/8_2.png) [@Sherrerd](https://boards.straightdope.com/u/Sherrerd)\ **Post date:** [September 18, 2016, 10:57pm UTC](https://boards.straightdope.com/t/pollyvote/766311/5 "2016-09-18T22:57:52Z")
> [@DSeid](#): > > [Denied by Weld.](http://boards.straightdope.com/sdmb/showpost.php?p=19637069&postcount=26) Interesting. (And there’s Carl Bernstein with egg on his face.) ---
**Author:** ![iiandyiiii](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/iiandyiiii/32/7924_2.png) [@iiandyiiii](https://boards.straightdope.com/u/iiandyiiii)\ **Post date:** [September 18, 2016, 11:15pm UTC](https://boards.straightdope.com/t/pollyvote/766311/6 "2016-09-18T23:15:02Z")
Another aggregator (polls plus other stuff). If their track record is [as good as they say it is](http://pollyvote.com/en/track-record/), then I’ll add them to my list (which currently has 538 on top, followed by PEC, along with the prediction markets. If after the election they turn out to be as good or better than the others, then they’ll move to the top of the list. Pleased to see the good prediction, but I won’t trust it until I can see and trust their track record (which means I can’t trust it until after this election, assuming they do well). ---
**Author:** ![DSeid](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/dseid/32/20194_2.png) [@DSeid](https://boards.straightdope.com/u/DSeid)\ **Post date:** [September 18, 2016, 11:19pm UTC](https://boards.straightdope.com/t/pollyvote/766311/7 "2016-09-18T23:19:31Z")
> [@Chronos](#): > > … meta-studies like this have a tendency to just amplify noise. You’d be much better off figuring out which predictor is most reliable, and then just following that predictor. Do you have any basis for that opinion? Not snarky, honestly curious… The data they supply seems to go against that proposition, and suggests that combining across methods decreases the potential errors in each other sort of method. Often that is a means of dealing with noisy signals, sample by different means (or at different angles, so to speak) and average it out: the idea being that the noise of different methods tends to cancel out leaving more true signal. Here’s [one](https://faculty.fuqua.duke.edu/~clemen/bio/Published%20Papers/13.CombiningReview-Clemen-IJOF-89.pdf) of their cites. It’s an old review but without data that contradicts it it seems likely still valid. No magic bullet but maybe a wee bit value added? **iiandyiiii** except that polling aggregation (in aggregate) is just 1/6th of the weighting … officially. Which I throw in because of course polls inform at least one of the other methods used. ---
**Author:** ![DSeid](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/dseid/32/20194_2.png) [@DSeid](https://boards.straightdope.com/u/DSeid)\ **Post date:** [September 18, 2016, 11:35pm UTC](https://boards.straightdope.com/t/pollyvote/766311/8 "2016-09-18T23:35:16Z")
[Track record so far](http://pollyvote.com/en/track-record/): > [@](#): > > The PollyVote published highly accurate popular vote forecasts for the three U.S. presidential elections from 2004 to 2012. … From January 1, 2012 until Election Day, the PollyVote forecast missed the final election results by only 0.3 percentage points. > > The forecast was also substantially more accurate than the closely watched forecasts at [FiveThirtyEight.com](http://FiveThirtyEight.com). That is, relying on the PollyVote rather than on FiveThirtyEight, across the period when forecasts from both methods were available, would have reduced forecast error by 42% when predicting the popular vote. … > > … The 2008 PollyVote was launched in August 2007 and forecast a victory for Barack Obama over the 14 months that it was making daily forecasts. On Election Eve, it predicted that Obama would receive 53.0% of the popular two-party vote, an error of 0.7 percentage points. … > > … The 2004 PollyVote was launched in March 2004 and forecast a victory for President Bush over the 8 months that it was making forecasts. The final forecast published on the morning of the election predicted that President would receive 51.5% of the popular two-party vote, an error of 0.3 percentage points. … > > … We also conducted ex post analyses by adding the three elections from 1992 to 2000 to the sample in order to provide additional evidence. … Across the last 100 days before each of the six elections from 1992 to 2012, the PollyVote provided more accurate forecasts than each of its component methods. … The chart below shows how the PollyVote would have performed from 1992 to 2012 when using its current 2016 specification, adding citizen forecasts and index models. The 2016 specification also adds forecasts from citizen forecasts when combining forecasts within and across six (instead of four) components (adding citizen forecasts and index models). > > The chart shows the mean absolute error across the six elections from 1992 to 2012, across the remaining days to election. That is, at each day, the chart shows you the error that you would have received by relying on that component method’s (already combined) forecast until Election Day. … > > \ ---
**Author:** ![DSeid](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/dseid/32/20194_2.png) [@DSeid](https://boards.straightdope.com/u/DSeid)\ **Post date:** [September 19, 2016, 1:31am UTC](https://boards.straightdope.com/t/pollyvote/766311/9 "2016-09-19T01:31:37Z")
Ah. Of the two-party vote. So if other parties get 5% then the prediction calls for Clinton to get 53% of the remainder, i.e. 50.35%, round to 50.4%, and Trump to get 47% of the remainder, i.e. 44.65%, round to 44.6%. Reading reviews here seems to be little debate over their past accuracy. Still as the[Wiki](https://en.wikipedia.org/wiki/PollyVote) article about them suggests as a possible reason for their lack of popularity with the press: “PollyVote predictions are very stable and rarely change, whereas election observers and journalists are interested in excitement and newsworthiness.”