How did the 2014 predictions compare?

[Princess Nyan Cat]

I never did get around to posting these results, comparing the accuracy of the various predictions for races at the Senate, House and Governor levels in the 2014 cycle. I previously posted something similar for the 2012 elections, but there are some new players this time around.

As much stock as everyone puts in Nate Silver, I consider Larry Sabato best of the pros and Scott Elliott best of the amateurs. Newcomer PredictWise also did quite well, and would have done even better if not for a few last-minute changes they made on Election Day.

Enjoy!

-----

The "score" is a simple one based on whether the prediction picked the right party to win and does not attempt to assess strength of victory.  It is often difficult to quantify terms like Tilt, Lean, Likely, Moderate, Strong, Solid, and so on when many of the sites do not clearly define what those qualifications mean.  A prediction of "tossup" is automatically classified as wrong since the prognosticator was unwilling to make a stand one way or the other.

-----

Senate races - graded on 36 races

97.2% - Daily Kos - 35 correct
97.2% - Dave Leip's Atlas aggregate - 35 correct (31/33 in 2012)
97.2% - ElectoralMap - 35 correct
97.2% - Huffington Post - 35 correct (28/33 in 2012)
97.2% - Larry Sabato's Crystal Ball - 35 correct (31/33 in 2012)
97.2% - New York Times - 35 correct (24/33 in 2012)
97.2% - People's Pundit Daily - 35 correct
97.2% - PredictWise - 35 correct
97.2% - Washington Post - 35 correct (28/33 in 2012)
97.2% - Your precious Nyan Cat! =^..^= - 35 correct (31/33 in 2012)

94.4% - 538 (Nate Silver) - 34 correct (31/33 in 2012)
94.4% - Election Projection (Scott Elliott) - 34 correct (31/33 in 2012)
94.4% - Electoral Vote Predictor (Andrew Tanenbaum) - 34 correct (29/33 in 2012)
94.4% - Princeton Election Consortium - 34 correct
94.4% - Real Clear Politics - 34 correct (26/33 in 2012)

88.9% - Rothenberg Political Report - 32 correct (29/33 in 2012)
80.6% - USA Today - 29 correct
77.8% - Wall Street Journal - 28 correct (22/33 in 2012)
75.0% - Cook Political Report - 27 correct (23/33 in 2012)

Every prediction was wrong for North Carolina.

[Princess Nyan Cat]

House races - graded on 435 races (number wrong listed instead of correct)

97.7% - Election Projection (Scott Elliott) - 10 wrong (12 in 2012)
97.7% - Larry Sabato's Crystal Ball - 10 wrong (13 in 2012)

97.0% - PredictWise - 13 wrong
97.0% - Your precious Nyan Cat! =^..^= - 13 wrong (11 in 2012)

96.6% - Washington Post - 15 wrong (26 in 2012)
96.1% - Rothenberg Political Report - 17 wrong (24 in 2012)
95.9% - ElectoralMap - 18 wrong
95.2% - Daily Kos - 21 wrong
94.5% - Cook Political Report - 24 wrong (31 in 2012)

92.9% - Real Clear Politics - 31 wrong (35 in 2012)
92.9% - USA Today - 31 wrong
92.9% - Wall Street Journal - 31 wrong

Every prediction was wrong for these 4 districts: Arizona-2, Iowa-1, Nevada-4, Texas-23

Only one prediction was right for these 4 districts: Arizona-1 (PredictWise), Florida-2 (Larry Sabato), Georgia-12 (PredictWise), New York-24 (EVP-Scott Elliott)

[Princess Nyan Cat]

Governor races - graded on 36 races

88.9% - Election Projection (Scott Elliott) - 32 correct (10/11 in 2012)
88.9% - PredictWise - 32 correct
88.9% - Real Clear Politics - 32 correct (9/11 in 2012)

86.1% - 538 (Nate Silver) - 31 correct
86.1% - Daily Kos - 31 correct
86.1% - Dave Leip's Atlas aggregate - 31 correct (11/11 in 2012)
86.1% - Huffington Post - 31 correct
86.1% - Larry Sabato's Crystal Ball - 31 correct (10/11 in 2012)
86.1% - Your precious Nyan Cat! =^..^= - 31 correct (10/11 in 2012)

83.3% - People's Pundit Daily - 30 correct
80.6% - ElectoralMap - 29 correct
80.6% - Rothenberg Political Report - 29 correct (9/11 in 2012)
63.9% - Cook Political Report - 23 correct (8/11 in 2012)
63.9% - Governing - 23 correct
63.9% - USA Today - 23 correct
61.1% - Wall Street Journal - 22 correct (8/11 in 2012)

Every prediction was wrong for Illinois and Maryland.

Only one prediction was right for Florida (Daily Kos) and Kansas (ElectoralMap).

[pbrower2a]

Few predicted how effective the Koch fronts would be at buying the election except the insiders, and they gladly let us believe that Democrats had a chance.

[OSR stands with Israel]

Few predicted how effective the Koch fronts would be at buying the election except the insiders, and they gladly let us believe that Democrats had a chance.

For the Senate I predicted a + 10 gain for Republicans with Virginia and New Hampshire also going republican with the rest that actually did and then losing Kentucky

For governor I expected +1 Republican gain with Arkansas going Republican and everything else staying the same

[Princess Nyan Cat]

Few predicted how effective the Koch fronts would be at buying the election except the insiders, and they gladly let us believe that Democrats had a chance.

Yes, that's what happened. If only not for evil capitalists, all would gladly accept the implausible utopianism of the Marxist-Leninist worker's paradise. *eye roll*

[bobloblaw]

Few predicted how effective the Koch fronts would be at buying the election except the insiders, and they gladly let us believe that Democrats had a chance.

This folks is an example of the mental cases that infect the Dem party

[Amenhotep Bakari-Sellers]

Landrieu; Begich, Hagen and Udall had a chance

[Oldiesfreak1854]

I don't think I've seen an election where the polls got it more wrong.

[Amenhotep Bakari-Sellers]

2004 incorrectly predicted a Kerry victory in OH and got it wrong. It was only Mason Dixon that predicted Bushie to win. Even Columbus Dispatch predicted Kerry win in OH.

[Young Conservative]

Few predicted how effective the Koch fronts would be at buying the election except the insiders, and they gladly let us believe that Democrats had a chance.

I can't tell if this comment is trolling. The Koch's actually donated LESS than the democrats top donor Tom steyer. Source: opensecrets.org

[Young Conservative]

Landrieu; Begich, Hagen and Udall had a chance

Landrieu really didn't though...