Elections that were prematurely called for the wrong person?

[Tekken_Guy]

What elections were prematurely called for the wrong winner?

[Mr. Smith]

Dewey, obviously.

Also GA SEN 1980.

[nclib]

FL in 2000.

I think NH Senate was once called for the Democrat but the Republican won.

[RJ]

Didn't Christine Gregoire defeat Dino Rossi for the governor in Washington in 2004 by a recount? I thought the media originally claimed that Rossi won.

[Rhode Islander First, American Second]

TJ Cox vs. David Valadao in 2018

[Dwarven Dragon]

IL-13 in 2018 at least on some stations.

[RBH]

Dick Swett vs Bob Smith in 1996 was called for Swett off of an exit poll. Smith won by 3%

[Mr.Phips]

NC-SEN 2014 was called for Kay Hagan relatively early by a network or two.  Also FL-SEN in 1988.

[Roll Roons]

Haven't you made this thread before?

[GeorgiaModerate]

FL in 2000.

I think NH Senate was once called for the Democrat but the Republican won.

If you mean the 1974 NH Senate election, it was much more complicated than that.  See https://indepthnh.org/2024/11/05/look-back-wyman-durkin-1974-the-closest-u-s-senate-race-in-history/ for a very good writeup (a must-read for all fans of Congressional elections).

[PALiberal]

I'm pretty sure the networks called the GA Sen race for Herman Talmadge early in 1980, just for him to narrowly lose to Mack Mattingly. Which is impressive given Carter won his home state by 15 that year.

[RBH]

Florida 2000 really ended the days of calling close states based off of exit polling, didn't it?

[minisoldr]

DDHQ calling the MO-01 Dem primary for Lacy Clay in 2020.

[Talleyrand]

DDHQ calling the MO-01 Dem primary for Lacy Clay in 2020.

And the Michigan Democratic primary for Clinton in 2016.

[Steve from Lambeth]

LTElections called CA-45 for Michelle Steel for a week or two (!)

[nclib]

FL in 2000.

I think NH Senate was once called for the Democrat but the Republican won.

If you mean the 1974 NH Senate election, it was much more complicated than that.  See https://indepthnh.org/2024/11/05/look-back-wyman-durkin-1974-the-closest-u-s-senate-race-in-history/ for a very good writeup (a must-read for all fans of Congressional elections).

I was actually referring to 1996.

[Benjamin Frank 2.0]

Aside from that, I remember one case, unfortunately I can't remember what state, where I noticed (because I had written them down originally) where the final certified results were changed from the original certified results. I saw this because I had noticed some other discrepancy.

It didn't change the results all that much, but I was so surprised, at the same time I was doing this I was speaking to a friend online, and said to them 'I've never seen this before, the certified results in an election have been changed. What if that had effected the outcome?'

And my friend replied "We're sorry to inform you, we had told you you had won the election, you actually lost."

[Pollster]

Decision Desk HQ is bound to do this one of these days if they haven't already -- they call races WAY too early and have mostly lucked out (much like the networks did when they called the House incredibly prematurely for the Democrats in 2020).

[Mr.Phips]

Decision Desk HQ is bound to do this one of these days if they haven't already -- they call races WAY too early and have mostly lucked out (much like the networks did when they called the House incredibly prematurely for the Democrats in 2020).

But the networks got the House right in 2020.  They also get AZ-PRESIDENT right that year. 

[MT Treasurer]

Other GA races besides GA-SEN 1980 include GA-SEN 1992 (called for Wyche Fowler on election night only to go to a runoff where he was defeated by Paul "jingle" Coverdell) & GA-SEN 2008 (called for Chambliss on election night by at least one network - FOX News, IIRC - only to go to a runoff where Chambliss won nonetheless).

GA-PRES 1992 (called at polling closing time when Clinton won the state by roughly half a point) was also very premature and later retracted by a few networks because the count was so close at that point.

[GeorgiaModerate]

Decision Desk HQ is bound to do this one of these days if they haven't already -- they call races WAY too early and have mostly lucked out (much like the networks did when they called the House incredibly prematurely for the Democrats in 2020).

But the networks got the House right in 2020.  They also get AZ-PRESIDENT right that year. 

But "right" and "premature" aren't mutually exclusive.  Fox and AP's call of Arizona in 2020 was premature -- they were lucky it held up.  The other networks didn't call it until much later.  I've read that the networks use a standard of 99.5% likelihood to win before calling a race.  DDHQ clearly uses a lower standard, and they have indeed gotten at least one call wrong and had to retract it:

https://twitter.com/DecisionDeskHQ/status/1290860670083313664