When did the betting markets/projections/cw turn on Election Night?

[TC 25]

Looking back at Election Night, I'm curious to find out when the tide turned and when the markets, convention wisdom turned toward Trump.

From my own recollection, between 7 and 8 pm ET, the talking heads on TV were speaking in code as if Clinton would win, based on the exits, their talk of a Latino surge, the idea that FL was going Clinton's way.  I can recall even at 7:30 or 7:45 ET, the betting market odds on RCP had Clinton still at an 85 percent chance to win. As a Trump supporter, I was down.  I hadn't given up hope but from appearances, tweets (Frank Luntz), etc., it seemed like Clinton was on her way to a comfortable win.

For me, the realization that Trump could indeed win set in around 8:30 PM ET, with Trump starting to pull ahead in FL.   Numerous networks were still going with the spin that Clinton would win FL (even Krauthammer on FOX) even as the vote totals showed Trump gaining.  The Virginia vote also gave me some optimism.

Between, 9 PM and 10 PM it became obvious to me that Trump would win FL, NC and OH and this started to sink in with the TV analysts, much their shock.  Hearing that my own state, MN, was too close to call, and hearing about the Trump strength in WI, PA, MI and even NH made me a believer.  The stock market plunge, while unsettling, provided further confirmation.

[angus]

Interesting post.  I didn't turn on the TV till after about 9:30, when I finally got the boy into bed.  By that time it was unclear who would win.  I watched for about an hour, then got bored and went to bed myself.  I remember not knowing who would win at the time I drifted off to sleep.  That was probably after 10.  It was not particularly surprising to me that they didn't know who was going to win.  I posted in this thread that it would happen some time after 10pm EST.  Admittedly, it was a safe bet.  After 10 can mean 10:01pm, or it can mean New Year's Day.  Oddly, a bunch of people, like more than a third, said between 8 and 9 pm.  Polls aren't even closed by that time. 

[100% pro-life no matter what]

I was confident sometime around 7:30-8 CT that Trump was going to win.  Once it was clear that he was going to win Florida, Ohio, and North Carolina, and that he certainly looked decent in some other Midwestern states, our watch party pretty much became a victory party hours before it was officially called.

[SirMuxALot]

I had both the NY Times projection meter on one tab and the electionbettingodds.com on another tab going all night.

The betting markets were very slow to move, which surprised me.  I don't remember the exact times, but when the NYT projection moved to 51% for Trump, the betting market odds still had Hillary in the 80-85% range.  And I think that was at least 30 minutes after it appeared like FL and OH were going to Trump.

I didn't have any money parked in the betting market sites this time, so couldn't take an arbitrage opportunity this time.  I did in 2004 and made a good chunk of pizza money on election.  When the early (erroneous) exits leaked in '04, the betting markets spiked to Kerry at 90% or more (Bush at 10%).

And the markets were similarly slow to respond to early count totals that indicated something was off with the leaked exits.  I picked up some shares of Bush at around 20-25% and sold when it hit 90%.  If I remember right, that move only took about 2-3 hours, hitting 90% once FL returns started looking solidly Bush.

[alomas]

I have had very similar experiences to you, TC25. It feels like me on Election Night

I was confident sometime around 7:30-8 CT that Trump was going to win.

Yes, that was the turning point to me. When Trump started clearly overperforming in Florida, then not only he won the state but the those leaning-Clinton states were no longer leaning her way.

[jaichind]

What I experience was that when the exit polls came out and one saw the breakdown by demographics I was sure Trump will most likely lose by 3%-4% but avoided a blowout.

Then when IN results came it I was shocked by the size of Trump leads in rural IN.  I said to myself "well, that is because of Pence."  Then I saw the same thing for KY and then I concluded that IW and OH will go Trump for sure and places like MN WI PA and MI will be close.  Of course then I said to myself "too bad, exit polls seems to indicate that Trump will most likely lose FL and perhaps even NC so even if he overperforms in the Midwest and even take PA or MI (WI in my mind was just never going to happen) it will not matter."

Then around 7:30PM enough of FL came in to have Trump actually ahead.  All the talk on TV was about how SE FL vote will come in and them crush Trump.  But I said to myself "wait, even though the sizes of leads in the FL counties are totally different, but in 2014 Rick Scott was but at this stage and ratio of the SE FL vote coming in relative to the rest of FL is around the same as at this stage in 2014."  In other word Trump has most likely a 50/50 shot or even better to win FL.  Then around 7:45 I found the "dial" in NY times website which pretty much showed a projection for FL close to my conclusions.  Then I said, "well if FL is possible Trump then NC is for sure for Trump so no need to even dig into the results there."  I even told my DW at that stage that Trump could win and at worse losses narrowly.  She said "No way."

I then spent the next 30 min looking are FL where it got better and better for Trump.  The OH started to com in in larger numbers so by 8:30 I was convinced that the Trump Midwest rural surge was real (even as Clinton was ahead).  So for me it was around 8:20 or so when I concluded that it should be somewhat better than even money that Trump will win. It took the next hour to validate that assumption of course.

[TC 25]

Between 7 and 8 pm Central, Florida was obviously trending Trump but the TV analyst seemed to ignore the evidence and were sticking with the narrative that the Latino surge and Broward, Miami Dade would pull Clinton through.

The numbers showed Trump ahead but Krauthammer was on with O'Reilly saying Clinton essentially had Florida in the bag.

MSNBC of all networks seemed to have figured out around 8:40 ET or 7:40 CT that Trump was on his way to winning Florida.

The same thing occurred with Ohio.  Most of the networks talked about a tight race there that could go either way, even when Trump was 7-9 points ahead.  It was like they couldn't believe the numbers.

[The Other Castro]

On PredictIt, I observed a similar phenomenon to Brexit night, in which an observer without any expectations of the winner would rate a higher price than the price offered. The preconceived notions of likelihood, along with possibly the severity of the outcome should the result go the other way (Brexit Yes, Trump winning) might have made it harder for people to part with their Clinton Yes/Trump No shares, or to buy up shares for the opposite side.

[Mr. Morden]

On election night, I was tracking the live state-by-state forecast models on Upshot.  While it was clear pretty early on that Trump was likely to win in states like FL and NC, Clinton didn't really need them.  So I didn't think her struggles there meant she was going to lose nationally.  The real issue was her losing the "firewall" states of MI, PA, and WI.  IIRC, Michigan was actually the first of those to go south for her.  Here's the Upshot live odds on Michigan:

http://www.nytimes.com/elections/forecast/president/michigan

When Clinton's odds there started nosediving, and Trump took the lead (which I guess was around 8pm?) that's when I think it looked like Trump was the frontrunner nationally.

[Absentee Voting Ghost of Ruin]

On PredictIt, I observed a similar phenomenon to Brexit night, in which an observer without any expectations of the winner would rate a higher price than the price offered. The preconceived notions of likelihood, along with possibly the severity of the outcome should the result go the other way (Brexit Yes, Trump winning) might have made it harder for people to part with their Clinton Yes/Trump No shares, or to buy up shares for the opposite side.

Predict it also has a max # of traders on a particular prediction, AND tended to get log-jammed in the middle of really busy primary nights. Those would both contribute to making the markets slow to respond.