Resolves to the person who wins the majority of votes for US President in the Electoral College, or selected by Congress following the contingency procedure in the Twelfth Amendment.
(May resolve provisionally if both the Associated Press projects a winner and the losing major party candidate concedes; if Manifold allows provisional resolutions.)
1,000
1.00
After two NYT/Siena polls in a row showing Kamala up 4 points in Pennsylvannia (1) (2), we should be pretty confident that she is actually leading there. Other polls showing a smaller lead may be suffering from the unintentional herding of weighting by recalled vote that Cohn describes here.
Pollsters are using this weighting method to make their polls look more like the very close 2020 election, and so are failing to capture the trends in the electorate we saw begin in 2022 and continuing now in the NYT polling.
Democrats are doing better with white voters, and worse with non-white voters. Kamala is likely to lose Arizona, but win Pennsylvania. And if she wins Pennsylvania, she likely wins the whole thing.
Things could change in the next 24 days, but the NYT/Siena polls are the gold standard. If we go to the 538 "what if" tool, we can set PA to Kamala and AZ to Trump and see what the model predicts:
After 2(!!) consecutive polls by godl standard WSJ and Emerson showing Trump +1 in Pennsylvania, we should be pretty confident that he's actually leading there. Other polls showing a smaller or negative lead may be suffering from the same follies that led them to fumble 2020 and 2016 badly as described by Nate Cohn here.
Trump is doing better with every kind of voter, except neocons and suburban wine moms meaning he's likely to lose DC by 95 points as opposed to 92 points. He is however likely to win all 7 swing states. And if he does that, he likely wins the whole thing.
Things will not change much over the next 24 days because Mr. Trump has the protective power of St. Michael behind him, and an army of patriots looking out for any Democratic mules. My latest sources confided in me that he is actually measuring new white house drapes and playing golf as we speak. Pack it up, Democrats - the election is over.
A move like this in the last month doesn't look good.
https://www.axios.com/2024/10/11/trump-campaign-communications-new
@jim walz is still a good pick. He may not be good at debates but he's good at everything else vance is still a bad pick. If trump had half his cognitive abilities he would be doing better but he's a deranged old man.
This whole Harris would be doing better with Shapiro thing doesn't make sense.
@jim Of the handful of options Harris explored for VP, Walz was the best choice. Shapiro and Kelly are not relatable to any segment of the Republican electorate. Picking Walz was an attempt to relate to the working class Republican voter. I would have targeted another segment of the Republican electorate and picked a seasoned Republican to split the hardline Republican loyalists.
Of Trump's 20+ options for VP, Vance wasn't even in the top 5. He looks good on paper and he did very well in the debate but has mostly sucked at everything else. He's not great with messaging, he's really boring at rallies, and he lacks a certain masculinity to his appearance even while being the only candidate with a beard in 100 years.
@becauseyoudo exactly walz has that relatability. He's very good on the ground. Vance is extremely good at debates. Which there was just one so it wasn't a wise thing to put a lot of weight on.
At the end of the day if harrise doesn't win it wasn't becyof walz. She's running a near perfect campaign. If she looses I'm just gonna attribute it to racism and sexism.
@jim I didn't think it would be Vance, but I was pleasantly surprised. Vance was always a great pick.
@AndrewCrosse I wouldn't call Vance "extremely good at debates." If you look at the post-debate polling, he won by like... 2 points.
@jim you mean like, today when I thought JD Vance was a slick talking fascist who can't even order a donut right? and Walz was out there campaigning like a normal human being? Yeah I remember.
@FergusArgyll Kalshi is being arbed. Here we're basically tracking Silver Bulletin, which I think underestimates Harris but is still a reasonable thing to track.
@Joshua All the same, the movement is clear, if silver bulletin is moving (I don't know, I don't pay...) as well as this market, polymarket and Kalshi are moving, that's not just one crypto site getting gamed
@Joshua I can spot the difference between these two boxes in a millisecond though. The one on the left is clearly about 4% more blue on margin
@Joshua I can't see your reply, manifolds UI often breaks for me (it's probably my fault, I have all kinds of funny things running....)
That's why I often use the API for betting.....
Kalshi Presidential Election Odds live in Times Square: 52% Trump, 48% Harris.
https://x.com/mansourtarek_/status/1844137640934899809?t=DJ538hkblHQRjdhKEMXNuw&s=19
Kalshi runs the only legal, real-money, regulated Presidential Election Market, under the Commodity Futures Trading Commission (CFTC) which is overseen by Congress, protecting the public from market manipulation and fraud.
@SimranRahman Supposing leftist conspiracy nuts like Paul Graham are right that this is manipulation by right wing money, what is the play? People see Trump 5 percentage points up and suddenly decide Harris is a communist after all?
@skibidist and if that were the case, why not swoop in and take all the free money from the market manipulators?
@ChrisLloyd There isn't any free money, considering time value of money, platform risk, and crypto fees.
@skibidist There was a poll yesterday showing Trump up in Pennsylvania. A lot of people who hold hard-left views, especially those on Manifold, are placing too much weight on the national polls that show Harris up 4.
As I and others have been saying for months, there's only one state that matters in this election. Harris isn't there, and Trump is everywhere. That's why he's winning.
Trump's efforts in the state are paying off, and the poll broke through and turned the markets. This market is not representative of the actual odds because people are not spending real money here.
@SteveSokolowski Neither Trump nor Vance are in Pennsylvania today. Vance will be back there on Saturday. Trump is traveling through CO, AZ, CA... And that is why... it's a toss up.
@SteveSokolowski Why would a single poll showing Trump up shift the market? We've had plenty of polls showing Trump up in PA already. People who believe polls respond to polling averages, not whatever the latest single poll is.
@PlasmaBallin There haven't been many polls in Pennsylvania. Like with stock options, time decays the value of polls. So much time has elapsed since the last poll that a new poll should be more impactful.
@skibidist Stuff like this has happened before. In 2012, there was a single trader who lost $7 million betting on Romney in a prediction market, and he significantly inflated the odds. It's impossible to know his motivations for sure, but it was speculated that he was intentionally trying to make Romney's odds look better to help him in fundraising, enthusiasm, campaign morale, and turnout, or to get favorable media coverage based on the perception that he had momentum. So it's very easy to see what the play could be, and believing that a lone actor's motivation might be to help Trump in some way is far from "conspiracy nut" territory.
@SteveSokolowski There are polls quite often in Pennsylvania, and the polling averages take time decay into account. 538 still has Harris ahead: https://projects.fivethirtyeight.com/polls/president-general/2024/pennsylvania/
As does the Silver Bulletin: https://www.natesilver.net/p/nate-silver-2024-president-election-polls-model
@PlasmaBallin One of the most important things that I take into account in my stock trading models is the trend. Today, I cut my losses selling NVDA calls at -1500 and sold puts instead, because the market wasn't moving down fast enough as the model predicted and the trend is up.
I think a similar thing is occurring here. The trend has been, for this entire election, that in the absence of news or debates, Trump gains. Nothing is happening for Harris at the moment.
Markets are range bound 66% of the time, and then there are violent moves when the range is broken. There was "smart money" accumulating for Trump based on internal polls for some time, until the A/D line became too high for the normal sellers to bear.
What's happening here is entirely expected - the markets will move violently up for Trump until they settle at a new range in a few days; my guess right now would be 57-60% as a first target. Those are real money odds, not Manifold odds.
@SteveSokolowski Coincidentally Nate Silver just released a newsletter about why internal polls are not very useful data: https://www.natesilver.net/p/why-you-should-mostly-ignore-internal
So I don't think there is any smart money for Trump accumulating based on that. Especially because any internal polling is either not released publicly (and therefore not known to the traders unless they're campaign insiders) or already included in polling averages.
As for the point about trends, the election models take that into account too. I know that the Silver Bulletin model explicitly models candidates' momentum. And it hasn't suddenly shifted towards Trump in the way prediction markets have. It's stayed stable and still slightly favors Harris. You can't blame market dynamics or conventional wisdom taking too long to update for this.
@PlasmaBallin Still, if you supported Harris, wouldn't you be very concerned that the most recent public polls in Pennsylvania show the state has flipped? We know from past experience that there are Trump supporters who don't answer Internet polls and show up to vote on Election Day, and many polls are now conducted online.
@SteveSokolowski Polls don't "show that the state has flipped". The polls show PA being very close, and just as a matter of statistics (not even including systematic errors), that means some polls will show Harris ahead and some will show Trump ahead.
Am I concerned that the polls are close? Yes. Am I concerned that the last couple of polls were slightly redder than the previous round? No, not really. Or at least, not any more than I already was. The difference this makes to her chances of winning is very small, and the supposed trend has a significant chance of just being noise. I will be more concerned if lots of polls come out showing Trump ahead, such that the polling average shifts in his favor.
@PlasmaBallin Yes, they do show that Trump has taken a decisive lead.
Every poll before this week showed Harris in the lead. Every poll since has shown Trump winning. There are now multiple polls showing Trump ahead, including some by reputable pollsters like the Wall Street Journal. You can eliminate that one poll that some pollsters have criticized for underrepresenting black people in Philadelphia and still find that all polls taken recently have Trump leading in Pennsylvania.
I don't know why Manifold is so stuck on this. If you, or others, want Harris to win, then that's fine and it's what elections are for - go campaign for her. But this is a prediction market, and the data is very clear, with multiple confirmations that Trump is now favored to win.