Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:00):
Prediction markets have been anything but quiet over the past
forty eight hours, especially as we get deeper into election
season and closer to key international developments. On polymarket, the
top volume markets remain heavily dominated by US political questions.
As of this morning, the will Trump win the twenty
twenty four presidential election market had over eight million dollars
(00:23):
in volume, with Trump's probability currently sitting at fifty six percent,
a slight dip from fifty eight percent just two days ago.
Joe Biden's corresponding market has inched up slightly to thirty
nine percent. The change seems modest until you consider that
these markets have remained stubbornly flat for over a week,
(00:43):
making a two point swing in that short a window notable.
But perhaps the most surprising movement came from the polymarket
contract asking will Joe Biden be the Democratic nominee in
twenty twenty four? That market dipped sharply from eighty three
cents to seventy six sense on Tuesday, triggering significant chatter
across Discord and Reddit prediction forums. The drop followed multiple
(01:07):
viral clips of Biden appearing confused during public appearances and
renewed scrutiny of his approval ratings. While there's no confirmed
effort to replace him on the ballot, the sudden shift
suggests growing uncertainty among traders. In parallel, Predicted saw a
strong surge in activity around House and Senate control. The
(01:28):
market on which party will control the Senate after twenty
twenty four showed Democrats dropping from fifty seven to fifty
two per cent in less than twenty four hours, largely
fueled by recent polling in Nevada and Ohio that showed
Republican candidates outperforming expectations. Meanwhile, the control of the House
remains tilted slightly toward Republicans at fifty three per cent,
(01:52):
essentially unchanged over the past week. Metaculous, with its community
driven forecasting model, offered some contrasts. Their aggregated forecast for
Trump winning in twenty twenty four still hovers around forty
nine percent, but forecasters are increasingly split. Interestingly, a question
asking whether a major third party candidate will win more
(02:14):
than five percent in the general election has jumped from
just under eight percent to over twelve percent this week.
That rise appears to follow the latest developments surrounding independent
candidate Robert f. Kennedy Junior, who recently secured ballot access
in two more swing states. While five percent may not
seem like much in a tight race, it could prove pivotal.
(02:37):
One emerging trend worth keeping an eye on is how
prediction markets are starting to respond faster to online sentiment
than traditional polls. We saw this dramatically with the Biden
nominee market this week, which adjusted sharply within hours of
the viral videos. That kind of rapid shift suggests that
traders are now waiting real time digital behavior as much
(03:00):
as polling data, if not more. This gives prediction markets
a unique sensitivity to narrative momentum, potentially allowing them to
pick up on political tectonic shifts before mainstream media even
takes notice. Thanks for tuning in and be sure to
subscribe for more updates like this. This has been a
quiet please production. For more check out Quiet please dot
(03:22):
ai