This is a continuation of thread: Nonlinear odds-to-probs conversion
Drudge links to an
odds-implied probability study (7/1) that utilizes the conventional
linear odds-to-probs conversion, where each candidate’s American-format odds are used to estimate their probability independently, and then all the candidates’ probabilities are normalized to sum to 100%.
With the
nonlinear odds-to-probs determination, Joe Biden has over a
74% probability in Election 2020.
The combined probabilities of the
Republican candidates – Trump, Pence, Haley and
The Rock(?) – total a
meager 24.2%.
On the other hand, a Stony Brook professor projects that Trump has a
91% chance of reelection (7/2), claiming to have a model that’s been correct over 90% of the time, looking back over 100 years … but he doesn’t share his crystal-ball metrics, other than Biden placing 5th in the Democratic primary in New Hampshire back in February, as well as his being correct in predicting Trump's election in 2016, as opposed to everyone else.
-----------[ WAIT ... THERE'S MORE!!! ]------------
Kanye West jumps into the race, and clocks the same probability as Hillary Clinton. Biden's slidin' somewhat in probability, but remains way ahead.
Kanye’s odds jumped by a factor of 10X – from 500/1 to the 50/1 above -- after
getting supportive tweets from tech moguls Musk and Cuban (7/5). In fact, the music celebrity garnered the majority of bets in the market since throwing his hat in the ring. Pundits suggest that a third-party run by
Kanye could strip votes from the Biden campaign (7/5). Then again, perhaps Biden could ask Kanye to be his VP candidate [
no odds offered on that action yet], though he's not the right gender.
It would be interesting to see a betting line on who’s President in, say, 2022:
The coming 'Biden coup' (7/6). More crazytalk on
the 'COVID coup' (7/17).
-----------[ WAIT ... THERE'S MORE!!! ]------------
The betting odds and their implied probabilities on Election 2020 are surprisingly stagnant, perhaps tracking one party's innovative campaign strategy, preferring to keep the focus on the '
referendum' on their
uber-active opponent, who "is not a legitimate candidate", despite being the incumbent. Heck, things appear to be going swimmingly for the challengers, so why even bother to introduce any uncertainty into the proceedings ... like quaint, outmoded concepts such as
publicly debating policy, as that could inject messy uncertainty (8/2), along with providing memorable entertainment.
NOTE: In the interest of reducing uncertainty, the
parties' odds are used as the metric here, rather than the odds of specific individuals -- enough with this
‘identity politics’! -- as in the
Time of COVID-19, anything can happen. Or
not happen, of course.
-----------[ WAIT ... THERE'S MORE!!! ]------------
Drudge headlines a
Reuters article on implied probabilities from
Ladbrokes odds on Election 2020:
Betting markets favor Biden over Trump, but odds narrow in U.S. race (8/6). Reuters utilizes the
conventional linear odds-to-probs conversion, and so
understates Biden’s advantage (see above), claiming only a
61% probability for Biden and 36% for Trump.
Ladbrokes head of political betting notes that 1) Biden's surge in odds and implied probability were concurrent with
BLM-inspired protests, and that 2) betting markets are "overwhelmingly dominated by men", and that may provide a bias towards Trump, who generally polls better with men than women.
Furthermore, as a show of his robust Electoral strength, Biden’s odds-implied chances were only minimally impacted by the recent airing of his innovative views on
cognitive testing (8/5) and the
diversity of his core constituency (8/6). Then again, from the Big Picture standpoint,
the same mysterious forces – bewildering to our Coastal Elites -- are still at play now (8/2), as
they were in Election 2016.
