All the polls said that Trump would lose last time. The problem with polls is that people lie to themselves and the pollsters. Or the pollster never reaches them in the first place. And there is no effective way to predict turn out. This is particularly true in the COVID era which will change behavior significantly.
Trump as a majority in the only place it counts given that the election will end up in court most likely the Supreme Court. BTW I am a bible thumper (or certainly would like to be).
(CNN) Six in 10 debate watchers said former Vice President Joe Biden did the best job in Tuesday's debate, and just 28% say President Donald Trump did, according a CNN Poll of debate watchers conducted by SSRS.
So, in the ratio of “best job” in the debate, Biden had about twice as many approvals (60%/28%, or 2.1X).
Another technique being used is to make voting against Trump the default option if you wish to escape the insanity of 2020. Many states, for example, have promised to end the Covid restrictions after the election. The media is injecting the term “return to normalcy” when discussing a Biden victory. In other words, having created a highly unpleasant environment for the public, they are now talking about the one door through which people can escape the madness of the current year.
With nonlinear odds-to-probs conversion, the Democrats’ lead in probability is considerably more substantial: D: 74.9%, R: 25.0% (based on Betfair decimal odds: D: 1.5, R: 2.5, I: 201). This is an almost a +50% advantage for Biden, vs the linear RCP edge of +16%, if one uses this (unadvised) linear metric. For the more-meaningful nonlinear ratio metric, see below the chart.
NOTE: From a ratio standpoint, these odds imply that Biden is now 3X more likely to win the election than Trump. Just a day before, pre-debate, Biden was only about 1.6X more likely (Betfair, R/D odds-ratio of 1.2/0.73).
President Trump and the First Lady are diagnosed with COVID-19, which confounds the oddsmakers … only one betting house, Smarkets, has a line on Election 2020 at press time.
Like Real Clear Politics, Smarkets helpfully calculates -- linearly -- the implied probabilities from their odds, which understates the likelihood of the favorite.
UPDATE (10/2): Just a few hours later, the Smarkets odds are now D:0.51, R:2.2, I:299 ... which implies over an 81% probability of the Democrats winning the Election. (Other betting houses still have suspended wagering on this event with all the uncertainty, what a bunch of weenies.)
NOTE: From a ratio standpoint, these odds (R/D = 2.2/0.51) imply that Biden is now 4.3X more likely to win the election than Trump.
The Tale of Two Women (10/16) … perhaps we’ve reached a crossroads, and maybe a turn is preferable to just barreling straight ahead. Forward!Speaking of barreling-straight-ahead, business-as-usual … an unfortunate historical analogy, re: passively accepting oblivion (10/16).
With President Trump out of the hospital and back in the saddle, let’s revisit the odds-implied probabilities ... Smarkets is now joined by Betfair and four other betting houses in providing odds on Election 2020.
RCP uses the linear odds-to-probs conversion in determining the average implied probability: 61.8% for Biden, and 37.0% for Trump. These linearly-derived percentages can be corrected into nonlinear probabilities with the simple formula, for two-outcome events (see post):
This algorithm translates RCP’s linear probabilities into the proper nonlinear odds-to-probs: about 74% for Biden, and 26% for Trump … or a 2.8Xadvantage in likelihood of election (this is calculated from the average of six inputs, not just Smarkets, as the above post).
Then again, the Trump campaign has been here before, at the same juncture -- in 2016 -- but then even more so ... see below a chart on RCP’s probability of a Trump victory in both 2016 and 2020 (linearly derived, so overstate the underdog’s implied probability) plotted against the number of days until the Elections.
NOTE: Just before the 2016 Election, RCP's linear probability for Donald Trump was about 13%, which had converted to a nonlinear probability of only 2.2%... which translated to more than a 44X advantage for Hillary Clinton in odds-implied probability.
Vicereported on Monday that the Zoom call, which was described as an "election simulation," featured Toobin's New Yorker colleagues Jane Mayer, Masha Gessen, Andrew Marantz, Jelani Cobb, Evan Osnos, Sue Halpern and Dexter Filkins playing various roles including President Trump, Joe Biden, "establishment Republicans," 'establishment Democrats," and "the military." Toobin was playing "the courts." [How fitting]
Modern theorists argue that morality is the spring of democracy. It may be that the driving force of liberal democracy is the absence of any moral code. The liberal democratic ruling class is one without scruples. The heart of the empire is an open air market where everything is for sale.
Concerned about a possible future backlash from disgruntled Americans if this conspiracy natural chain-of-events plays out, just as projected? Well, don’t worry … the Department of Homeland Security has that potential ‘extremist scenario’ already covered (9/4, more here and also here), and has identified a domestic group that poses a threat “more significant than the immediate danger from foreign terrorists.” [from earlier post]
THAT CERTAINLY DIDN'T TAKE LONG Category (Part 1):
Just a few weeks ago, the Department of Homeland Security declared that ‘white supremacists’ are the biggest domestic terrorist threat. Well, through “the efforts of more than 200 state and federal law enforcement officials”, we finally can see what these nefarious types look like, as the FBI uncovers an armed militia plot to abduct Michigan Gov. Whitmer (10/9; more here, including reference to ‘FBI handlers’ in the plotting, similar to what's been done to snare Muslim wannabe domestic terrorists).
Their 'supremacy' appears to be covert ... but not for eager news outlets...
THAT CERTAINLY DIDN'T TAKE LONG Category (Part 2): 'White Supremacist' Narrative Unravels(10/11): Whitmer kidnap suspect attended BLM rally, another called Trump a 'tyrant'.
Something that should bring Americans together, in these trying times ... "it appears that the FBI busted an anarchist, anti-government militia which plotted violence against elected officials - yet hated both sides of the aisle."
UPDATE (10/18): Why Facebooksuppressed Hunter Biden revelations ... that FBI FB decision was made by an ex-advisor on Ukraine to VP Biden, of all places and people. Gee, what a small cozy world, at least at the top.
Everything is designed to affect the election, and if Biden wins, he’ll need a strong economy to win again (or Harris will); thus, everything will return to normal and America will treat COVID just like Biden and Obama treated the Swine Flu…like it never existed.
With the dust settled, we find that odds-implied probabilities now stand at Biden 76.6% and Trump 23.4% (nonlinear correction of Real Clear PoliticsAverage, 10/31; updates here), or almost a 3.3X advantagefor Biden.
You say you want a revolution..?!?
Well, this Brit says we can't pull off thatsh*t nowadays (10/17), thanks to our dependence on government and separation from tangible production (agriculture, industrial; see chart below). The critical skillset of self-sufficiency is not who-we-are.
Despite most Americans no longer having the skills and capability to farm nor manufacture goods, our nascent industry of 'PREEMPTIVE DEBUNKING' is very healthy and growing (10/20, video; perhaps that's our 'revolution').
This hasn't yet been approved by Big Brother, but ... Economic considerations in Election 2020, in light of the “largest social and economic experiment in American history: the COVID-19 lockdowns” (10/29).
Post-Election nostalgia:Trump as Nixon, the focus of fading radicalism? (11/5) … “The generation that made a fetish of free speech, good government and democracy, is now closing the show with censorship, corruption and now the undermining of democratic legitimacy.” Earlier, the same pessimist also offered that perhaps we ought to have the UN provide adult supervision of our chaotic Elections (11/4). Here’s his scarier pre-Election take on the country's dynamics (11/3).
Then again, nationwide, the odds-implied probability shows a Biden victory as still the heavy favorite, with Biden 77.6% and Trump 22.4% (nonlinear correction of Real Clear PoliticsAverage, 11/1), or almost a 3.5X advantagefor Biden.
A leading cliodynamist -- with an excellent track record with data-based predictions – playfully posits himself as an alien from Alpha Centauri for the ultimate nonpartisan perspective in this structural demographic view of our “post-truth world” (11/1).
‘It’ sees the United States as an “extremely fragile state, which in technical terms is known as the revolutionary situation” ..and.. that we are “on the cusp with a highly positive Lyapunov exponent”, where “unless there is a clean win, we will be in situation where possible trajectories start diverging dramatically.”
The alien doesn’t see ‘the problem’ as coming from the masses … rather the division of our elites in their worldviews is driving this. It also notes that secession(s) and/or Civil War II is also possible. And violence could always be delayed … the American-history-buff alien points out that while the abolitionist Republican Abe Lincoln was elected in late 1860 -- outraging the slave-owning Democrats, leading to the South seceding -- the first battle wasn’t waged for 7 months. But that was before the internet.
Yes, it all seems far-fetched, and certainly couldn't happen here, but surveying today's sociopolitical dynamics can lead folks to differing viewpoints on what the present is, and what the future holds:
UPDATE (11/9): Like with that election, we may not know what’s next, but Team Deep State probably has a pretty solid idea. And they've certainly factored in that rebels from Team Red, loyal to the despotic ruler in the process of being deposed, may be, uh, problematic.
[ЗДЕСЬ НЕЧЕГО СМОТРЕТЬ, ДВИГАЙТЕСЬ ДАЛЬШЕ]
Товарищи... Скажите Владу, что его идея использовать "теорию Ляпунова" товарища Турчина была блестящим маневром и, как предполагала проклятая теория, полностью вызвала желаемую реакцию. Благодаря временному запрету на этом сайте эти бедняги пропустили всплеск вероятности в ночь выборов, когда Трамп достиг пика в 98% после закрытия избирательных участков, и это число резко упало на следующее утро, когда по почте пришли сомнительные бюллетени, почти все для Байдена. -- очень "неожиданно", но достойно Сталина! -- и это было очень неловко ... и очень смешно! (Но коррупция глубока, так что не волнуйтесь.) Продвигаясь вперед, особенно теперь, когда элита считает, что у них есть "мандат с Байденом" (забавный каламбур на английском языке), подавление даже умеренного инакомыслия американским медиаклассом и глобальной элитой обеспечивает долгосрочную нестабильность в стране, что, несомненно, порадует тех, кто извлекает выгоду из угасающей американской гегемонии. Наряду с их постоянной поддержкой усилий президента Си по сеянию социального раскола, инакомыслия и хаоса в Америке, полезные слуги из администрации Байдена, средств массовой информации и корпораций доступны по запросу. Кстати, вам действительно нужно, чтобы я снова работал под прикрытием в Давосе, чтобы записывать достойные шантажа разговоры о теневом процессе Западной глобальной элиты, таком как выборы 2020 года? Это поможет вам получить выгоду со всех сторон. Всего наилучшего олигархам. С уважением... Нелинейный агент с нулевой суммой
It’s safe to assume that a fair number of TalkStats members are highly-trained statisticians, and perhaps practice the art/science vocationally, maybe even in the political realm, and so may be able to offer some perspective and guidance on the following:
With the incorporation of said homogeneous data batch into a larger set of heterogeneous incrementally-incorporated data, how does that impact the credibility of the agglomerated database, especially when the competing outcomes are very close in their relative count? (Perhaps we should consult with experts on polling-quarantine methodology.)
NOTE: As has been earlier reviewed elsewhere ... while percentages may be convenient in relaying a piece or set of information, it is generally inadvisable to use these percentages in further calculations, as that can obscure, convolute and even corrupt more-general information, and therefore conclusions. To manipulate information beyond a singular statement, one should use ratios ... RATIONAL numbers (more here, on a parallel topic).
Trump improved his relative support from 2016 to 2020 with every demographic group -- exceptWhite Men -- so he apparently is a pretty rotten 'sexist' and 'racist'. In addition, turns out that Trump was a crummy authoritarian too, as he was “easily beaten” (11/6).Orange Man Bad, indeed.
‘Red Flag’ Nation (11/5; rude words) … with tomfoolery suspiciously all in one direction -- another statistical anomaly:
When you are auditing you see mistakes happen all the time. Humans make errors. Except in real life, mistakes usually go in different directions. When all the mistakes go in the same direction and benefit the same parties, they probably aren’t mistakes. They’re malfeasance.
On the statistics-and-probability front, ‘voting irregularities’ only appear in critical swing states … such a befuddling puzzlement, that will surely be glossed over and soon, since Reasons (11/5).
On a less-statistical note but still pretty unidirectional, the shocking lack of landslide shows that Americans have rejected Science! … say American scientists (11/6) ... "With so many votes cast for Trump in US election, some researchers conclude that they must work harder to communicate the importance of facts, science and truth."
With Election 2020 being 'called' for Biden (11/7), in a race he "made about character", perhaps Benford's Law and associated statistics-and-probability theorems and methodology need to be under lockdown until further notice ... as this type of credible number-theory study meaningless drivel starts to trickle in...
On (probable) over-optimism in 'D.J.'s eventual favor, re: state ballot challenges (11/8), from a guy who claims to have a degree in Statistics, so you know he’s credible. He continues with a possible scenario, if the pending ballot investigations can not be done quickly and cleanly: Electoral College score: 26-24 ... “All politics are local” … more (delusional?) D.J.-optimism on election procedures as set by the Constitution (11/9).
On knowing things aren't looking good, freedom-of-analysis wise: And then they came for Benford’s Law..? (11/9, after the social-media authorities looked up the term, presumably).
On why you shouldn't care about a 'caring government' ... Hitler really really cared about things, a whole heck of a lot. You should not want your government to care that much. (‘Caring’ is meant to be on a local level, but don’t get me started on empathy logistics.) Anyway … thankfully we have incompetence (11/12). [In this case, that old rhetorical-boogeyman ‘Citing Hitler’ may actually win the argument, for a change. That whole Third Reich thing started within a democracy, after all.]
On how -- plus ça change, plus c'est la même chose -- ‘the more things change, the more they stay the same’: the danger of seeing your world through those darn flat glowing screens, getting your information and opinions from talking heads (flashback, 1976; note the high-quality monologue and cinematography, you certainly don’t see that anymore out of CGI-obsessed comic-book Hollywood).
On parallels with the USSR collapse: “Watch this!” … as our gerontocracy drives US into a ditch (11/11).
On suspecting any voter fraud, this guy certainly sounds like a crazy conspiracy theorist in a tin-foil hat (11/16): “The point is there is plenty of evidence suggesting something happened, more than enough to warrant asking rational questions and expecting reasonable answers … evidence, both circumstantial and direct, breaks down into six basic categories: Precedent, Motive, Foreknowledge, Opportunity, Voting irregularity, and Cover-up.” Obviously, he's deranged.
Since the American media ‘called’ the race for Biden, the conventional betting markets on Election 2020 have been shut down, probably for Reasons, and good ones at that.
However, PredictIt keeps the flame alive, with their cents-pricing of the possible outcomes, including Election 2020. Of course, one should not conflate cents-pricing with implied probability, despite both metrics being on the same 0-100 scale. Cents-pricing is linear, while the true monatomic odds-probability relationship is nonlinear … like with the conventional linear conversion of odds-to-probs, cents-pricing overstates the underdog’s probability, while understating the favorite’s probability.
Trump's implied probability of just over 2% (11/9) does kinda match about where he stood in odds-implied probability, converted from regular odds, when the other betting markets closed shop. In light of the political and constitutional nuances reviewed above, this number could be fun to watch (see below).