How about if we start the day with one thing serious and educational and one thing idiotic and hysterical? Idiotic takes precedence, of course. This is America... although the piece comes from the U.K.'s Daily Mail: Embattled QAnon congresswoman Marjorie Taylor Greene 'openly cheated' on her husband of 25 years with a polyamorous tantric sex guru and then moved on to another affair with the manager at her gym. Martin Gould and Ruth Styles reported yesterday that "Neither man denied the affairs when approached by DailyMail.com. Craig Ivey, the tantric sex practitioner, said: 'I will not respond to anything about this,' while the other man, Justin Tway, said: 'I have no interest in talking about anything to do with that woman. Everything with her comes to no good.' But others say the new representative from Georgia's 14th Congressional District was brazen about her affairs which she carried on a decade ago while working in gyms in Alpharetta, Georgia."
'It wasn't a secret. Everyone who moved in her circles knew about both the affairs,' one man who worked closely with Taylor Greene told DailyMail.com.
And her boss, billionaire's grandson Jim Chambers, added: 'She socialized a lot with us. I remember one particular pool party where she was lying draped over Craig's lap drinking a beer.
'She was quite open about it,' he added. 'We all thought her marriage was falling apart.
'But then I took my son to her son's birthday party and things seemed totally normal between her and her husband-- even if they were a little lukewarm.'
In response to DailyMail.com's request for comment, Taylor Greene called the story ‘ridiculous tabloid garbage spread by an avowed Communist,’ and ‘another attempt to smear my name because I’m the biggest threat to the Democrats’ Socialist agenda.’
...Taylor Greene, 46, has become notorious since being elected to Congress. She calls herself a 'strong conservative Christian,' who planned to take her 'family values' to Washington.
...Rumors of Taylor Greene's affairs have been around since she hit the headlines in the summer by winning the Republican primary for her seat, which covers a wide swath of rural north west Georgia.
Chambers even tweeted out at the time: '@mtgreenee so when you worked for me at CrossFit Alpharetta in 2012, the extramarital affairs you had with Justin and Craig, those were Jesus-approved, right?'
He then added: 'I would never choose to doc someone's infidelities...unless they're this person,' and linked to a Marjorie Taylor Greene for Congress page.
A New Yorker profile of her said a reporter had seen texts in which Taylor Greene admitted sleeping with one man outside her marriage.
Turning the page rapidly, serious story comes from Change Research and you may or may not find it as fascinating as I do-- and if you don't... just read about Christian Family Values idiocy again. This one tries to get to the bottom of why polling sucked so badly again in the 2020 election. Examples of suckage:
The question now is why, and what is there to do about it? Thirty years ago, two out of three Americans would interrupt dinner, pick up their landline phone, and answer questions from a stranger. Today, only about one in 100 adults will answer a call from an unknown number to take a survey over the phone. With Americans not answering their phones, the tables have turned-- it’s survey takers who are choosing to take polls, not pollsters choosing survey takers. The fundamental assumption that has historically underscored political polling-- that pollsters can reach a random sample of voters-- is no longer valid. When millions of people refuse to participate in certain types of polls, it is mathematically impossible for those surveys to provide random samples. The solution is a wholesale rethink, drawing on technologies and business practices now widely deployed by consumer-facing businesses on digital platforms. The path forward requires pollsters to embrace the mentality of modern data science. We know that approach is possible because Change Research has taken a data science approach to polling since its founding-- with success. In 2020, Change Research’s private and public polls were nearly 25% more accurate than other pollsters in battleground states-- with an average absolute error of 3.5 points in the presidential race across all polls in Arizona, Florida, Michigan, North Carolina, Pennsylvania & Wisconsin in the two weeks leading up to November 3rd. Overall, live phone polling, once considered a “gold standard,” fell especially short: live phone pollsters were 23% less accurate than polls conducted online and 48% less accurate than Change Research’s polls.
Just before Election Day, Change Research also released state-by-state election forecasts. The forecasts used a machine learning model trained on data from hundreds of thousands of survey responses to compute election predictions for each state. The model broke the population into different subgroups, stratifying voters along a number of different qualities including age, ethnicity, and education. For each subgroup, we estimated the percentage of people in that subgroup we believed would vote for Joe Biden. Our model incorporated data from hundreds of polls run in 2020 -- dozens in some states, none at all in a few others -- to come up with figures for each state. This machine learning approach picked up political signals that most poll-based forecasts did not, including Democratic losses in the Northeast and races closer than most polls forecast in the Midwest. That model had a statewide average error of 3.6 points, more accurate than poll-based models from The Economist and FiveThirtyEight.
However, this campaign season, we also experienced some of the challenges of relying on the certainty of random sampling alone. In 2020, we surveyed over 10,000 people in Miami-Dade County in South Florida. We saw that Democrats were losing ground: eight percent of Miami-Dade County 2016 Hillary Clinton voters would flip to Trump, our polling suggested, a number far higher than every other major county in the U.S. (for most counties, between two and four percent of 2016 Clinton voters voted for Trump). On the surface, that sounded smart: Polling successfully uncovers rightward shift in South Florida. However, election returns showed that our polling - and almost all other polling of Miami-Dade-- underestimated the size of that shift. Change Research’s numbers pointed to a 20-point Biden lead in Miami-Dade County, a modest change from Hillary Clinton’s 29-point lead there in 2016. On election night, Joe Biden won by only seven points in Miami-Dade County. Election returns suggest that about twenty percent of 2016 Clinton voters in Miami-Dade County voted for Donald Trump in 2020-- more than double what polls suggested. In other words, survey takers in Miami-Dade County were simply not reflective of the full electorate. That’s why we believe that an overreliance on the certainty of random samples alone can limit the ability to see the full picture, and that innovation in the polling sphere is much-needed. It’s time for a different approach. For Change Research, that means we’re focusing on deeply integrating data science, machine learning, and the wealth of data from our over 3 million survey respondents into every poll and every offering. We will embrace new tools - like custom data modeling for campaigns of all sizes, and continue to lead the way in finding new ways of reaching voters. A modern approach to polling will allow our leaders-- political, social, corporate-- to better hear and understand what customers, voters, workers, and others are going through. With millions of data points, modern polling will yield a deep quantitative view of the population and lead to better strategies, better communications, and a more humane, scientific, and just America.
