I'm a big fan of Gelman usually. His new regression book is fantastic and I wish it existed 10 years ago. That said he is behind the Economist model and it's ridiculous #'s. As someone who actually bets real money on his models, if I have something where my model wants me to bet 60% of my roll on in a large liquid market my first instinct is to trash the model not to run and bet it everywhere.

Holiday, your attempts to try and portray Biden as some sort of evil pedophile are hilarious quite frankly.

In my opinion, this year's election will be an anomaly and in the end a court will decide who the next president is. Let’s not fret the 2004 elections when Gore decided to drop the case after 40 days of recounting in Florida. Gore had the chance to use more constitutional tools, but decided not to divide the nation further. Let’s not forget also the 2016 elections when Clinton got 3.000.000 more votes than Trump, but failed to be elected.
With the nation divided to extremes, there is always the possibility of a constitutional anomaly. A very interesting post I read very recently talks about the possibility of Nancy Pelosi, Chuck Grassley, Mike Pompeo and Steven Mnuchin as next Presidents. You will not even find these names in any bookmaker, but according to the constitution, if no President is elected by January the 20th, these people have good odds to be the next POTUS.
Read here: https://arbusers.com/index.php?topic...79060#msg79060
In the end, I simply hope that everything goes smoothly and we have a POTUS on time with no adventures. Be extremely careful with the T&Cs of some bookmakers as they might hide surprises.

This T&C comment is on point. The manager of my local's service admitted in an email exchange that their rules* COULD result in Trump bets being graded as losers, yet he gets inaugurated on Jan 20.

* - Their official policy will be to check the New York Times at 9 AM Eastern each day for a declaration of the winner.

Discuss.

The most recent post does point out that that his model is designed to predict "voter intent" and not actual "outcome" and criticizes the economist for not highlighting that fact. https://statmodeling.stat.columbia.e...e-predictions/

Their models seem to perform really well in backtests (https://github.com/TheEconomist/us-potus-model), which would indicate that the current fair market prices on BM and the such are where they are because they bake in a very high chance of shenanigans. My amateur guess is that the polls we are getting data for now are not as relevant as they were in 2012, or even 2016.

Also interesting is that the prices on pokershares are +189/-222, which implies poker players (a more data driven demographic than most) are coming in harder on Biden, so they must trust the numbers more. Or perhaps all the other books are getting tons of Trump action at +163 and are loading up on Biden at that price point, which is also quite possible, as Bovada has it at +145/-175.

I'm not well-versed in stats at all, but all of this has been super interesting to read and follow. What a week we have in store for us.

"Despite featuring the duo of LeBron James and Anthony Davis, and having two major-market teams playing in the Los Angeles Lakers and Miami Heat, Game 1 of the NBA Finals on Wednesday night drew record-low TV ratings.

The game -- a 116-98 Lakers win -- drew 7.41 million viewers to ABC, according to overnight ratings. The previous low was 8.06 million viewers for Game 3 of the 2003 NBA Finals between the New Jersey Nets and San Antonio Spurs.

The number is also a 45 percent drop compared to the audience for last year’s Finals opener between the Toronto Raptors and Golden State Warriors."

https://www.cnnphilippines.com/sport...g-history.html

I feel like these type of things matter. The NBA completely embraced leftist politics and China and it is a complete bloodbath for them. The viewership tanking is a pretty realistic looking poll. BLM and the riots were wildly unpopular. The riots actually stopped once it started showing up in the polls. That was the Biden low point on chance to win by the betting odds.

i mean, i have a 50k bet with a 2p2 poster just IN THIS THREAD, but sure bud, i am definitely lying. YA GOT ME.

lol facebook

nailed that read. trumpers and bern outs are as predictable as the sun rising. you can type their responses before they make them!

gotcha, no yeah, you're definitely apolitical and just stumbled across gateway pundit, which is in the hall of fame of most vile, most lie-filled websites on the internet. ������

and then you decided to come to a betting thread and post said gateway pundit link as you furiously JAQ yourself off to pictures of Hunter Biden and someone you (obsessively) claim is a young teenager. you are Just Asking Questions™ as you ramble incoherently, like a raving and roving lunatic, about hunter biden. **** off back to your facebook "news" feed and chain emails. here's another read: i'd say it is a mortal lock that you've sent an email before that contained "funny images" of the obamas next to pictures of monkeys.

look Holiday In The Sun, if you're still with me and your brain hasn't yet melted into a slop of gelatinous goo, then please, i implore you to listen: you are braindead. stop spamming us, stop staring at pictures of very young teens, and seek help.

JSmith27 enters the chat.

I accepted your Trump bet a couple months ago in a different thread, then when we went to DMs you demurred.

Now that you're posting in the....hang on, let me double check here...okay....betting thread, would you like to now make a bet on Trump to win the election?

(Per electionbettingodds.com)

From 9am Weds to 9am Thursday, Biden's win percentage is up:
2.3% in Arizona
2.2% in Florida
7.5% in Georgia
3.8% in Iowa
3.6% in Michigan
3.7% in Minnesota
3.7% in Nevada
3.7% in North Carolina
4.2% in Pennsylvania
6.1% in Wisconsin

His win percentage went down 0.4% in Ohio.

Overall, his win percentage went up, from 62.4% to 62.6%.


To me, there are two reasonable theories why the betting markets aren't closely following the polling:
A. Sophisticated gamblers factoring in polling errors and the potential for fraud
B. People gambling on this election have a Trump bias

The irrational moves between the overall election odds and the underlying states odds tells me that we are seeing a lot more of group B than group A.

i think it is mostly demographics

trump does best with men, he does best with GenX, and he does best with non-college educated. prime SPORTS BETTING demographics.

these groups likely also have low "social trust" -- which would measure whether, for instance, they believe in things like the news and polling. low social trust and voting for trump are well-correlated, which is one reason that polling was off in 2016. trump's voters were harder to reach, and this was not accounted for in polling (now many polls weight by education, with the goal of recapturing these white working class whites with low social trust that polls missed in 2016).

At the very minimum, a majority of people who bet on sports are going to be Trump supporters or at least Trump-curious. It might even be a large majority of people who gamble as a hobby on sports are amenable to Trump.

Betting markets move on sharp money. And Trump bettors are more likely to have actual money available to bet.

Maybe I should have put more emphasis on this, but the big moves in the underlying states did not make a big move in the overall line. This was a very irrational 24 hours within the betting markets.

So if the big question is:
Are the betting markets doing a rational job of digesting the polling data, but correcting for polling bias and potential fraud? Is it rational for Nate Silver to say 90% and have the betting market say 60%? Or is it just a big Trump bias in the betting markets?

This is evidence that there is a lot of irrational behavior in the betting markets. If it was a genius pulling levers and making big bets, you would expect the whole to equal the sum of the parts regarding the overall election vs the underlying states.

Yup, that is the big thing here. When i have a model that likes a bet by 40 percent off from the market, i usually assume the data didnt scrape correctly or some bug happened.. but hey, we will see soon enough! I agree with the earlier post saying this could go to the supreme court and be careful who and where you bet it with, could be some weird rules, be awhile before this bet is settled

Nate has the Broncos at 57% on Sunday, are the gambling markets irrational at +155?

Comparing registered voters to players on a sports team is a poor analogy. In your analogy I think registered voters is the sum of players, coaches, staff and fans.

In game 6 of the world series the Dodgers had the same number of hits as the Rays and for some reason the Dodgers were given the title and that is after a game 5 where the Rays had 7 hits to the Dodgers 6 and the Dodgers were given the win in that one too. I think Hillary and the Rays are getting screwed.

It's very unlikely that the election is decided Tuesday unless Biden wins big. The media won't call it for Trump under any circumstance and if it's close neither is going to concede.

where is this? for a while i wasn't following this thread so i missed that.

i literally said it's a fringe outlet at best. but what can you do when the main media outlets have a blackout on any negative information on their chosen candidate, and all the popular polling experts say biden is going to win 538-0? have to dig a little deeper. i would think if you're yolo'ing $300k on the old boy you ought to be doing your due diligence on what could potentially sink him.

speaking of -- when you say you've got 300k on the election, is that on biden? is that your liability? or have you got it spread across both sides? what's your max loss here?

i don't think there were many good obama memes and especially not any with monkeys. my favourite biden meme, however, is this one:



honestly, the meme of the campaign imo. what do you think? did you get a chuckle out of it? i hope you weren't too uptight due to your obsession with the old boy to find the humour in this one.

So if trump does pull it off, all you super condescending smart guys in here will never post again, right?

To answer your question Poogs, NO. And why will I continue posting you ask? Because I'm not super smart

Tucker Carlson supposedly lost the only copy of the emails ha had. Come on guys. Everyone with half a brain knows that it's a hoax. There is nothing. It enough for the right to spread the rumors and have people talking about it. That's their whole strategy.
And besides, anyone with a brain knows Trump is the creep.
Lol

Edit: I thought the people here were smarter than that.

This factoid keeps getting repeated but polling wasn't really off in 2016, at least not polling that weighted by 2012 vote.

Final 2016 you gov poll (real results in parentheses)
Trump 41 (46)
Johnson 5 (3)
Other 3 (1)

Trump + Johnson + Other (lot of McMullin) = 49 (actual 50)

Clinton 45 (48)
Stein 2 (1)

Clinton + Stein = 47 (actual 49)

Not sure 4 (0 - nobody actually votes "not sure")

There is no obvious over-sampling or under-sampling of right or left in that poll.

Either
1) right through the campaign polls from all companies consistently 2x-ed the actual support for third party candidates - libertarians, greens and evangelicals - perhaps their voters have very high levels of social trust and there is a serious problem causing them to get sampled a hell of a lot at the expense of the main party on their part of the political spectrum

2) polls are just a snapshots not predictions and voters read them too. The above (and more importantly all other chatter going around) was sufficient motivation for third party voters to get behind the candidate from the larger party on the right or left of the spectrum on election day, i.e. after the polls were taken, particularly in the states where it mattered.

I think 2).

Saw this IG post from the Kinder Institute at Rice University. The counties in which the cities of SA, Houston, Austin and Dallas are in is about 50% of state population. These counties are blue. I also drove through rural Texas a few weeks was surprised how many ranches had Biden signs.

https://www.instagram.com/p/CGp9zXTHvEj/?hl=en

Highly unlikely that Biden wins but it’s getting closer. That is unless the state starts seeing an exodus of people. Houston which is so tied into O&G is already seeing a lot of the educated work force in this field leaving the state. O&G bleeding jobs plus people getting tired of terrible summers full of hurricanes and leaving might reverse the trend.

Not nearly as hilarious as your attempts to argue gambling markets against guys who have made tens of millions of $ betting when you need help with a 1st grade math problem https://forumserver.twoplustwo.com/4...right-1780501/

I'm still not sure how it's possible to irrecoverably lose emails.

not at all what he said. Also shows that perhaps you're not aware of the situation? If its the laptop emails, there are multiple copies of the hdd floating around.

His sports models are garbage his political models are not garbage

That seems to be a hard concept to understand for some people ITT