I think there's enough interest for a betting thread. Now I won't have to sift through pages of debating the finer points of Trump's racism to find out what 2+2 thinks of his chances to win South Dakota.

The two main betting sites that have been getting chatter on 2+2 are Predictit and betfair. As an American I only know about Predictit.
Pros:
The interface is nice and the betting is similar to intrade(really miss intrade). They will let you fund your account with an American credit card.
They have pretty liquid markets across a number of different bets. (like who will place 3rd in a state)
Cons:
One major downside is a 5% cashout fee.
Another downside is a 10% winnings fee, but that's not too bad because it's only on your winnings. If you make a $100 bet and win $0.10, you pay $0.01 unlike sports betting where the whole bet is taxed.
There is an $850 limit per contract.
The website is very slow on election days. Don't assume you can easily cashout a falling position during election result reporting. I hope they buy some new servers with their heavy fees for the upcoming general election.
Workarounds:
To avoid the 5% cashout fee you could play several 98-99c markets in a row to make 5% before cashing out. Sometimes after an outcome is decided bettors want to cash out their winnings at 99c to avoid waiting on predictit to assign the winner.
You can sometimes workaround the $850 contract limit by making the same bet multiple ways. Take Hillary to win the election and take the "next president is female" position.

Free money:
Right now on predicitit you can buy Hillary to win at 48c and buy "next prez is not female" for 44c. That's 92c for a $1. But like I said, if that's your only bet predictit will take 1c for the win and 5c for the cashout.
You're welcome.

Betfair:

Pros:
The software actually works, including on election night
Liquidity is great in the markets that they have
No silly $850 limits
Deposits are instant. Literally instant, for any amount up to five figures
Fees are a very reasonable 5% on winning markets, discounted through a rakeback mechanism for bigger amounts

Cons:
Very few markets. Like, right now, it's Nevada, nomination/GE bets, or bust.
Once you get into five figure bets you have to keep emailing them to raise caps for no apparent reason

Free money:
There's less of it on Betfair as of the moment of me typing this but that varies based on the phases of the moon. However, if you can get on Betfair (or Matchbook) and PI at the same time, there are huge arb opportunities. Lines can vary up to 10% between exchanges.

I just put in a small bet on Rubio at 10:1 in Nevada tonight. Twitter is going on about low turnout and there really hasn't been much polling in the state (none after the Jeb dropout). With the caucus format, lack of polling and unknown Jeb effect, I feel like a Rubio upset could be a good long shot bet.

Yes, right now Bernie is 10-11% on Betfair and 15% on PI to win the nom.

Rubio is also off about 5% for his party nom.

I think in the primary the odds to back sanders at betfalir is like 9 to 1 or something atm. I think this is a bit out of line since we cant know yet when the prosecution will come. I think its the president over at Juridical Watch thinks the prosecution will eventully come because the fact that emails are being released shows that the intelligence establishment and FBI essentially expects that it will happen.

in

Guess the early twitter turnout indicators were incredibly wrong and the Jeb support didn't help the non-Trumps.

Trump is going to take this down.

Morning Consult just released a new national poll. Does anyone know how to get the state-by-state data from that? link ?

National polls don't have state by state crosstabs, the sample size would be way too low.

I basically want to recreate this chart for Republicans.
http://fivethirtyeight.com/features/...he-nomination/

In the article Nate says he got the data from MorningConsult state-by-state results.

You are correct that some low pop states will not have enough samples, but I only need a few states to be able to gamble.

You bet Rubio to win?

this chart has since been updated btw, based on demographics and other factors.

http://fivethirtyeight.com/features/...-these-states/

No, I don't think he will win. But I would like to know how hard I should pound the Trump state-by-state lines given national polling.

It seemed like Nate's chart required a lot of work wrt matching each state's demographics with voting patterns. I think you're gonna have to do that work yourself.

X-post. Why would any sane person bet the markets ??

Yes, but he starts with MorningConsult's state-by-state data. I didn't see this on their website. Maybe he has some special arrangement.

Did you see the free money section I posted in the OP ?

Betfair has a lot less fees and a lot higher possible amounts. At many of the books (not sure about Paddy) you'd be lucky to get £100 down on US markets without a line change, and I also really don't want to worry about getting backed off after winning 500 pounds or something ridiculous.

The really good money is also on state lines which don't have a book equivalent at all.

I missed the cash-out fee. I ship $100 to Predictit and bet Trump gets the nom @73c. If I win, then cash, I cash 0.95*(100+100*(0.9*0.23/0.73)) = $121.93. So my odds are -456. Am I doing this right?

If so, this is what we've got...

2016-2-25	Predictit	Paddy Power
Trump	-456	-250
Rubio	+265	+200
Cruz	+1179	+3300
Carson	--	+15000
Kasich	+2759	+3300
Ryan	+2759	+10000
Romney	--	+10000

Yikes.

I'm not sure what this means. Can an account holder have more than one 'contract'?

Err no. Cashing out, the Hillary "do" is -115, the Hillary "don't" is +103. It isn't even an overlay. But even if you do find an arbitrage situation, it's always going to be more +ev to get one side of the action at PP and one at Predictit -vs- getting both sides at Predictit.

And I'll ask again, how much can I even get down on Hillary at those prices? How much volume can I arbitrage anyways? I'm guessing like $20 max... so my free money is like what, 6 cents?

If you're trying to make me feel bad for being an American, it's working. Thanks.

I think your math is wrong. You pay 10% of the gain. So if you buy Trump for 73c and he wins, your gain is 1-.73 = 27c. The fee on this share is 10% or 2.7c. So a $100 bet on Trump returns ~$133.

Yeah, I had a typo. 0.95*(100+100*(0.9*0.27/0.73)) = 126.62. I'm assuming that '5% cash out' applies to both the profit and the original bet. That's a line of -376. PP had the same horse at -250.

Or let's put a dime on a pick'em two horse race. You pick one side, Ill pick the other. I lose and get nothing. You win and cash, and you get $1805. That's a line of -125|-125, yep a 50 cent line, with a vig of 9.75%.

Why would anyone, who isn't an action junky, bet into a 50 cent line?

There are honestly bets there that are > 50c, but like I said, the real money is in all the state lines and there's nowhere else you could book those.

American betting options do really suck, though.

Sure, that makes sense. I'm not really sure what you mean by 'state lines'. POTUS Primaries, or Governors, etc.

Peeps in SoCal have an option most might not know of. LF Caliente in Tijuana does limited US & UK Politics betting. Right now they have...

Trump -281, Rubio +196, Kasich +3234, Cruz +3234, Carson +24500. I have no idea what kinda action they take. IIRC MX has a 2% tax on the profit, which I have already adjusted the above lines for.

State lines as in individual state primary outcomes.