Predict the US midterm results.

[Ming]

Just voted
Dems in the House 215
Dems in the Senate 48

I think the Dems will be the first party ever to gain seats in the house but lose in the Senate at a mid term when not in power

[Lorizael]

I think the Dems will be the first party ever to gain seats in the house but lose in the Senate at a mid term when not in power

Nope!

[Ming]

I'll have to go back and look at the story I read this morning to see how the "phrased it"
But, I will admit I'm wrong on that claim
Good Source

[Broken_Erika]

The Republicans baby. The republicans.

The future supreme overlords of this country. Armed troops on every streetcorner. Mandatory curfews. No lawyers, life in prison for petty theft.
*sigh*

I can't wait.



Sent from my HTC Desire 626 using Tapatalk

Chris Tucker 2020
Make America Super-Green Again!

[Ben Kenobi]

Crazy for putting money down on Heitcamp if you expect her to win.

[Ben Kenobi]

As for the Trump election, funny I seem to recall Nate Silver also predicting Hillary would win. Nobody had Trump winning, and yet everyone is backpeddling now and say, "yeah it was a real possibility" LOL. Easy to CYA, but his call was Hillary and so were almost everyone's picks here.

[Ben Kenobi]

For this one. I have Rs at 58, Ds at 42 in the Senate. Gain of 4 in the House for Republicans.

538 was 5 points off last election. This flips WI and the MN special. I have WV for Manchin, but that could flip too. All three are really close. OH with Brown could also flip, alongside MI with Jones, but I think Brown and Stabenow hold on, as Democrat incumbants in purple states.

Rs win ND, TX, TN, MO, IN, MT, FL, AZ, WI and the MN special. Lose in WV, OH, and MI, but close. NJ won't be close.

R 56 D 44 is pretty much just as likely as R 58, WI and MN are quite close.

[Ben Kenobi]

Christ. 20 percent Democrat chances to win the Senate? It's more like 1 percent. Anyone here want to take that bet up with me? Seriously, 538 is taking advantage of people's lack of numeracy.

[spambot]

As for the Trump election, funny I seem to recall Nate Silver also predicting Hillary would win. Nobody had Trump winning, and yet everyone is backpeddling now and say, "yeah it was a real possibility" LOL. Easy to CYA, but his call was Hillary and so were almost everyone's picks here.

Do you understand how probability works?

[spambot]

Christ. 20 percent Democrat chances to win the Senate? It's more like 1 percent. Anyone here want to take that bet up with me? Seriously, 538 is taking advantage of people's lack of numeracy.

Why would someone bet on the democrats taking the Senate if they believe it's somewhere in the ballpark of 1 in 5. I'd be happy to bet against your claim that the Republicans will gain seats in the House.

[Ben Kenobi]

Do you understand how probability works?

Yep, sure do.

ND is an R Lock, at around R+10 so maybe 2,3 percent by itself chance of a Democrat hold. Let's be generous and say 5.

TX is an R Likely at around R+8, so maybe 20 percent chance of flipping democrat

TN is the same, around R+8 maybe 25 to account for it being less polar than TX.

That's already 0.25 * 0.2 * 0.05 to flip all three.

We have WV, IN, AZ, MO, MT, FL, which is six tossups. Ds would need to win all six. Assuming they are true tossups that's 0.5^6 = 0.015 or 1.5%.

0.015 * 0.2 * 0.25 * 0.05 = 0.000039 or 0.0039 percent odds of a Democrat win in the Senate.

at 5:1 your odds should be 25600:1.

Your odds are about 5k:1 of making back your money. I feel confident betting my house against 5:1 odds on the Democrats winning the Senate.

[Ben Kenobi]

If OH is at 35 percent, MI at about 30, WI at about 40 and MN special at about 40, that's

0.65 * 0.7 * 0.6 * 0.6 = 0.16 or a 16 percent chance of holding all 4.

My averages work out to around 50 percent odds of R+5 based on these numbers. R+2 is around 20 percent.

[Ben Kenobi]

I'd be happy to bet against your claim that the Republicans will gain seats in the House.

Just looking at the tossups it's about 40 percent chance of retention for the Republicans, assuming that RCP is right about which races are tossup. R+4 is probably around 10 percent or so, assuming that RCP is correct about the races that it is calling. If there's systematic bias of about 5 percent in their numbers (as there was in 2016), then we'd be looking at a congressional ballot of about +1 Democrat which means that Rs would be odds on to gain a few.

[spambot]

You're assuming all of those probabilities are completely independent? Now that's just silly.

[Ben Kenobi]

You're assuming all of those probabilities are completely independent? Now that's just silly.

If I were to quote the predictit odds on ND alone it would be around 5 percent, and that's just for one state. All that has to happen is that one of those states flips the other way and you lose. 20 percent? My ass. Again, he's taking advantage of people not understanding how probabilities work.