For 538 watchers

[DemocratSinceBirth]

http://projects.fivethirtyeight.com/2016-election-forecast/updates/



The Selzer IA poll actually increased Hillary's probability of victory.

[vadermike]

1. That's weird ?

Why?

[still_one]

3. The Selzer poll has Hillary still losing Iowa, am I seeing different?

[tarheelsunc]

5. It caused the previous Selzer poll to drop off.

Nate had adjusted the previous Selzer poll to Trump +7, and he adjusted this one to +6 Trump.

[Dem2]

6. I thought that was interesting too

I dunno.

[budkin]

7. Predicting that his final number will be 60-40 Hillary

[regnaD kciN]

9. If he were crafting the results…

…he'd have the final one be as close to a tie as possible while still having HRC ahead -- that way, should Trump pull an upset, he could claim his was the only model that caught that the race was close enough that it could go either way, while, if Clinton wins, hey, he predicted she probably would, right?

[qdouble]

10. I wouldn't be surprised if he put it at 55%. He's definitely trying to set up a scenario

where he's right regardless of what happens. The odds going further and further towards toss up when they are positive for Hillary makes little sense.

[Dem2]

11. Gravis (Breitbart) poll goes from +1 to +2 (or +1 to +3 using Nates "adjustment"

Hillary's odds decrease slightly.

It the inverse poll ration transfer function - yeah, that's it!