Reuters/Ipsos: (Most) States

[mencken]

I thought Reuters was a good pollster...?

By definition 2 out of every 50 polls is going to have a result outside of the margin of error of the actual result. If we assume a sample size of ~350 (which seems typical for a medium-sided state in these polls), then a fair amount of these polls should be 5% or further from the actual result, with smaller states even less precise.

[mencken]

On FiveThirtyEight, an Ipsos poll has Trump leading here. Is this legit or no?

I wouldn't dismiss the possibility of Trump having a lead there or drawing things close in Colorado as nonsense. The Ipsos polls of the state actually have a statistically significant sample and a two-week frame of reference.

Given the state of the national race, it could be closer than some of the older data suggests.

I am not saying Trump is ahead in the state, but it could be closer than you think. There hasn't been a legit landline/cell poll of the state in a while.

Quinnipiac and Marist had Clinton up at 8 and 12 respectively back in early August. Since then Trump has picked up ~4 points nationally, so a modest single digit lead is about what should be expected here.

[mencken]

This also has WV as Trump +10, for what it's worth.

Probably best to treat with significant skepticism subsamples with lower than traditional sample sizes (I wish people would apply this logic to demographic crosstabs...)

[mencken]

FL Clinton+4 here too. What are the chances of so many polls of different qualities showing Clinton+4 in FL this many times? Has the race stabilized/locked in that much there?

Cough cough herding