Trump cheats at golf - the ONE and ONLY politics thread

I think you could easily argue polls taken over a longer period of time are a more accurate representation of the electorate, and not a hot flash, in-the-moment decision. We are 5+ months away from the election, not 5 minutes.

 

This is the point at which I'd need to see some past examples of which turned out to be more accurate. I'm guessing, though, that the various bounces and such are so influenced by the particular facts of a particular election that it would be difficult to suss out a definite pattern. That said, it is long, long past time that the Bernie Bros knocked it off.

 

Looking at polls has become a pointless exercise. It's pablum for the respective bases of
the candidates and a conversation starter for the cables.

 

In OCTOBER 2012, one MONTH before the election, all major polls following the first debate had Romney leading Obama, because that debate was a thing that happened. The effect wore off pretty quickly. Trump clinching the nomination is a thing that happened, and he is getting its benefits -- six months before the election. Polls can be useful, but their limits are pretty clearcut. People are big liars and people have no idea what they're gonna do next week, let alone next month or year. No sampling method ever devised can account for those facts.

 

Yeah, but you're discounting the horse-race facet of it. Suppose those others weren't divvying up the not-Trump vote and he, therefore, wasn't the top-polling candidate at the outset. Completely different story (probably).

 

I get all my analysis from Peggy Noonan's feelings. And yard signs.

 

Ah, the point at which you tell us that polls never mean anything, because Michigan.

I think in two elections, Nate Silver has hit 99 of 100 states in the general. So he would politely disagree with you.

 

The 99 of 100 thing sounds more impressive than it really is. I suspect had I (or you) spent maybe half an hour on it, 94 or 96 out of 100 would have been a distinct possibility. There simply aren't that many states that are truly in play.

 

Tell that to Peggy Noonan and Karl Rove.

But the point is: No, it's really not that impressive. But Gee keeps beating this drum that polls are meaningless. They aren't. In fact, things usually go relatively predictably.

 

Polls aren't meaningless, but polls in November mean more than ones in May.

 

It assumes that some great majority of folks not voting for Trump were voting against him.

It also assumes that even folks who didn't like Trump would not have chosen him in a two-way race against an unlikable candidate like Cruz, or someone whose last name is Bush.

 

Arguing about which polls are the best is pointless. Incorporate them all.