[Artemis Webb]

I would be careful of criticizing this. FReepers often complained about “weighting” in the last two Presidential elections yet the weighted polls proved correct.

[Voluntaryist]

While that’s true, I don’t think Romney was really ever up in any poll at this point in time.

[GilGil]

See post #23

[GilGil]

And post 25

[Sans-Culotte]

I would be careful of criticizing this. FReepers often complained about “weighting” in the last two Presidential elections yet the weighted polls proved correct.

It is normal to weight polls. However, it is not proper to over-weight polls. It is like handicapping in golf. You would not give a bunch of free handicap strokes to a guy who is a scratch golfer. I would add weight to the dem side in a poll, but not 14% more.

[Yashcheritsiy]

I would be careful of criticizing this. FReepers often complained about “weighting” in the last two Presidential elections yet the weighted polls proved correct.

Yes and no.

The professional houses that relied on realistic partisan weights reflective of actual partisan alignment, and which generally predicted a 2-5% Obama victory, were generally very close to the final result.

The MSM polls that relied on heavy Dem overweighing and predicted Obama victories in the 8-10-15% range were way off. They were only "right" in the sense that they called an Obama victory. Yet, they were often further away from the actual final result (Obama +4.1%) than were the FReepers who were "revising" their numbers using more accurate partisan weights.