How accurate were the state level LV polling models in 2008 and 2012 vs the RV models?

[uponit7771]

Remember this is for state level polling... Is there some kind of accuracy metric?

Id like some historic context

For instance the Bloomberg poll uses the 2004 LV model for Ohio when it came to black voters.. That seems stupid relative to what happened in 2004 with minority polling places.... The voting places state wide were a disgrace in Ohio... So why use that model?!?!

It also was 12 yrs ago... Seems overtly flawed

Either way...

I'm thinking there's some historical context here to explain the LV vs RV descrepency in so many polls.

Thx in advance

[democrattotheend]

2. 12 years ago was the last time we had an all-white ticket

They probably figure 2008 and 2012 provide too favorable of a model for Democrats, both in terms of turnout and percentages. Blacks came out in unprecedented levels for Obama, and he also got a higher share of the black vote than Democrats historically have (93-95% versus about 88-90% typically).

2004 may be too conservative, but using 2008 or 2012 would probably be overly optimistic.

[uponit7771]

3. I pray blacks come out higher than 93℅ so the posters will stop thier stupid low balling

[democrattotheend]

4. That would be great

But 93% was unusually high. I think the 95% he got in 2008 was a record. There were some black Republicans who voted for Obama but didn't become Democrats permanently.