I am addressing it, mhaze.

Past directional error can justify predictive caution - but only if the cause of that error is stable.

If CNN's underestimation mirrored the broader industry pattern, that suggests a shared modelling issue rather than something specific to CNN.

Without isolating CNN's performance relative to the field, the projection from "they did it before" to "they are likely doing it again" remains an assumption about cause.

I'm going with lack of cajones.

"Without isolating CNN's performance relative to the field, the projection from "they did it before" to "they are likely doing it again" remains an assumption about cause."

Lots of words looking for a coherent thought.

It's simple, mhaze:

- If everyone made the same error, that suggests a shared modelling issue.
- If CNN made a materially larger error than others, that suggests something specific to CNN.

Without that comparison, you can't distinguish between the two.

Just making the same erroneous claim over and over doesn't improve it.

CNN is being bought by Zionist Larry Ellis.

It isn't erroneous, mhaze.

Distinguishing shared industry error from outlet-specific deviation requires comparison.

Without that comparison, you can't determine whether the pattern reflects CNN specifically or the broader polling environment.