Two takeaways. 1) Small sample size and few polls means House polls can be off by an average of 8 points either way. 2) Dems in early voting are overrepresented compared to polling models, which suggests a systematic polling error on our favor.
https://www.nytimes.com/2018/11/05/upshot/two-vastly-different-election-outcomes-that-hinge-on-a-few-dozen-close-contests.html#click=https://t.co/2oWXClIij5
