OSC House Model

[voice_of_resistance]

Check out this new House model from OSC Elections, a bipartisan group of Yale students who have strong stats backgrounds.

https://oscelections.shinyapps.io/OSCElections/

Currently, projected for D+10 GCB with D+7 seat gain. TX-32 is currently the tipping point although I'm sure it could change.

[voice_of_resistance]

Must be using Garcia's overperformance relative to demographics and trends to inflate his vote share. Overall, I think this is a good model though.

[voice_of_resistance]

A bit skeptical of IN-5 being Dem favored,  maybe a tossup district sure, but 56% chance to flip seems really generous.

In general the percentages see to be oriented in a very steep bell curve. Moving 1% from one candidate to another shifts the percentages drastically.

Also, yeah, TX25 is the statistical error, since it looks like they are using the previous election as a baseline, and the previous election is the weird special one.

How is TX25 an error? Most ratings places seem to put it at Likely R.