In 2016 my electoral map guess was literally a copy of the 2012 one except with like a 6 point HRC popular vote win, I'm not getting burned like that again. Also the dem overconfidence on this forum is painfully obvious, I don't know how one can't see that. As for the tilt, I hate putting things at pure tossups so I use tilt instead mostly, because it usually is not strong enough for lean. I don't get the whole problem with the tilt rating though, I really like it.
Oh I'm aware of the very high confidence users on here are placing in Democrats, I just don't think it's
overconfidence except in some select cases. Considering the average user on Atlas is young, late HS - college aged, that means most people here have no good memory of a Democratic midterm wave, of which the last one was in 2006. For a while I think that tended to skew people's opinions towards Republicans because it was impossible to fathom the idea that Democrats could do well in a midterm. If Democrats really did get a D+8 or D+9 win in the House popular vote, the overconfidence you claim would turn out to be reasonable predictions.
The problem with "tilt" ratings is it gives people too much wiggle room to avoid committing to a real prediction. It's kind of like decimals on poll returns.