these are 538 numbers regarding probabilities of trump winning. yes, the math works in harris' favour as well, but the existence of a likely bradley effect and a general polling error in favour of harris make the opposite exceedingly more likely and everybody realizes that.

50% - pennsylvania

50% - nevada

50% - michigan

50% - wisconsin

==============

53% - minnesota

53% - new hampshire

54% - virginia

54% - new mexico

54% - maine

=================

56% - colorado

56% - oregon

given the not just possibility or plausibility but actual likelihood of a roughly 5% error in the polling in harris's direction (and trump's favour), the election map actually probably looks like this going into the very end of the election:

there is a very real possibility of a 400+ ec vote massive trump blowout.