Things get dicey when the Dem is under 50. I'm sure President Hillary Clinton thanks Michigan for those key victories as shown by the polls.
Polling average of 2016 in Michigan: Clinton +3.6
Polling average of 2018 in Michigan: Whitmer +10.3
Totally the same. Also, this isn't even a tightening since Whitmer was up 5 in EPIC-MRA's last poll.
Well, subtract 4 from the Clinton average and get the actual result...so let's subtract 4 from 10.3...and it gets us 6.3. Still not a double digit victory.
You do realize that just because the polls overestimated Democrats by 4 in 2016, it doesn't mean they will do the same thing this year? Polling errors can happen to both parties. Polling errors are especially prone to happen when the race is lightly polled like Michigan in 2016 was, because absolutely nobody thought Clinton would lose there.
Examples of polls overestimating Republican strength:
The 2012 election in general (not just the presidency, some of the Senate races were off big time including ND-Sen and MO-Sen)
2017 Virginia governor (Northam was up 3.3 in the polling average, won by 9)
2017 Alabama Senate (Moore was ahead 2.2, he lost by 1.5)
Pretty much every competitive Nevada election in the past 15 years
Even the 2010 GOP wave was overestimated. The GOP was up by 9.4 in the GCB, but won the nationwide house vote by 6.8