"The Senate is a like form of historical gerrymander, forcing certain votes to count far less in the results. We can't do much about a 240 year old rule, but we should understand how a national popular vote win of 12.4% on Senate races results in the LOSS of three seats for Dems."
Is he really that stupid? The last time Class 1 Senators were elected the result was D 25, R 8, on a 11.6% D popular vote lead, in 2012. This time, the result is looking like being 22-11 on a 12.4% lead, which is less disproportionate in favour of the Ds, but still disproportionate.
And he might like to compare the California Senate race between 2012 and 2018:
2012 California
D 7.86 million
R 4.71 million
2018 California
D1 3.3 million
D2 2.8 million
So ~4.7 million Republicans were not able to vote for a Republican.
If you look at the big states (so the popular vote):
CA: 2012: D+25 2018: disenfranchised
TX: 2012: R+16 2018 R+3 (swing to D 1.033m votes)
FL: 2012 D+13 2018 R+0 (swing to R 1.031m votes)
NY: 2012 D+46, 2018 D+33 (swing to R 1.22 m votes)
PA: 2012 D+9 2018 D+13 (swing to D 0.10m votes)
OH: 2012 D+6 2018 D+6
NJ: 2012 D+20 2018 D+9 (swing to R 0.397m votes)
In terms of the popular vote the Republicans actually did considerably better than last time. Hence winning more seats
.