That does not prove randomness.
Even if you go by each head-to-head series, here's what it looks like:
Nat> Mil check
Rays> Oak (1 win difference, but Rays had higher total WAR)
STL>Atl (6 wins less but had a better 2nd half record 47-27 to 43-28)
WAS> LAD (13 wins less but was one of the best 2nd half team)
NYY> MN check
HOU> TBR check
WAS>STL check
HOU>NYY check
So far, 5 faves have won, and 3 did not, but one could call the TB-Oak play-in game as an even game.
If you don't count the one game play-in game, which is certainly more "random" than a 5 or 7 game series, the record is Faves 4- Underdogs 2.
It's still not random.
2nd half records:
50-22 HOU
47-27 STL
46-26 NYM did not make playoffs
46-24 LAD
46-27 WAS
46-24 OAK
45-28 MN
44-27 TB
43-21 CLE no playoffs
43-28 ATL
42-29 MIL
I'm not trying to move the goal posts with this second half stuff. I mentioned season W-L records may not be the best way to determine who were the best teams each year and then show if the best teams win, but to me it it pretty obvious the better teams win more often.
I still have yet to see any clear evidence to show it's 50-50. I'm open to change my mind, but give me something. Overall WAR totals and playoff records? Second half records vs PO wins. Last 30 game records vs PO wins.
Does anyone really believe the better team doesn't have a better chance at winning a 5 or 7 game series? You gotta show me more than 2 or 3 upsets out of 8 match-ups this year. If anything that should be used as evidence it is NOT random.