To me, it's more of a predictor of how well a pitcher should do based historical data on stats that normally lead to good performance.
It's making it more complicated than it has to be.
Great pitching is not allowing men to get on base, and when they do, to limit the amount of XBHs (SLG%). One can argue that keeping runners from scoring (having a higher LOB%) shows that a pitcher buckles down when he needs to, but the end result is allowing less runs. Now, playing in different parks, and I'm not just talking home field, and facing different opponents has to be factored in, as it is in ERA-.
I'm not saying ERA- is "the stat", but when you couple it with WHIP (maybe we need a WHIP-) and IP ("Bulk"), then I think you get a true sense of who did the best. Had Porcello not let up the HR and maybe lasted another inning to get the win, I think he'd be the slightly better choice. Now, it's a toss up.
ERA-
71 Porcello
73 Kluber
75 Sale
ERA+
149 Kluber
145 Porcello
125 Sale
WHIP
1.01 Porcello
1.03 Sale
1.06 Kluber
OPS Against (OBP/SLG) note: OBP>SLG
.631 Kluber (.274/.357)
.635 Porcello (.268/.367)
.642 Sale (.278/.364)
GS%
79% Porcello
74% Sale
69% Kluber
GS/IP/IP per GS
33/223.0/ 6.76 Porcello
32/215.0/ 6.72 Kluber
31/221.2/7.15 Sale
WAR
5.3 Sale
5.2 Porcello
5.1 Kluber
xFIP
3.50 Kluber
3.59 Sale
3.90 Porcello
Can't get much closer than this.
Perhaps the win-loses will tip the balance, wrongly or rightly, to Porcello.