"Relief pitcher is in favor of a rule that favors relief pitchers"

Shocking.

Yup, keep the inherited runners away from me, I want me some saves!

How about the situation where the starting pitcher gives up a bunch of runs (say 5 in 5 IP) and leaves trailing but the offense comes back to take the lead while the pen throws 2 scoreless innings but then a reliever gives up a run then the offense hits a 2-run jack but then some random AAAA reliever comes in and blows the game, we give out a loss to the starter, two relievers, 5 holds, 2 wins to the offense and 5 game-winning RBI?

I'm with Hoffman here. You want to change the win definition in some meaningful and useful way, that's worth discussing. You want to do it in a manner to "punish" bullpens (who are probably holding more leads for the starter than the starter would be able to do himself) is stupid. Rewarding starters who "pitch well" was why the Devil created the quality start stat.

And you can write it off to relief pitcher selfishness if you want, but something tells me Hoffman's career 54-66 record never earned him one penny and won't play a part in his HoF case. (And no, he wouldn't be the first pitcher with a losing record to make the HoF -- Sutter was 68-71 and Fingers was 114-118. And technically Satchel Paige was 28-31 but I'll guess he was better than 3 games over 500 in the Negro Leagues. :-)

For all of the problem with WPA and LI, I'm looking forward to the day when those stats are prominent enough to motivate decision-making the way the Win and Save stats do.