Originally Posted by Homer Bailey
We clearly will have to agree to disagree, but the bolded part of what you said is just absolutely not true. Back in the day, teams fell victim time and time again to ERA's that were bloated due to random luck over a certain period of time. FIP and xFIP didn't just magically become better predictors of future success, or better indicators of true talent in recent years. Just because no one knew about it then, didn't mean that pitchers that struck out more, walked less, and gave up fewer homer weren't better pitchers than those that relied on excellent defense or luck to get them their outs.
BAPIP doesn't apply to relievers the same way as it does to starters. Relievers have been shown to be able to maintain low BABIP over many seasons, even whole careers.
Relievers and starters should not be treated the same way via stats. Even though they are both pitchers, what they do is very different, and nets different results statistically.
Relievers face only a handful of batters a game and can throw as hard as they want the entire time. Starters have to face the same lineup multiple times each game, and must pace themselves in terms of velocity and pitch selection. Relievers also usually come in during high leverage situations, causing them to pitch differently than when starting an inning clean and/or a big lead.
All of this and more leads to different stats and changes the importance of each stat for reliever and starters. You are using starter stats to talk about Ondusek, and they aren't as relevant for him since he's a reliever.