We examined pitchers being undervalued last week, so let’s flip the script this time around and look at guys to avoid at their current draft price. One of the ways we look at pitching sleepers is to note their luck on hits, measured by BABIP (batting average on balls in play). But that ignores homers and doesn’t even distinguish doubles and triples from singles.
The questions is whether ISO allowed (isolated slugging, which is slugging average minus batting average) is remotely predictive. Some think it takes so long to stabilize (meaning it’s at least half skill or lack of skill) that it’s worthless.
Yet there are many relievers in history who post similar ISO numbers for their entire careers. The skeptics say that since those numbers are random at that sample size, that should never happen. Mariano Rivera for example had an ISO allowed of .039 to .114 every year. A statistician may call that random, but I call that predictably well above league average at limiting damaging hits every year. Aroldis Chapman has had an ISO below .100 for four straight years and only one over .100 (.112 in 2013). Do we have to pretend that Chapman’s ability to limit damaging hits is not real because it has never in any one of his seasons reached stabilization?
But let’s set the debate aside now and look at the pitchers who had a problem last year in hit quality as opposed to hit frequency, something few factor into their valuation. Thus I will fade all at their average draft position (ADP).
Madison Bumgarner, ADP 27.2: (ISO allowed rank 98th out of 134): He’s not going at any discount this year despite a much poorer performance in 2017. This is despite a shoulder injury that if it was caused pitching instead of dare-deviling would have us gravely concerned. Let’s cross reference ISO assuming ISO is fluky; Bumgarner’s well-hit rate of all at bats via Inside Edge was .186 compared with the MLB average of .155. So Bumgarner got pounded last year, case closed. I don’t get everyone giving the all-clear sign on him in 2018. He’s way more of a gamble than he was at this time last year. His ISO allowed in the second half was .211 (awful) vs. .122 (better than average) in the first half. Was he healthy? Is he healthy now?
Yu Darvish, ADP 51.4 (71st): It wasn’t just the postseason: He was hit hard in the regular season, too. Why? Was he tipping his pitches? His well-hit average in the regular season was .162, about a C-minus. We’re supposed to like Darvish because his strand rate will revert to 80% from 73% but ignore the fact that runners scored at an above-average rate because he was giving up above-average hit quality. Hmm…. This makes zero sense. So your case for Darvish is really, “I think Darvish was unluckily hit hard all of last season and will revert to 2016 when his rate was just .124.” Okay. That is a gamble and I can see it paying off but you’re paying to gamble on Darvish at his ADP. You’re supposed to get paid to gamble (meaning pay a much lower ADP than the player’s plausible upside).
Jake Arrieta, ADP 105 (78th): His well-hit allowed was .170. More badness. In his 2015 heyday, it was .196. Remember this is of all at bats so Ks count as “not well hit.” I can’t sign off on betting that Arrieta will be anything more than a fifth fantasy starter and he’s going to cost you well more than that. I’ll shock you and say I’ll take Sean Newcomb way later than Arrieta and feel I have a similar guy. I like Newcomb and you may like someone else but the point is that Arrieta is now in that very deep third tier where you have to reach to draft him instead of waiting for one of those guys to fall much later.
I find it very interesting how well-hit is just a proxy seemingly for ISO allowed. Looking just at well-hit average, some more names pop up.
Johnny Cueto, ADP 131.8. The original poster boy for my use of ISO allowed, was crushed last year, ranking 30th worst in well-hit average (.191) or even worse than Bumgarner (36th). Can you tell yourself a story and hope Cueto is an ace? Sure. Maybe it was all the blisters. But I see longer odds that this ends happily ever after than most.
Chris Archer, ADP 66.6: He underperforms his peripherals and the answer is obvious: he’s a two-pitch pitcher and if the slider isn’t working he’s going to get raked. Archer was 35th worst in well-hit (.186) and his ISO allowed the past two years is .163 and .165, which sure seems non-random. Yes, Archer was great at ISO allowed (and well-hit allowed) in 2014 and 2015 — but if you want to bet on 2015 stats, just go make Matt Harvey your ace in the reserve rounds. The point is that 2015 is a long time ago and things for pitchers change very, very fast. So maybe what we think is random variance isn’t so much the stat changing radically but the pitcher changing. Maybe Archer’s stuff has lost that extra something special or hitters have just figured him out. Either way, I’m not paying ADP to find out.