Let’s start this week with the poster boy for Statcast skepticism, James Paxton. The data says he’s bad. But he keeps posting good results. So if you want to say he’s proof we should ignore Statcast, you have an argument. Butyou’re also then a nihilist(language warning for that link). You believe in nothing except the actual results, which have zero utility value/edge.
Advertisem*nt
Look at the differencehere in his pitching value (actual) and his pitching metrics(mostly expected).Evel Knievel couldn’t jump that chasm. I’m not going to try. If you want to increase Paxton’s rostership from his current 44% on Yahoo because you believe what he’s doing is sustainable, then you’re just not a persuadable for me. If your league is big brained, could there be an edge in actually buying the players doing well for real rather than hypothetically? Maybe. But most leagues have three or four managers — max — who think like we advise in this space.
It goes without saying that Paxton is going to be on my not-good list. We’ll start there. I am not maintaining that expected stats are predictive. I’m saying they’re descriptive of how the guy is actually pitching. So if you wouldn’t roster a pitcher right now who’s 35 with a low K rate and an ERA over 5.00, then don’t roster Paxton. Just give the expected stats the weight of the actual ones. There is edge here because you’re playing a different game than most of your league.
Who else should you dismiss despite valuable stats so far? And by dismiss I mean try to trade. There’s no point in just cutting a guy with great surface stats. (All data is as of Monday. The full spreadsheet is here.)
The pitchers to avoid
Tyler Anderson’s expected ERA is 4.83, or 2.11 better than actual. He’s 33% rostered.
Logan Webb clocks in at 4.86, not a surprise given his WHIP is over 1.30. And he’s 98% rostered, so he’s extremely tradable with an actual ERA of 3.03.
Jose Berrios is 4.61 in expected ERA, but his actual ERA of 2.82 has him 93% rostered. I do not believe.
Carlos Rodón is viewed as a Comeback Kid but, really? His expected ERA is 4.89, ninth worstamong pitchers with at least 150 balls in play. So see if a leaguemate buys his actual ERA (3.27). I stipulate there’s a chance Rodón will pitch better given his injury woes of last year, which may still be making him a little rusty. But why wait for that when you can sell him at that price right now?
Advertisem*nt
The pitchers to target
More important for us are the pitchers who are much better than they appear looking at their scoreboard stats. These types tend to be more readily available, via trade and even on waivers.
Brandon Pfaadt has an elite expected ERA of 3.07, more than a run better than actual. But he’s 66% rostered. Could you trade Rodón and a weaker hitter for Pfaadt and a stronger one and win both sides of the deal? Absolutely. I’ll dare say it’s even borderline likely.
Nestor Cortés is so ugly to roster given he is gimmicky with his motion and barely can break a pane of glass with his fastball. But his expected ERA is second lowest. Yes, he’s 82% rostered. But none of his managers love him as much as his expected ERA, I guarantee it. He also doesn’t hurt you much in Ks.
Here’s a guy to pick up off waivers: Patrick Sandoval is just 8% rostered. His expected ERA is 3.62, which is playable in all formats. Some, like Aaron Nola, have a certain something, or more precisely lack it, that keeps them underachieving their peripherals. Maybe Sandoval just can’t leverage his above average raw ability for some reason. But when I can’t think of a reason why that would be true, I keep thinking it’s random and expect it to change.
Similarly, Reid Detmers is just 35% rostered. He’s substantively performing almost exactly like Yoshi Yamamoto. I’m saying you can trade Yamamoto and a borderline hitter for Detmers and a top 20 hitter. Not only can you, but you should (if you need hitting). That’s how you win leagues.
Ronel Blanco is good and I personally believe that it’sarbitrary to be suspendedfor sticky fingers (best Rolling Stones album and probably the best album ever, btw). What is the standard being applied by umps? There’s no science involved. Blanco is being dropped by some moralists and I believe he’ll pitch as well as he has going forward. Don’t drop him.
Advertisem*nt
Michael Wacha is being added like crazy but his roster rate still is below 40% now. His 3.67 expected ERA says he’s still a borderline mixed-league starter. So it’s not just the last two starts but the whole year where he’s pitched well enough to be on the back end of your rotation for the price of a waiver pick up.
(Top photo: Matt Krohn/Getty Images; Chris Coduto/Getty Images; pic of Paxton: Matt Krohn/Getty Images)
Michael Salfino writes about fantasy sports for The Athletic. His numbers-driven fantasy analysis began with a nationally syndicated newspaper column in 2004. He now covers a variety of sports for FiveThirtyEight and The Wall Street Journal, for whom he also writes about movies. Michael helped Cade Massey of the Wharton School of Business originate an NFL prediction model https://massey-peabody.com that understands context and chance and avoids the trap of overconfidence. He strives to do the same when projecting player performance. Follow Michael on Twitter @MichaelSalfino
