RADIO
![By Erik Drost on Flickr [CC BY 2.0 (http://creativecommons.org/licenses/by/2.0)], via Wikimedia Commons](https://cdn.rotoballer.com/2014/12/MLB-Fantasy-Baseball-Advice-Analysis-Danny-Salazar-e1467171723990.png)
As Opening Day rapidly approaches, we close the preseason version of Champ or Chump by looking at a trio of pitchers. Pitchers have a unique set of data points to consider. Stats like BABIP and SwStr% continue to matter, but we want our pitchers to have low BABIPs and high whiff rates whereas with hitters it's the opposite. Pitchers have their own luck-driven stat in strand rate (LOB%), and ERA estimators such as FIP that have proven more predictive of future performance than ERA alone.
PITCHf/x is a powerful tool when evaluating pitchers, and I plan to use it frequently both below and once the season starts. If you're unfamiliar with its basic applications, I provide a rundown here. Without further ado, let's look at some arms.
Editor's note: Be sure to also check out our 2016 fantasy baseball rankings dashboard. Our rankings assistant tool combines all our staff's ranks in one place. You can easily filter and export all sorts of rankings and tiers - mixed leagues, points leagues, AL/NL only, top prospects, dynasty ranks, keeper values, and more.
Starting Pitcher Champs and Chumps
Johnny Cueto (SP, SF, ADP: 72)
Cueto had a bizarre 2015, pitching like a true ace in the first half (2.73 ERA) only to implode with the Royals in the second half (4.34) and postseason. For fantasy owners, it all added up to a 3.44 ERA, supported by a 3.53 FIP, and a record of 11-13. His strikeout rate was essentially league average at 20.3%, leaving owners that paid for an ace unsatisfied.
Their mistake was in paying Cueto as an ace. Since his breakout in 2011, Cueto has consistently over performed FIP. This fact is clearly demonstrated by the table below.
Year ERA FIP xFIP
2011 2.31 3.45 3.90
2012 2.78 3.27 3.60
2013 2.82 3.81 3.23
2014 2.25 3.30 3.21
Even allowing for the fact that 2013's sample is only 60.2 IP, a clear trend emerges. Sky high strand rates and microscopic BABIPs can only continue for so long, however. FIP catches up to most FIP beaters eventually, and it looks like eventually has arrived for Johnny Cueto. His ERA last season fits perfectly in the FIP or xFIP columns, making it a believable result despite never happening before.
Cueto was never a fantasy ace, even when defying FIP. FIP is rooted in strikeouts, walks and homers allowed, the things pitchers have the most direct control over. The first of those is a category in most roto leagues, and worth a substantial number of points in other formats. The fact that FIP never liked Cueto tells us that he never really posted the elite strikeout rates required of fantasy aces.
By PITCHf/x, we see that Cueto has one good strikeout pitch - a changeup with a 20.7% SwStr% and 47.7% O-Swing%. No other offering stands out, however, leaving Cueto with only one wipeout option. True aces, like Clayton Kershaw, Carlos Carrasco and Max Scherzer, rack up Ks by fooling hitters with a variety of deadly offerings. The change gives Cueto a strikeout floor, but that just makes him rosterable.
Many will point to Cueto's strong 25.2% K% in 2014 as evidence of the fact that the Ks will come back, rendering last year as nothing more than a blip. In reality, 2014 was the blip. Cueto's career K% is just 19.9%, below the MLB average. No other Cueto season, besides 2014, has produced especially high strikeout totals. He is just not a K guy, which hurts his value even if he returns to his FIP beating ways.
Why are fantasy owners drafting a rosterable arm as if he is a borderline #1 starter? Sure he's back in the National League, but it is hard to find a better supporting cast than the eventual world champs and Kauffman Stadium is great for pitching. AT&T Park also has that reputation, but it won't make him into a strikeout machine.
Obviously, the Giants disagree with my conclusion based on all of the money they gave Cueto. They made a similar call with Matt Cain, another "proven" FIP beater. How has that worked out?
Verdict: Chump
Danny Salazar (SP, CLE, ADP: 82.2)
Salazar is Cueto's polar opposite in many ways. While Cueto historically pairs great ERAs with uninspiring strikeout totals, Salazar posted a 3.45 ERA despite a 25.3% K% in 2015. The dominance makes it easy to like Salazar, but he still managed only a 3.62 FIP last season. Why?
Somehow, he was lucky on batted balls despite Cleveland's porous defense. Salazar's BABIP against was just .278 last year. The best explanation for this is his LD%, which shrank to 18.7% last season after landing over the MLB average in his previous major league experience. Some pitchers can sustain low LD% rates, but Salazar needs to do it a lot more to prove he's one of them. Until then, we have to assume that he allows more hits this year, bringing his ERA closer to his FIP.
More problematically, his strikeouts are trending in the wrong direction. Salazar changed up his pitch selection to include more 2-seamers (6.7% in 2014 to 18.1% last year) and changeups (11.5% to 18.3%) at the expense of fastballs (69.3% to 51.3%) and sliders (12% to 8.5%). The 2-seamers offer nothing in terms of SwStr% (5.1%), while the heater offers a respectable 8.4% figure. He appears to be trading dominance for grounders, a trade most fantasy owners shouldn't be interested in.
His slider is even more concerning. Serving as the secondary strikeout weapon that Cueto lacks, the pitch posted an above average 38.3% O-Swing% in 2014. Last season, it remained effective by SwStr% (16.5%) but was chased much less often (32.1%). With a zone% of just 36.5%, hitters need to chase the slider so it is not an automatic ball. If hitters refuse to chase it, Salazar won't bother throwing it - potentially the rationale for the decreased usage above.
While Salazar's changeup is much better than Cueto's (26.8% SwStr%, 53.2% O-Swing%), nothing else in his repertoire seems like the complement necessary for elite K totals. We may have gotten a sneak peak at what a Salazar without a slider would look like in the second half last season, as his K% dropped from 28.6% in the first half to 22.5% in the second. His BB% also increased from 6.4% to 7.7%. Could this have been the result of opposing hitters learning to spit on the slider?
Salazar has great potential, but is no lock to figure everything out in 2016. His current ADP does not adequately price in his risk.
Verdict: Chump
James Shields (SP, SD, ADP: 144.6)
If anyone had a stranger season than Cueto, it was Shields. His K% exploded from 19.2% to 25.1%, his BB% followed suit by jumping to 9.4% from 4.7%, and his ERA ballooned to 3.91 thanks to a ridiculously high 17.6% HR/FB. It added up to a 13-7 record and confused advanced metrics, as FIP (4.45) thought he should have been worse while xFIP (3.70) thought he pitched better.
Shields's career HR/FB is 11.7%, and he posted a 9.7% mark in 2014. His HR/FB was inflated on all of his offerings, not just one, so it wouldn't seem to be a pitch selection issue. His batted ball profile was also unchanged despite all of the dingers. Therefore, I'm inclined to side with xFIP in regressing his HR/FB and saying that Shields will post a better ERA this year.
The explosion in the other two true outcomes is fascinating, especially since Shields lost 1.4 mph on his fastball. Decreased velocity is not usually the catalyst for an elite strikeout rate, though his changeup lost 0.9 mph to maintain most of the differential. It's possible that Shields knew his fastball was going, and spent some time figuring out how to compensate.
Shields had two great K pitches last season, a change and knucklecurve. Eerily, both offerings had exactly the same SwStr% (19.5%) with an O-Swing% a hair higher than 43%. The change has been progressing to wipeout status for three years now, posting a zone% of 40% in 2013, 34.2% in 2014 and 27.3% last year. This seems to suggest that Shields is increasingly looking at it as a put away pitch, rather than a change of pace in the zone.
The knucklecurve did not enter Shields's repertoire until 2014, and almost doubled in usage last year (11.3% in 2014 to 18.8% last season). This gave Shields the second weapon necessary to post elite K numbers, while prominently featuring two pitches with zone percentages around 30% helps to explain some of the walks.
Not all of them, however. Shields threw his 2-seamer more often last season as well (12% last year to 16.6%) but less often for strikes (42.6% zone% last year vs. 47.4% career). It was the only pitch in his repertoire to post a high GB% (59.4%), so it seems likely that Shields threw it more often and lower in the zone in an attempt to stop the ball from flying out of the park. It didn't work, but it seems logical that Shields won't have the same incentive to bury it with neutral HR/FB luck.
Shields also found the zone less often with his cutter last season (49.3% vs. 52% career), but the resulting uptick in SwStr% (9.5% in 2014 to 11.4% last year) may make the change worth keeping. It says here that Shields can maintain enough of his K% to strike out 200 batters next season while bringing his walk rate down to 7% or so.
That many Ks at his current draft slot represent a tremendous value, and neutral HR/FB luck should improve his ERA as well. He throws enough innings to collect his share of wins as well, even on a team that isn't very good. He might walk a few more guys, but 2014's 4.7% BB% was an outlier anyway (6.1% BB% career). Overall, he's a fine selection at his current price point.
Verdict: Champ
MLB & Fantasy Baseball Chat Room
[iflychat_embed id="c-12" hide_user_list="yes" hide_popup_chat="no" height="400px"]
