Mediating projections, Part I: Jerad Eickhoff

Baseball projections are a tricky beast. They seem to have improved in recent years, as more data have become publicly available and the algorithms have advanced, but they still have a long way to go. With so many Michael Brantley-esque players who explode out of nowhere, along with the Pablo Sandoval-types who collapse without warning, the world of baseball repeatedly spits in their face. For this reason, we should regard projections with caution, not unwavering credulity. [whistles nonchalantly]

One of the interesting things about projections is their variety. Steamer, ZiPS, Oliver, PECOTA, Marcel — all of them can present unique appraisals of a player. In the cases where they disagree significantly, we encounter an intriguing thought exercise. Last offseason, I looked into a few of these cases, in a popular noteworthy recurring series; today, it starts anew, with Phillies starting pitcher Jerad Eickhoff.

Back in July, Philadelphia dealt an ace pitcher to an aggressive Texas team, bringing back a tantalizing pile of prospects. (A few days ago, it would seem that history repeated itself.) Eickhoff, one of the players who came to Philadelphia in the Cole Hamels trade, dominated in a brief showing at the major-league level later in the year, with a 2.65 ERA and 3.25 FIP over 51.0 innings. Such play from a 25-year-old rookie obviously warrants some skepticism, but how much? Steamer and ZiPS can't come to a conclusion:

System	IP	K%	BB%	HR%	BABIP	LOB%	ERA	FIP	fWAR	fWAR/200
Steamer	155.0	20.7%	7.7%	3.0%	.286	73.0%	4.00	4.16	1.6	2.1
ZiPS	149.0	20.2%	7.8%	3.7%	.293	70.8%*	4.77	4.65	0.3	0.4

*ZiPS doesn't project strand rate directly; I created this from the numbers it does project, using Eickhoff's minor-league hit-by-pitch rate.

In terms of strikeouts (slightly), home runs, batting average on balls in play, and strand rate, the projections diverge in their opinion of Eickhoff. The fWAR disparity makes the significance of those disagreements clear — Steamer foresees a solidly average campaign, while ZiPS thinks he'll become a replacement-level sophomore. What should we expect? Let's delve into each metric and reach a verdict.

Strikeouts

As a farmhand, Eickhoff ran mediocre strikeout numbers, fanning 19.7 percent of the 2,406 men to step in against him. He jacked that up to 24.1 percent upon arriving at the show, and Steamer and ZiPS don't see eye to eye on the extent of his regression. How much of that clout will Eickhoff manage to retain?

So long as his repertoire doesn't go away, I'd expect him to hang onto the punchouts. Eickhoff has five pitches — a four-seam fastball, a sinker, a curveball, a slider, and a changeup, all of which he threw at least 6 percent of the time in 2015 — and although none of them possess notable velocity, their movement sets them apart:

Eickhoff's four-seam fastball, sinker, changeup, and curveball each had significantly more movement — in both directions — than average offerings of their type. That arsenal unsurprisingly succeeded across the board:

Pitch	Eickhoff Whiff%	Standard Whiff%
Fourseam	6.0%	5.6%
Sinker	1.3%	4.5%
Change	13.7%	11.9%
Slider	25.6%	13.0%
Curve	16.9%	9.5%

Standard whiff rates via Jeff Zimmerman and Eno Sarris's research.

In terms of whiffs, every pitch except the sinker held its own. This gave Eickhoff an overall swinging-strike rate of 10.4 percent, which played a large role in his ability to set batters down on strikes.

Some of these whiffs were clear flukes. A slider with mediocre velocity and subpar bite shouldn't fool batters a quarter of the time; that will almost certainly plummet in the years to come. With that said, the curveball certainly earned its production, and Eickhoff will spend his winter perfecting the changeup. This kid's arm has the potential to make hitters look silly for a long time, so I'll take Steamer here.

Home Runs

The strikeout projection gap, however, pales in comparison to the home run difference. After a minor-league career in which 2.9 percent of his opponents hit the ball out of the yard, Eickhoff held hitters to a 2.5 percent long ball rate in 2015. Steamer predicts he'll pitch more to his minor league level of play, and ZiPS feels he'll fall back even further. To what extent will he Harang-ify?

Well, a falling strikeout rate will translate to more balls in play, meaning more changes for a round-tripper. More importantly, Eickhoff has always been a fly ball pitcher: According to Minor League Central, he owned a 38.3 percent ground ball rate down on the farm. That translated to a 37.9 percent mark in the majors, which doesn't bode well for home run avoidance. In a home ballpark that amplifies four-baggers by as much as 14 percent, he'll have to limit his fly balls to preserve this kind of play.

Of course, Eickhoff managed to accomplish just that in 2015. With an average fly ball distance of 274.3 feet aiding him, he allowed a home run on 8.9 percent of his fly balls, the best figure on the team among pitchers with as many innings. But it's one thing to post a number like that in 50 frames of work; it's quite another to keep it up when given 30 trips through the rotation. A sample this small leaves me with too much doubt to pick anything other than ZiPS.

Batting average on balls in play

Fly balls do come with some advantages — namely, that when they don't travel over the fences, they tend to fall into the gloves of defenders. Eickhoff benefited from this in both the minors (.281 lifetime BABIP) and the majors (.257 BABIP this year). Will he continue to do so going forward?

The 34.0 percent hard-hit rate he gave up in 2015 says no. We shouldn't stop there, though — not only do statistics like that fluctuate in small samples, Eickhoff still has the batted-ball profile on his side. He accrues fly balls while also collecting the best type: popups. Following an 8.6 percent popup rate in the minors, he goaded hitters into an infield fly 5.0 percent of the time as a major leaguer. That many automatic outs will help anyone suppress his BABIP, especially if accompanied by similarly routine fly balls to the outfield.

And, of course, there's the matter of defense. For a fly-baller such as Eickhoff, we care about the outfield. Unless the Phillies bring in a free agent, they know who will man two-thirds of theirs — 2015 rookies Odubel Herrera and Aaron Altherr, both of whom excelled with the glove and should continue to do so. If Cody Asche holds down the other spot, the team's defense will struggle, as he's never distinguished himself in the field. Nevertheless, the former two should play well enough to negate the latter. Together with Eickhoff's tendency for easy flies, this should help him turn balls in play into outs.

This shouldn't come as a surprise. Recent research by Jeff Zimmerman (hooray BtBS!) has suggested that pitchers with significantly low ground ball rates will outperform their peripherals, aligning with what Matthew Murphy had postulated earlier in the year. Eickhoff presumably won't put up a BABIP in the .250s again, but something in line with Steamer looks reasonable.

Strand rate

This is the weirdest area of Eickhoff's 2015 breakout. Through his minor-league tenure, he prevented runners from scoring only 69.7 percent of the time. Yet somehow, he managed to leave on 80.7 percent of his runners during his big-league cup of coffee. While no one in their right mind thinks he'll maintain that sort of output, Steamer and ZiPS don't concur on the degree to which he'll regress, with the former taking a more optimistic one than the latter. With which system should we align our predictions?

On the one hand, Eickhoff prevented runners from moving ahead this season. Despite none of his 19 pickoff attempts paying off directly, the opposition only tried to steal three times in 65 chances — a 4.6 percent clip that trailed the MLB average of 5.4 percent. In terms of Defensive Runs Saved's rSB component, this saved the Phillies a run over an average pitcher. If Eickhoff continues to pitch to Cameron Rupp and his 37.0 percent caught-stealing rate in 2016, he could sustain this.

On the other hand, Eickhoff struggled significantly with runners on base this season. His wOBA against rose from .249 to .309, the sort of spike that allows everyone to come around and score. His peripherals show a near-uniform decline:

Situational	K%	BB%	BABIP	HR%
Bases Empty	23.1%	3.9%	.239	2.3%
Men on Base	26.0%	11.0%	.296	2.7%

Like most pitchers, Eickhoff gave out more free passes with men on base. Unlike most pitchers, he also saw batters make better contact against him in those situations. His ground ball, line drive, and hard-hit rates all increased once runners had reached, so this doesn't shock me.

Not all hurlers can remain cool in the clutch. Eickhoff never displayed the ability to do this in the minor leagues, which tells me that his major-league results won't happen again. Here, the pessimism of ZiPS wins out.

Conclusion

We'll take Steamer (i.e. the over) on the strikeouts and BABIP, and ZiPS (i.e. the under) on the home runs and LOB%. This makes our hypothetical Eickhoff a below-average yet still adequate starter. While he will almost certainly pitch worse in 2016 than he did in 2015, he can still snag a spot in the starting rotation. The rebuilding Phillies will give many fringe types an opportunity next season, and hopefully Eickhoff will make the most of his.

. . .

Ryan Romano is an editor for Beyond the Box Score. He also writes about the Orioles on Camden Depot (and on Camden Chat that one time) and about the Brewers on BP Milwaukee.