Mediating projections, Part IV: Rajai Davis

Welcome back to Mediating Projections, wherein I examine the notable disagreements between projection systems. If, for some reason, this ridiculously esoteric concept intrigues you, look through the first three installments for more. Today, we'll take the case of Rajai Davis, who joined the Indians this offseason after two years with the Tigers. With one outfielder on the mend and another still acquainting himself to the position (albeit pretty well), Cleveland may have to rely heavily on the 35-year-old Davis. Will he produce?

System	PA	K%	BB%	ISO	BABIP	wOBA	wRC+	Def	fWAR	fWAR/600
Steamer	364	20.3%	5.4%	.116	.291	.282	76	-5.4	-0.2	-0.3
ZiPS	389	19.0%	5.1%	.144	.304	.325	93*	3.0	1.5	2.3

*ZiPS projects OPS+ instead of wRC+; the difference should be minimal.

Well, that depends on which projection you believe. Steamer foresees his age catching up with him; ZiPS thinks he'll keep chugging along. What will come to pass? Let's find out!

Strikeouts

Opponents have never fanned Davis too often, as he's posted a lifetime strikeout rate of 17.2 percent. In 2015, however, that figure climbed to 20.5 percent, far too high for a hitter who lacks power. Will he remain a strikeout threat, like Steamer predicts, or will his punchouts regress toward his career norm a bit, as ZiPS feels they will?

Oddly enough, Davis's peripherals held pretty steady last year. Per Baseball-Reference, his looking strike rate increased from 16.4 percent for his career to 17.0 percent in 2015, and his swinging strike rate decreased from 10.3 percent to 10.1 percent. With a lower rate of strikes as well — which fell from 67.1 percent to 66.7 percent — he didn't appear to earn his higher strikeout rate at all.

Looks and whiffs alone don't set a hitter down on strikes. Mike Podhorzer's research has connected fouls to strikeout rate, since extending an at-bat adds strikes to the count and pushes it in the pitcher's favor. And indeed, Davis fouled off more pitches last year, raising his foul rate from 19.0 percent to 20.2 percent. Unsurprisingly, his pitches per plate appearance went from 3.56 to 3.70, which likely helped him strike out more.

But does it justify that degree of increase? The expected strikeout equation from Podhorzer's article yields a 2015 mark of 19.1 percent — which means Davis was unlucky. Even if he continues to pile up the fouls, his respectable looking and swinging strike rates should keep his strikeout rate respectable. In other words, given the probability of some regression here, I'll side with ZiPS.

Walks

Interestingly, the two systems flip-flop here: Steamer takes the optimistic view, ZiPS the pessimistic one. A career free pass rate of 5.9 percent — and a 2015 mark of 5.6 percent — illustrates Davis's paucity of patience. Still, the absence of a few walks can make a difference. Should we expect his bases on balls to stabilize or drop?

As noted above, Davis took more balls in 2015 — not a great deal more, but more nonetheless. Plus, his PITCHf/x plate discipline metrics back that up. With an O-Swing rate of 31.5 percent, 2015 Davis chased less often than pre-2015 Davis did (33.4 percent). Because his zone rate stayed about the same, his expected strike rate plummeted by a full percentage point, from 67.8 percent to 66.8 percent. While the unfettered expansion of the strike zone will probably prevent him from reaping the benefits of this, he still shouldn't have a notably lower walk rate. And, of course, the uptick in pitches per plate appearance would aid his free passes too, since he would work his way deeper into the count.

With this in mind, why would anyone (or any projection system) think Davis' walk rate will decline? Perhaps the specter of 2014, when he took a free trip to first 4.5 percent of the time, haunts their decision-making. That output, though, didn't have the backing of underlying metrics. Davis took strikes 67.2 percent of the time — in line with his career clip — and posted an expected strike rate of 66.0 percent. His recent aggressive campaign appears to be a fluke, so the Steamer prediction looks the most reasonable.

Isolated Power

This may be the most odd aspect of Davis's case. Prior to last season, Davis owned a lifetime ISO of .110; his highest mark, .121, had occurred in 2012. Then he burst out with a .182 ISO, giving him the clout to make up for his anomalous spike in strikeouts. 2015 has come and gone, and 2016 will likely bring one of two things: a return to the ways of old (Steamer) or a bit of power retention (ZiPS).

Broadly, Davis' new home park hurts power. Via John Choiniere, right-handed hitters in Progressive Field have seen their ISO decrease by four to six percent in recent years. Comerica Park, which Davis used to inhabit, played more neutral. On a more minute level, Progressive really harms triples and home runs for righties, whereas Comerica inflates the former and doesn't impact the latter.

Why do three- and four-baggers concern us? They caused Davis's breakout:

Year(s)	2B/PA	3B/PA	HR/PA
2006-2014	5.1%	0.9%	1.1%
2015	4.3%	3.0%	2.2%

That jump in triples should set off the luck alarms. Davis had never hit more than two percent triples before last season, and as one of the flukier offensive statistics, they can bounce around like crazy without any underlying change in skill. Davis has always run fast, a skill that he should hold on to, but I can't imagine him cranking out triples at this rate.

Similar logic applies to the long balls. Davis put up a career-high 9.1 percent home run-fly ball rate in 2015, supported by an also-career-high 288.0 foot average fly ball distance. A late explosion such as that, however, usually doesn't stick over the long-term. Even with the fly ball distance to back it up, Davis just doesn't have the resume of a power hitter, giving Steamer the edge here.

Batting average on balls in play

For speedsters in Davis's mold, this facet of offense matters the most. How many hits can they churn out? Will they do so at an above-average level, fulfilling ZiPS' vision, or will they falter, to satisfy Steamer? ("They", obviously, being Davis.)

Across his 3,504 major-league plate appearances, Davis has put together a respectable .317 BABIP. That fell a bit to .308 in his most recent effort; still, he managed to top the major-league average of .299. He's notched a mark above .300 in every season except 2008 and 2011, so he doesn't have an ominous track record in this regard. With a career ground ball rate of 45.6 percent (44.1 percent in 2015) and an infield hit rate of 11.0 percent (10.3 percent in 2015), he's always put his fleet feet to good use, and I don't see why that would cease for 2016.

Older players always carry the risk of a total BABIP collapse. That often happens, though, to the hitters who rely on bat speed and blazing line drives, which can diminish with age. Davis and his lifetime 23.2 percent hard-hit rate (MLB average is 29 to 30 percent) thus won't have to worry about that too much. Nor should Davis's home impinge him: The Tigers and Indians occupy similar stadiums when it comes to BABIP. All in all, ZiPS takes the cake here.

Defense

On a 600-plate appearance basis, Steamer thinks Davis's glove will cost the Indians 8.9 runs. After the same playing time adjustment, ZiPS pegs him as a 4.6-run defender — in the positive direction. That makes for more than a win of difference, meaning this will influence him immensely. Up to which projection will Davis live?

In 2015, Davis performed pretty well in the field. According to UZR, he saved 4.3 runs, translating to a 2.5 Def — which, over 600 trips to the dish, comes close to ZiPS. There's just one problem: Last year stood out for Davis, because he hadn't played that well in years. His last positive campaign with the glove came in 2009; for his career, he's lost 16.6 runs and accrued a -30.6 Def. Further decline for a player in his mid-thirties wouldn't surprise anyone, especially considering the unreliability of one-year defensive samples.

Don't believe the metrics? Let's look at scouting opinions. Over the past four seasons, 64 outfielders have compiled at least 3,000 innings. According to Inside Edge scouts, Davis ranks:

• 56th in "routine" percentage;
• 41st in "likely" percentage;
• 40th in "even" percentage;
• 26th in "unlikely" percentage; and
• 33rd in "remote" percentage.

Across the board, he falls short — particularly where it counts the most. (Routine plays comprise the majority of most fielders' opportunities; they've made up 67.8 percent of Davis's chances since 2012.) Both objective and subjective evaluators have dinged his defense, which makes me want to side with Steamer here.

Conclusion

Adding it all up, we find that Steamer wins three contests, while ZiPS wins two. That belies the results, though — more accurately, Davis has a positive outlook in three areas (strikeouts, walks, BABIP) and a negative future in two (ISO, defense). Should he maintain the relative prosperity of his Detroit days, Davis will help Cleveland tremendously. And even if he falls off a bit, the satisfactory speedster should remain a useful player.

. . .

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.