# The Z Files: Making the Sausage Part Two, Pitcher Projections

Last week we looked at some "what if" hitting scenarios, using an Excel tool that helped project a batter's homers and batting average based on skills. This week we'll do the same for pitching, using ERA and WHIP as our target metrics.

It's important to review some differences between the hitting and pitching tool as pitching isn't as straightforward. To keep things simple, the expected ERA formula designed by researchers Dwight Gil and Tad Reeve will be employed. This projects ERA based on hit, home run, walk and strikeout rates. Granted, there have been improved means of estimating ERA, but for the scope of this discussion, this will suffice.

To use batting average on balls in play (BABIP) in our analysis, at-bats (AB) have to be part of the calculation. The catch is, the Gil and Reeve calculation uses innings in the denominator and the number of AB per inning pitched is different for every pitcher. As such, if you decide to work with the downloadable Excel tool, you'll need to set an innings factor to make the pitcher's WHIP match the projected WHIP based on his BABIP and other peripherals.

Like with the hitting tool, BABIP will be divvied into grounders, fly balls and line drives. We won't account for a pitcher inducing hard, medium or weak contact.

Like last week, plate appearances are the starting point with strikeout percent, walk percent and home run per fly ball (HR/FB) rate used to ultimately determine homers, hits, whiffs and walks.

Last week we looked at some "what if" hitting scenarios, using an Excel tool that helped project a batter's homers and batting average based on skills. This week we'll do the same for pitching, using ERA and WHIP as our target metrics.

It's important to review some differences between the hitting and pitching tool as pitching isn't as straightforward. To keep things simple, the expected ERA formula designed by researchers Dwight Gil and Tad Reeve will be employed. This projects ERA based on hit, home run, walk and strikeout rates. Granted, there have been improved means of estimating ERA, but for the scope of this discussion, this will suffice.

To use batting average on balls in play (BABIP) in our analysis, at-bats (AB) have to be part of the calculation. The catch is, the Gil and Reeve calculation uses innings in the denominator and the number of AB per inning pitched is different for every pitcher. As such, if you decide to work with the downloadable Excel tool, you'll need to set an innings factor to make the pitcher's WHIP match the projected WHIP based on his BABIP and other peripherals.

Like with the hitting tool, BABIP will be divvied into grounders, fly balls and line drives. We won't account for a pitcher inducing hard, medium or weak contact.

Like last week, plate appearances are the starting point with strikeout percent, walk percent and home run per fly ball (HR/FB) rate used to ultimately determine homers, hits, whiffs and walks.

Before we really dig in, it's best to be grounded with respect to plausible BABIP projections. We're at the point where the original contention that a pitcher has almost no control over his BABIP is considered a misnomer. Ignoring a pitcher's ability to avoid hard contact, by simply allowing more or fewer fly balls his BABIP can change as evidenced by the following table:

 GB FB LD BABIP extreme fly ball 35.3 43.8 20.9 .282 league average 45.3 33.8 20.9 .292 extreme ground ball 55.3 23.8 20.9 .302

Leaving the line-drive rate constant and adjusting the ground ball to fly ball ratio shows a 0.010 difference in BABIP from the league average.

Sometimes it helps to see what others have done as we attempt to frame what we expect. Here's a look at the 10 best career BABIP for pitchers with over 1,500 IP since 1995:

 RANK PLAYER BABIP GB% 1 Matt Cain .265 37.6 2 Ted Lilly .270 34.1 3 Jered Weaver .271 33.3 4 Clayton Kershaw .272 45.6 5 Tim Wakefield .273 39.6 6 Jarrod Washburn .273 36.2 7 Barry Zito .274 38.0 8 Johan Santana .276 37.2 9 Woody Williams .276 37.7 10 Carlos Zambrano .277 48.4

With the exception of Clayton Kershaw and the enigmatic Carlos Zambrano, the 10-lowest qualifying BABIPs of the last two decades are all fly-ball pitchers. Now let's check the 10 highest:

 RANK PLAYER BABIP GB% 1 John Burkett .316 40.8 2 Aaron Sele .315 41.5 3 Ricky Nolasco .314 41.7 4 Shane Reynolds .312 45.6 5 Jeff Fassero .310 48.7 6 Edwin Jackson .310 44.2 7 Sidney Ponson .309 51.2 8 Brian Moehler .309 44.5 9 Andy Pettitte .308 48.5 10 Esteban Loaiza .308 42.2

While there is a ground ball tilt, they're not all extreme ground-ball pitchers, suggesting a higher line-drive rate or hard contact is responsible. Now here are some of the more recognizable qualifiers:

 PLAYER BABIP GB% Pedro Martinez .279 38.8 Tom Glavine .281 45.3 Greg Maddux .284 51.5 Kevin Brown .286 53.6 Cole Hamels .286 44.0 Roger Clemens .289 46.8 Justin Verlander .290 39.7 Roy Halladay .292 54.1 Mark Buehrle .292 45.4 John Smoltz .294 45.9 Bartolo Colon .294 41.6 Felix Hernandez .295 54.4 Tim Lincecum .296 46.6 Curt Schilling .297 39.5 Chris Carpenter .297 51.2 Matt Morris .297 48.3 Randy Johnson .298 43.2 Cliff Lee .298 40.7 Zack Greinke .298 43.4 James Shields .298 44.7 Mike Mussina .299 43.3 Roy Oswalt .301 47.1 John Lackey .305 43.8

Keep this list in mind when you're deciding where to regress BABIP. It's simply not logical to expect anything below .280 unless the hurler is an extreme fly-ball pitcher. On the flip side, unless the pitcher is really horrible, anything over .315 isn't fair. A range of .280 to .315 isn't huge, but it's better than forcing everyone to .295 or so.

Now let's evaluate some of last season's unexpected performances, a couple good and one not so good.

Jake Arrieta, Chicago Cubs

Here are 2015's vitals for the reigning National League Cy Young Award winner.

 ERA WHIP K/9 BB/9 HR/9 BABIP HR/FB 1.77 0.86 9.3 1.9 0.39 .246 7.8

Keeping everything the same and taking his BABIP to .280, Arrieta's projected ERA rises to 2.28 with a 0.94 WHIP. The right-hander is an extreme ground-ball pitcher so there's a very good chance this season's mark is well north of that. Bringing it to a more logical .290 yields 2.43/0.97.

Arrieta was also a bit lucky with HR/FB. Edging that up to 9.0 (still below league average), we're now looking at 2.54/0.97.

Last season, Arrieta induced more grounders than ever. If he gives back only five percent of the worm burners, his numbers jump to 2.72/0.98.

We haven't even touched his strikeout and walk rate yet. Arrieta's strikeout percent has been consistent the last two seasons while his walk rate has been in decline, so there's a strong argument to be made we leave them alone. But for illustrative purposes, let's worsen each by a modest five percent. The surface stats are now 2.84/1.00.

RotoWire's projections is 2.55/0.99 while I have 2.80/1.08. Both can be fully supported and involve some give-back from last season's stellar campaign. Using the Excel tool helps visualize where the projection comes from and offers a basis for the small drop in skills and outcomes.

Zack Greinke, Arizona Diamondbacks

 ERA WHIP K/9 BB/9 HR/9 BABIP HR/FB 1.66 0.84 8.1 1.6 0.57 .229 7.3

Here's Greinke's digits from last year:

The veteran righty is moving to a more hitter-friendly park, which has some concerned. Let's be sensitive to that and raise his BABIP to .290 and HR/FB to league average. Greinke's ERA is now 2.91 with a 1.02 WHIP.

Penalizing the same five percent peripheral decline we did for Arrieta noses that to 3.01/1.04. RotoWire is more optimistic with ERA, checking on at 2.64/1.04 while I'm looking at a still outstanding 2.90/1.08.

Chris Sale, Chicago White Sox

Based on surface stats, last season was a huge disappointment for Sale owners. To wit:

 ERA WHIP K/9 BB/9 HR/9 BABIP HR/FB 3.41 1.09 11.8 1.8 1.00 .323 12.5

But check out those underlying metrics. By means of comparison, Clayton Kershaw's K/9 was 11.6 with a 1.6 BB/9. That's right, he and Sale have essentially identical skill sets. Park and league account for Sale possessing a higher home run rate, but if they switched teams, it wouldn't be shocking if Sale put up better numbers.

So what went wrong? Without blaming it all on bad luck, Sale's BABIP and HR/FB were both elevated last season. Let's revert them both to league average. Sale now spins a 2.88/1.01.

It wouldn't be fair to ding Arrieta's and Greinke's skills so applying the same five percent tax renders the lanky lefty at 3.03/1.04. RotoWire pegs him at 3.06/1.04, while I expect even better, 2.74/1.05.

One last thing ...

Before calling it a day, let's debunk one of the myths with respect to calling a lower ground-ball rate better. There are instances where it's actually more advantageous to be a fly-ball pitcher.

Here's a table showing the WHIP and ERA of several types of pitchers:

 ERA WHIP K/9 BB/9 FB% BABIP HR/FB Average 3.960 1.29 7.7 2.9 33.8 .292 .11 Extreme Ground Ball 3.719 1.30 7.7 2.9 23.8 .303 .11 Extreme Fly Ball 4.205 1.28 7.7 2.9 43.8 .282 .11 XFB, +K/9 3.971 1.25 8.7 2.9 43.8 .282 .11 XFB, +K/9, -BB/9 3.926 1.19 8.7 2.1 43.8 .282 .11 XFB, +K/9, -BB/9, -HR/9 3.443 1.16 8.7 2.1 43.8 .282 .08 XGB, +K/9, -BB/9, -HR/9 3.183 1.19 8.7 2.1 23.8 .303 .08

Walking you through, the top is a league average pitcher in every regard. Next we make him an extreme groundballer and an extreme fly-ball artist. As expected, the ERA is better for the ground-ball pitcher while the WHIP is superior for the fly-ball guy.

If we improve the strikeout rate of the fly-ball pitcher, his ERA is close to that of the league average pitcher with a better WHIP. Dropping the walks helps the WHIP even more.

Here's the real key. Put the fly-ball pitcher in a big park and the homers drop along with the ERA. Of course, mimicking these improved peripherals for a ground-ball pitcher really drops his ERA too.

Keeping all the skills the same, sure, the ground-ball pitcher will have a lower ERA than the fly-ball pitcher. But it doesn't take extenuating circumstances for a better skilled fly-ball pitcher to display superior outcomes to a lesser skilled ground-ball pitcher -- especially in a venue that depresses homers.

Remember, WHIP is more stable than ERA. A strong argument can be tendered that it's better to build a fantasy staff around fly-ball pitchers in big parks because the ERA will be very low and then late fate decide ERA.

For those interested, click HERE for the pitching tool, along with last week's hitting tool.