MLB

What Can We Expect From Rick Porcello in 2016?

Rick Porcello had an unusual 2015 season, so what can we expect from the Boston righty this year?

Rick Porcello is coming off arguably the worst season of his career, and the timing here was less than ideal if you're a Red Sox fan.

After three seasons in Detroit as a roughly three-win player, Boston paid him like one after trading for him last offseason, signing Porcello to a four-year, $82.5 contract.

Unfortunately for the Red Sox, he was worth half that in 2015, posting a 1.6 fWAR season in 172.0 innings, his lowest value since his rookie season in 2009.

The righty posted a 4.92 ERA (tied for a career worst) and a 104 FIP- (his worst since 2009).

Surprisingly, though, there are still reasons to be optimistic about Porcello.

In 2015, his biggest issues stemmed from a trifecta of " luck dragons:" balls in play, strand rate and home runs per fly ball rate. While these are all things Porcello has struggled with in the past, the degree with which they were problematic last season seems unsustainable.

Porcello also improved significantly in the second half, posting better strikeout, walk and ground ball rates after the all-star break, all of which seem linked (at least in part) to increased sinker usage.

Regression Coming....Right?

If you’ve taken SABR 101, you can see that Porcello’s BABIP (.332), strand rate (67.5%), and home run per fly ball rare (14.5%) were out of whack with the league averages (.293, 72.9%, 11.3%) last season.

So, if you paid attention in class, you would expect the first three numbers to regress towards the last three. And you would certainly be right.

Before closing the book here, though, it's worth noting Porcello has consistently underperformed his peripherals, and since we now have a 1245-inning/4000+-ball-in-play sample to work with, a closer look is probably warranted.

After his rookie season, when he ran a .277 BABIP and 75.5% strand rate, Porcello’s ERA has been higher than his FIP in every season, with the 0.53 gap between the two stats ranking 21st in the majors during this span.

While this could be cause for alarm, for his career, Porcello has a .313 BABIP, 69.9%, and 11.8% HR/FB rate. Worse than average, yes, but not exactly cause for riots in the street, especially because he's a ground ball pitcher (career 51.3% ground ball rate), and such hurlers tend to run slightly higher BABIP and HR/FB rates.

He was victimized by BABIP severely once before, in 2012, when he allowed a .344 BABIP. This though, was sandwiched between .316 and .315 BABIP seasons; aside from 2012 and 2015, his .316 BABIP in 2016 was a career-high, with career lows of .277 and .298 (in 2014).

While there is volatility in publicly available batted ball data, his career line drive rate of 20.4% is in line with American League averages, and his 26.8% hard-hit ball rate is better than average ( per FanGraphs).

It’s also worth noting he has not exactly had stellar defenses behind him, as from 2009 to 2014, Detroit ranked 24th in Defensive Runs Saved and 18th in Ultimate Zone Rating.

Last season, the Red Sox ranked 16th in DRS and 15th in UZR.

Considering all this, it would probably be best not to overstate Porcello's issues with contact.

As for his home run per fly ball rate, 2015 was the third time in Porcello’s career that he had a HR/FB rate higher than 14.0%, which on the surface seems like cause for panic.

Then again, after posting a 14.5% rate in his rookie season, he posted rates of 9.9%, 9.9%, and 11.5% over the next three years.

In 2013, 14.1% of his fly balls left the yard, but this dropped back down to 9.5% in 2014, so if there is actually some skill deficiency that causes Porcello to allow a higher than average HR/FB rate, it is a trait he apparently gains, loses, and then gains again (probably not actually the case).

In terms of his troubles with men on last year, while this has been a recurring problem for Porcello, it was nothing truly out of the ordinary in 2015.

Last season, his strikeout minus walk rate was 17.7% with the bases empty and 11.4% with men on, while his xFIP rose from 3.52 to 3.98 (the American League averages last year were 14.0% and 3.92 with no one on, and 10.5% and 4.20 with runners on base).

For his career, he has had a bigger gap in the K-BB% splits (12.4% with the bases empty, 5.4% with men on), so it is not unreasonable to assume his true talent strand rate is in fact lower than the league average (he has faced over 2,000 batters in both splits, while strikeout and walk rates stabilize after 200 batters).

Still, in 2015, his 67.5% LOB% was below his career rate of 69.6%, so it seems fair to chalk at least some of this up to random variation and expect regression this season.

Second Half Sinker Surge

Splits can be dangerous.

There are certainly times they can tell us something, but it is worth keeping in mind all splits inherently reduce the sample size we are looking at, increasing the likelihood there is no real trend but just random variation.

Jeff Zimmerman found that for both hitters and (especially) pitchers, full season stats in one season correlate with full season stats in the next season better than second half stats alone.

All things being equal, a larger sample is better than a smaller one.

With Porcello, who had a 5.90 ERA in 100.2 first-half innings and a 3.53 ERA in 71.1 innings in the second, however, there might be more here than random variation.

As a pitcher who has been an outlier in terms of factors like strand rate and BABIP, you might expect this dip in ERA to be a product of expected regression to the mean.

You would surprisingly only be partially right.

From the first to second half, Porcello’s left on base rate did increase (from 64.8% to 71.6%), but his BABIP (which went from .331 to .333) and HR/FB rate (14.5% to 14.3%) hardly moved.

The bigger factors in this ERA dip seem to be a significant rise in his strikeout rate -- which climbed to 23.3% from 18.1% -- which coincided with improvements in both walk rate (5.0% from 5.3%) and ground ball rate (47.6% from 44.4%).

All three stats are among the most reliable metrics in small samples, and seem to reflect Porcello pitching differently after the All-Star break.

Porcello has made a living off working down in the zone and inducing ground balls. From 2009 to 2014, he threw his sinker 46.8% of the time and his fourseam fastball just 19.5% of the time, according to data from Brooks Baseball (note that Pitchf/x seems to have identified what we'll refer to here as a sinker as a two-seam fastball).

By looking at his zone profile during this span (see below), it is very easy to see how he was among the leaders in ground ball rate among starting pitchers, inducing a worm killer on 52.1% of balls in play (and on nearly 60% of sinkers put in play).

Procello Raw Number of Pitches

During the first two months of 2015, however, a downward trend in Porcello’s sinker usage that began in 2013 and intensified in 2014 continued, as he only threw the pitch 31.4% of the time through May. After its usage rose slightly in June, it dropped again after two starts in July, giving his sinker a usage rate of 34.81% for the first half.

At the all-star break, his ground ball rate was 44.4%, which was both on pace to be a career low and was actually below the American League average of 44.6%. Porcello was also throwing his four-seamer almost 31.0% of the time (on pace to be a career high), while sporting a zone profile that was very different from that of the beginning of his career.

Porcello Raw Number of Pitches

In the second half, Porcello began to rely on his sinker again, which its usage spiking back up to 44.91. This naturally coincided with an increase in ground ball rate (47.6%) and a decrease in home run rate (from 1.43 to 1.14).

Predictably, this also produced a zone profile that bares much more resemblances to that of his first six big league seasons.

Porcello Raw Number of Pitches

In addition to helping lower the number of fly balls (which in turn lowered the number of balls which left the yard), Porcello’s sinker also helped keep him ahead in the count.

From 2009 to 2014, Porcello threw a first-pitch sinker 47.9% of the time, and got a called strike or foul ball on 50.7% of these pitches, while throwing a ball 35.0% of the time (the rest were put into play).

During the same span, he threw a first-pitch four-seamer 22.5% of the time, and while the pitch produced a strike/foul about 55.0% of the time, it also resulted in a first-pitch ball 37.7% of the time.

In the first half of 2015, this fastball became his pitch of choice on fresh counts, throwing it 37.9% compared to 29.9% for his sinker.

The results were similar, with the four-seamer producing an 0-1 count 55.1% of the time and a 1-0 count 37.2% of the time, compared to 53.9% and 30.0% for his sinker.

Things flipped in the second half, when Porcello threw his sinker 62.1% of the time on 0-0, reaching 0-1 on 57.7% of these pitches. 72.2% of these pitches resulted in a strike, foul, or ball in play, well above the AL average on all pitches of 60.8%.

Being ahead in the count obviously leads to better results, as AL batters had a .226/.266/.348 slash line with a 27.8% strikeout rate and 4.4% walk rate after an 0-1 count last season (compared to a .269/.373/.448 line, a 16.0 K% and 13.8% walk rate after 1-0).

By starting off more batters with his sinker, Porcello induced more ground balls and got ahead of batters at a greater rate. It's not hard to see this as being a factor in his second-half turnaround.

What the Projections Say

As you wait for us to release our numberFire projections for 2016, some other models have their forecasts set for the coming season.

These include Steamer and ZiPS, both of which can be found at FanGraphs.

Name ERA G IP HR SO BB WHIP K/9 BB/9 HR/9 FIP
Steamer 3.81 31 190 20 148 44 1.25 6.99 2.07 0.95 3.74
ZiPS 4.18 30 187.1 20 140 39 1.27 6.7 1.9 0.96 3.7


Steamer projects Porcello to be worth 2.8 fWAR, meaning he would provide value closer to what the Red Sox are paying for (by ZiPS, he is projected to be a 2.3-win player, though ZiPS uses a different WAR formula than the purely FIP-based model at FanGraphs).

As you can see, there isn't a huge difference between the models, aside from a disparity in ERA; given the similar WHIPs and home run rates, it’s likely the difference is due to difference strand rate predictions.

All things being considered, Porcello had a pretty bizarre 2015, making him an interesting player to keep an eye on in 2016.