# Beyond ERA: The Numbers Behind the Red Sox Rotation

If the Red Sox are as concerned with their rotation as themedia, they’re not showing it.

This year’s team is expected to trot out Clay Buchholz, Rick Porcello, Wade Miley, Justin Masterson and Joe Kelly in roughly that order. Given their recent history, those arms are unlikely to strike fear in the opposition, but manager John Farrell seems to be content with his group.

Strictly looking at ERA, it’s easy to see why so many are concerned — but the advanced stats show this rotation could be better at preventing runs than many think. Here’s a primer on a few numbers that aid in evaluating this personnel group.

Wins Above Replacement

This poster child of sabermetric stats is most often applied to hitters, but it can also be used to evaluate pitchers. At its most basic, WAR attempts to compare pitchers to a theoretical “league-average’’ hurler to determine how much a guy helps or hurts his team in a given span. The stat differs depending on whose metric you use, with Baseball-Reference and Fangraphs providing the two most-used WAR stats. Baseball-Reference arrives at its stat by taking the number of runs a pitcher allows and the innings he pitches and running them through a complicated equation that compensates for team defense, park factors, the differences between relief appearances and starts and the strength of the opposition. The Fangraphs stat is based on Fielding Independent Pitching, a separate stat that attempts to compensate for luck and defense to give a truer measure of pitching skill than ERA.

The two stats often arrive at different conclusions — Buchholz netted 1.5 wins by Fangraphs’ calculations and lost 1.6 by Baseball Reference’s, for example — but are generally useful for determining whether a pitcher helped or hurt his team.

FIP/xFIP/SIERA

Despite its ubiquity as a measure of pitching performance, ERA is viewed as a flawed statistic. Since it’s calculated simply by taking the number of earned runs a pitcher allows then extrapolating that number to a nine-inning average, ERA fails to take into account such factors as defense and luck. As such, several stats known as ERA estimators that aim to compensate for these factors and better explain performance have emerged.

Fielding Independent Pitching, or FIP, acknowledges that what happens to balls — fair, foul, line drive, etc. — after they’re hit is largely outside the pitcher’s control. As such, it takes the pitch outcomes fully under a pitcher’s control — strikeouts, walks, struck batters and homers — and runs them through a weighted formula to come up with a number that roughly corresponds to an in-a-vacuum ERA.

Expected FIP, or xFIP, is similar to its parent stat except it normalizes home run rates to league average in the equation. About 10 percent of fly balls become home runs, and if a pitcher deviates from that number in either direction it’s generally more due to luck than performance.

Skill-independent earned run average, or SIERA, is a newer, more complicated ERA estimator acknowledging that while pitchers generally can’t control what happens to a ball when it’s put in play, some achieve positive outcomes more than others. In contrast to FIP and xFIP, SIERA weighs strikeouts more heavily, penalizes pitchers with high walk rates, forgives pitchers with low walk rates, and takes into account whether balls in play are fly balls, grounders or line drives, each of which tend to have significantly different outcomes.

When it comes to the Red Sox rotation, the ERA estimators suggest Buchholz, Miley and Masterson were to some extent victims of bad luck in 2014, while Kelly and Porcello more-or-less saw the results they should have. That said, the estimators also suggest past great seasons like Buchholz’s remarkable (pre-injury) 2013 or Miley’s 2012 breakout were more luck than skill.

Strand Rate, HR/FB and BABIP

Three of the so-called “luck stats,’’ strand rate, home run to fly ball ratio and batting average on balls-in-play, are largely outside of pitchers’ control. Significant deviations from league average in any can make pitchers appear significantly better or worse than they really are.

Strand rate equals the number of batters a pitcher leaves on base. League average is roughly 70 percent, with numbers above and below that leading to misleadingly low and high ERAs respectively. As mentioned above, about ten percent of home runs become fly balls, so HR/FB ratios above or below that tend to be more luck than skill. League-average BABIP — the frequency with which anything other than a strikeout, walk, hit batter, catcher’s interference, sacrifice bunt or homer becomes a hit — is about .300, with deviations usually attributable to luck and team defense.

Pitchers’ skillsets, the quality of defense behind them, and the park they pitch in can influence any or all three luck stats, but in general a pitcher who significantly deviatiates from league average is likely to regress.

Batted Ball Distribution

Batted ball distribution — the percentage of pitches that become grounders, line drives and home runs — is a key factor in why the Red Sox assembled the collection of talent they did this offseason. With career ground ball rates of about 50 percent or higher, all the Red Sox starters generate grounders at significantly higher rates than league average. Grounders are generally the most desirable batted ball, as batters tend not to reach base much — in 2014, batters hit just .239 on grounders. Fly balls translated to base hits slightly less often (.207 last year) but also bring the possibility of home runs and more extra-base hits, making them more dangerous than grounders. Liners are by far the most dangerous of the three, as batters hit a whopping .685 on them in 2014. The more ground balls a pitcher gets, the fewer liners and fly balls he allows.

With Dustin Pedroia and Pablo Sandoval ranking as plus-defenders, Xander Bogaerts improving at short, and Mike Napoli about average at first, the Red Sox infield defense should play well to its staff’s strengths.

Rate Stats

Broadcast-friendly rate stats like strikeouts and walks per nine can help explain success or failure, but the advanced rate stats better tell the tale. Three of the most useful are swinging strike rate, outside swing percentage and contact percentage, all of which strongly influence pitching success. As their names suggest, swinging strike rate is the percentage of swings and misses a pitcher generates, outside swing percentage the frequency with which a pitcher gets hitters to chase outside the zone, and contact percentage the frequency with which hitters get bat on the ball.

Wade Miley was not much of strikeout pitcher until last year, when his K/9 jumped to 8.18 from 6.53. This is largely explainable by his swinging strike rate, which rose to 9.7, meaning his newfound swing-and-miss stuff is likely legit. One of Porcello’s greatest assets is his control, and he had career-best 1.80 BB/9 last season. Batters swung at his outside offerings a third of the time last year, meaning would-be balls often turned into strikes or poor contact. As far as the latter is concerned, Porcello is a case study for contact pitchers. With a career contact percentage of 84 percent (and upwards of 90 in the zone) Porcello lives and dies by weak contact, which leads to his high 52.1 percent career ground ball rate.

Pitch Type, Values and Velocity

Every pitcher brings with him a different repertoire, with these stats serving to help evaluate his bag of tricks. Pitch type tells how frequently a pitcher throws a pitch, value how many runs it saves or gives up versus a league-average version of the pitch and velocity is a measure of how fast the ball travels.

When a pitcher sees sudden success or decline, these stats often serve to show why. Masterson provides a textbook example, his success strongly correlating with the speed of his arsenal, which has been often hampered. According to data by pitch-tracking system PITCHf/x listed on Fangraphs, Masterson has a four-seam fastball, slider, sinker and changeup he’s used at various rates depending on the season. In his All-Star campaign of 2013 (14-10, 3.45 ERA) Masterson’s fastball averaged 93.1 miles per hour, his sinker 91.1, his slider 82.9 and his changeup 85.7. The slider was his most effective pitch, saving over two runs per 100 pitches per PITCHf/x pitch values, with the changeup, sinker and fastball following in that order. Each pitch other than the changeup lost two or more miles-per-hour in 2014, directly contributing to his lackluster follow-up campaign. His early-season velocity will go far in showing which Masterson the Red Sox got.

Innings Pitched

A high strikeout rate and low walk rate are nice, but they mean little if pitchers aren’t going deep in games. Both Porcello and Miley were iron men last season, each topping 200 innings and accounting for 4.4 WAR between them, per Fangraphs’ calculations. Each averaged over six innings per start, which has the added benefit of taking pressure off the bullpen. On the other side, neither Masterson nor Kelly topped 130 innings or six innings per start, meaning they were neither durable nor able to stay deep in games — the Red Sox will have to hope they improve in both regards if the team is to contend this season. At 170.1 innings pitched over 28 starts, Buchholz went relatively deep in games when he played, but often struggled.