Image Credit: Middlebury College
Before we begin this week’s Sunday Morning Sabermetrics article, let’s take a look at the Earned Run Average (ERA) leaders in Major League Baseball heading into games on Sunday, July 10th:
1. Clayton Kershaw* LAD 1.79
2. Madison Bumgarner* SFG 2.09
3. Drew Pomeranz* SDP 2.47
4. Johnny Cueto SFG 2.47
5. Jose Fernandez MIA 2.52
6. Kyle Hendricks CHC 2.55
7. Noah Syndergaard NYM 2.56
8. Jacob deGrom NYM 2.61
9. Stephen Strasburg WSN 2.62
10. Steven Wright BOS 2.68
For all of the lists included in this article, an asterisk (*) denotes a left-handed pitcher.
That list will change as we get into the more objective basic statistics to evaluate pitchers. Essentially, the thesis of this article is that the above list is not the top ten best pitchers in baseball thus far in 2016.
ERA is simple to calculate (9*(ER/IP)) has many flaws. It does not take into account the home ballpark of the pitcher, which plays a role in determining how many hits and home runs he allows. A pitcher with a 3.00 ERA at Coors Field in the thin air of mile-high Denver or the cozy confines of Fenway Park in Boston or Camden Yards in Baltimore is much more impressive than a pitcher with a 3.00 ERA in the generous dimensions of parks as Petco Park in San Diego (home of the 2016 MLB All-Star Game), Angels Stadium in Anaheim, or Safeco Field in Seattle.
The second flaw is that ERA is not robust when it comes to varying offensive environments throughout baseball history. An ERA of 3.00 in the high offense, PED era of the late 1990s and early 2000s is much more impressive than the same ERA held down during the Dead Ball Era of the 1910s or the 1960s. We will use two players whom I think Big Train fans will recognize at the end of this article (I warned you it was coming (http://bigtrain.org/news/?article_id=367)) to illustrate this concept. As a diehard Boston Red Sox fan, it is my all-time favorite pitcher comparison.
To fix these first two flaws, we turn to ERA+, which includes a ballpark adjustment, and is based on league-average ERA. An ERA+ of 100 always represents league average, regardless of the season. Anything above 100 is an above average pitcher; anything below 100 is below average. This is counterintuitive, so be sure to read ERA+ carefully and remember this rule when making judgments on pitchers’ abilities. The formula for ERA+ is below:
ERA+ = 100*(league-wide ERA/Pitcher’s ERA) + ballpark adjustment
“Pitcher friendly” parks (Marlins Park, U.S. Cellular Field, O.Co Coliseum, or Shirley Povich Field) get a penalty when determining ERA+, while “Hitter friendly” parks (Globe Life Park, Progressive Field, Citizens Bank Park, or Herndon High School) get a bonus.
Below is the top ten pitchers for ERA+:
1. Clayton Kershaw* LAD 220
2. Madison Bumgarner* SFG 192
3. Danny Salazar CLE 171
4. Steven Wright BOS 169
5. Tyler Chatwood COL 164
6. Johnny Cueto SFG 163
7. Stephen Strasburg WSN 162
8. Drew Pomeranz* SDP 161
9. Kyle Hendricks CHC 161
10. Jose Fernandez MIA 160
A few things stand out when considering this list compared to the original ERA rankings. Clayton Kershaw and Madison Bumgarner who place first and second, respectively, in ERA by a landslide, still lead in ERA+ even after ballpark adjustments. Third place is Danny Salazar, who did not even make the top ten in unadjusted ERA with a 2.75. However, Salazar pitches for Cleveland, home of the most hitter-friendly, difficult-to-pitch in park in the American League, Progressive Field. Those cozy confines lead to a generous bonus for Salazar and catapult him to third in ERA+, making him arguably the strongest candidate to start the All-Star Game for the American League. The same argument I just made for Salazar also applies to Steven Wright, who just barely cracked the top ten in unadjusted ERA but leaped to fourth since he also pitches in a hitter-friendly park, Fenway Park.
The best takeaway from this list is number 5, Tyler Chatwood, who finished a mere 21st in unadjusted ERA, with a 3.08. However, as mentioned earlier, a 3.08 in Coors Field is more impressive than a 2.47 in either San Francisco or San Diego, as evidenced by Chatwood’s higher ERA+ than both Johnny Cueto and Drew Pomeranz. Chatwood is an underrated pitcher who would be garnering much more attention if he pitched anywhere other than a mile above sea level for half of his starts. If more people paid attention to ballpark adjustments, Chatwood would begin to be mentioned in the same sentence as Bumgarner, Cueto, and Stephen Strasburg, and deservedly so.
There are two flaws with ERA+ that I feel merit attention. First, the minor one, is that the league-wide ERA is MLB league-wide, as opposed to discriminating American League versus National League ERA. The league-wide ERA in the NL is almost always lower than that of the AL, sometimes by as much as half a run. This is because the pitcher bats in the NL, and rarely, if ever, is he as much of a run-producer as an average DH in the AL.
The second flaw with ERA+ is the more problematic one. ERA+ is based on ERA. That means that ERA+, improved it may be, does not account for the fact that some pitchers have better defenses behind them than others and that affects how many hits, runs, and even earned runs they allow. This especially affects pitchers with low strikeout rates who ‘pitch to contact.’ A pitcher with a 120 ERA+ for the poor-fielding Oakland Athletics, who have cost their pitchers 52 runs this season via poor defensive play, is much more impressive than a 120 ERA+ for the stud-defense Kansas City Royals, who have saved 35 runs for their pitchers this season.
Defensive runs saved relies heavily on range, and certain players simply do not have the range that others do, and that leads to more hits, and, earned runs. A left side of the infield consisting of shortstop Andrelton Simmons (LAA, 124 career defensive runs saved) and third baseman Adrian Beltre (TEX, 221 career defensive runs saved) will stop far more potential hits than will former shortstop Derek Jeter (NYY, -246 defensive runs saved) or third baseman Pablo Sandoval (BOS, -12 defensive runs saved in 129 games with Boston). If I’m pitching, I want the Red Sox outfield behind me. I certainly do not want the outfield defense of the Baltimore Orioles. The Bosox “Soul Patrol,” currently consisting of the speedsters Brock Holt, Jackie Bradley Jr., and Mookie Betts, has saved 13 runs in 2016 for a pitching staff that has badly needed it this season. In contrast, the Orioles outfield, consisting of big guys Hyun Soo Kim, Adam Jones (who, dare I say, has actually not been a good centerfielder throughout his career), and Mark Trumbo, has cost O’s pitchers 12 runs thus far in 2016.
Two statistics aid us in fixing this issue. The first, RA9, is simply ERA but using both earned and unearned runs. RA9 is calculated as follows:
RA9 = 9*(R/IP)
The issue with RA9, though it does not care whether the play was scored a hit or an error, is that it still does not take into account the fact that better fielders stop more balls and prevent hits, extra bases, and runs, earned or unearned. One interesting case study is Steven Wright again, who leads the American League in unadjusted ERA with a 2.68, but has given up so many unearned runs (largely the result of passed balls from Boston catchers having trouble with the knuckleball) that his RA9 is more than a full run higher at 3.87.
There is a third statistic that we will cover today, that finally addresses the issue of fielding discrepancies among teams. Fielding Independent Pitching (FIP) takes into account only those statistics that the pitcher has a large amount of control over: home runs allowed, walks, hit batters, and strikeouts. The formula is as follows:
FIP = ((13*HR + 3*(BB + HBP) – 2*K)/IP) + Constant
The constant is to get FIP into ERA range to better compare the two stats, and is usually around 3.1.
The top ten leaders in Fielding Independent Pitching are as follows:
1. Clayton Kershaw* LAD 1.7
2. Noah Syndergaard NYM 2.06
3. Jose Fernandez MIA 2.13
4. Johnny Cueto SFG 2.7
5. Corey Kluber CLE 2.96
6. Stephen Strasburg WSN 2.97
7. Jake Arrieta CHC 3.06
8. Madison Bumgarner* SFG 3.15
9. Drew Pomeranz* SDP 3.18
10. Aaron Nola PHI 3.2
Attributes as high strikeout rates, low walk rates, low home runs allowed rates (biggest one), and fewer hit batters improve FIP. Non-home run hits allowed will have no impact on a pitcher’s FIP.
Walter Johnson or Pedro Martinez? ERA+ Revisited:
There is no denying that Walter Perry Johnson (WSH), the “Bethesda Big Train”, is one of the all-time greats. With a career ERA of 2.17, the all-time record for career shutouts with 110, an incredible 5914.1 career innings contributed, somehow just 97 home runs allowed in those 5914.1 innings (0.1 HR per 9 IP), 3509 career strikeouts, twelve strikeout titles, six ERA+ titles, nine FIP titles (the 0.1 HR/9 helped with that), and two MVP awards, Walter Johnson is on a short list of legends.
He did it in the worst era ever for offense. The Dead Ball Era of the 1910s overlapped with the prime of the Big Train, and helped him in putting up such staggering unadjusted statistics. Johnson’s career ERA+, which is dependent on the offense of the league, was 147, very good, no doubt, but not the greatest of all time.
Before I defend Pedro Martinez (LAD/MON/BOS/NYM/PHI), it is important to acknowledge that Mariano Rivera (NYY) is the official title-holder for all-time career ERA+, with a 205. However, Rivera was obviously a relief pitcher and played his career largely an inning at a time.
Pedro Martinez, like Rivera, pitched his prime in the PED era of the late 1990s and early 2000s, when offense was explosive and pitchers had nightmares. League-average ERA over Martinez’s career was more than two full runs higher than that over Johnson’s career. I stated earlier that a 3.00 ERA in the late ‘90s is more impressive than a 3.00 ERA in the 1910s. I don’t think anyone will oppose that.
What will draw arguments is the claim that a 2.93 ERA in the late ‘90s (Martinez) is more impressive than a 2.17 ERA in the 1910s (Johnson). The concept of ERA+ supports my claim. Martinez had a 154 career ERA+, exceeding Johnson’s 147 (recall that a higher ERA+ is better).
I’m not here to say Pedro Martinez was more accomplished than Walter Johnson. Walter Johnson had much more longevity, and his 152.3 career Wins Above Replacement (WAR) is one of the greatest of all time and far exceeds Martinez’s 86.0 career WAR. What I am saying is that if you told me I had one single game to win tonight, and I could have any pitcher in history in their prime on the hill for my team, and I knew nothing else, the size of the ballpark, the quality of bats and balls we’d be using, nor the hitting ability of the opposing lineup, treason as it may be, I want Pedro el Grande more than I want the Bethesda Big Train on that mound.
Next Week (7/17): The Bill James Game Score for pitchers. We will take a different angle on this article. The previous articles apply to the MLB far more than to the Cal Ripken Collegiate Baseball League, because 40 games is simply not enough to use stats as FIP, WAR, OPS+, and the like. But the Game Score is a game-by-game stat that we can use to evaluate even our own Big Train pitching staff, which, with a 3.31 team ERA that ranks second in the Cal Ripken League, should feature an abundance of acclaim.
Previous Sunday Morning Sabermetrics Articles:
June 5th: Pythagorean W-L:
June 12th: Wins Above Replacement:
June 19th: OBP/SLG/OPS/OPS+
June 26th: BABIP, ISO, and wOBA
July 1st: SecA, RC, and wRC+
There was no spreadsheet for this week’s article since all three metrics covered are available on baseballreference.com.