Yesterday, we looked at starting pitching over the last 100 years through a basic statistical lens. Today, we work on tools for measuring the quantity pitched in evaluating a pitcher's career.
I can already hear the question- "isn't it as simple as innings pitched, or if you want to get really fancy, number of batters faced?". Well, no. We will get to that in just a little bit. But first, I would like to explore some basic quantity measurements for batters, pitchers and fielders. For this, I will use Tommy John, Ozzie Smith and Rabbit Maranville. They are a long serving pitcher and two long-serving shortstops, perhaps the two greatest defensive players at the position.
In his career, Tommy John threw 4,710 innings and faced 19,692 batters. In his, The Wizard of Oz had 8,375 assists, 4,249 putouts and made 10,778 plate trips. Maranville had 8,957 assists, 6,413 putouts and 11,256 plate trips. Interesting, but what exactly is the point? Shortstops putouts are mostly routine- popups and flips and throws from the second and first basemen and pitcher on DPs and fielders' choices, and one looks at assists plus a small fraction of putouts to assess the quantity of work that a shortstop does defensively. In real terms, this means that John's quantity measure is almost double the Maranville and Smith defensive quantity measure. This significantly affects overall value. For instance, Baseball Prospectus measures pitching runs and fielding runs both as against average (PRAA and FRAA) respectively and as against replacement players (PRAR and FRAR). Here is how John, Maranville and Smith stack up:
This chart tells us that Maranville and Smith were much better shortstops compared to their peers than John was a pitcher compared to his peers, but that John saved more runs for his teams pitching than Maranville and Smith did for their teams fielding. That's simple quantity working for the old workhorse. I am not suggesting that this made him more valuable overall, as Smith particularly had considerable value at the plate and on the basepaths.
Now for the batters, let's look at Pete Rose. Rose made 15,861 plate trips, and is the career leader in this category. Despite facing almost as many pitchers as John faced batters, Rose's career BRAR (batting runs above replacement) of 979 is lower than John's 1188 PRAR, while his BRAA (batting runs above average) of 520 far outstrips John's 118. Rose was a better hitter relative to his peers than John, but added less value with his bat than John did with his arm. Rest easy, I am not suggesting for a moment that John was a better ballplayer. We will get to his Hall of Fame case in a week or two.
So, what does all this mean? Being able to throw a solid starter's workload with average effectiveness has quite a bit of value. Being able to throw 15-20 average or a wee bit better seasons together makes a starter one of the best, and extraordinarily valuable to his teams.
What is a solid starter's workload, anyway? It depends on when he pitched. Let's take Tommy John. He pitched from 1963 to 1989, with the great majority of his career being spent in the American League for the White Sox and Yankees. Yesterday's chart showed us that during the years 1965, 1970, 1975, 1980 and 1985, the average 3rd-5th best starters threw 265, 291, 300, 276 and 262 innings, or an average of 279 innings per season. John threw 4710 innings, or 16.9 seasons (based on 279 innings per season).
With Pete Alexander, whose career spanned 1911-1930 (with only 21 innings in 1930), I would use the 1910, 1915, 1920 and 1925 figures to generate an average of 294 innings. Alexander threw 5190 innings or 17.7 seasons. When I list pitcher's statistics for comparison purposes in the Hall Watch, there will be both innings and seasons listed.
Next week, we'll deal with quality measurements-ERA+, fielding independent pitching, and contributions to team wins above and beyond run prevention. Merry Christmas.
I can already hear the question- "isn't it as simple as innings pitched, or if you want to get really fancy, number of batters faced?". Well, no. We will get to that in just a little bit. But first, I would like to explore some basic quantity measurements for batters, pitchers and fielders. For this, I will use Tommy John, Ozzie Smith and Rabbit Maranville. They are a long serving pitcher and two long-serving shortstops, perhaps the two greatest defensive players at the position.
In his career, Tommy John threw 4,710 innings and faced 19,692 batters. In his, The Wizard of Oz had 8,375 assists, 4,249 putouts and made 10,778 plate trips. Maranville had 8,957 assists, 6,413 putouts and 11,256 plate trips. Interesting, but what exactly is the point? Shortstops putouts are mostly routine- popups and flips and throws from the second and first basemen and pitcher on DPs and fielders' choices, and one looks at assists plus a small fraction of putouts to assess the quantity of work that a shortstop does defensively. In real terms, this means that John's quantity measure is almost double the Maranville and Smith defensive quantity measure. This significantly affects overall value. For instance, Baseball Prospectus measures pitching runs and fielding runs both as against average (PRAA and FRAA) respectively and as against replacement players (PRAR and FRAR). Here is how John, Maranville and Smith stack up:
John 111 PRAA 1188 PRAR Maranville 167 FRAA 991 FRAR Smith 236 FRAA 753 FRAR
This chart tells us that Maranville and Smith were much better shortstops compared to their peers than John was a pitcher compared to his peers, but that John saved more runs for his teams pitching than Maranville and Smith did for their teams fielding. That's simple quantity working for the old workhorse. I am not suggesting that this made him more valuable overall, as Smith particularly had considerable value at the plate and on the basepaths.
Now for the batters, let's look at Pete Rose. Rose made 15,861 plate trips, and is the career leader in this category. Despite facing almost as many pitchers as John faced batters, Rose's career BRAR (batting runs above replacement) of 979 is lower than John's 1188 PRAR, while his BRAA (batting runs above average) of 520 far outstrips John's 118. Rose was a better hitter relative to his peers than John, but added less value with his bat than John did with his arm. Rest easy, I am not suggesting for a moment that John was a better ballplayer. We will get to his Hall of Fame case in a week or two.
So, what does all this mean? Being able to throw a solid starter's workload with average effectiveness has quite a bit of value. Being able to throw 15-20 average or a wee bit better seasons together makes a starter one of the best, and extraordinarily valuable to his teams.
What is a solid starter's workload, anyway? It depends on when he pitched. Let's take Tommy John. He pitched from 1963 to 1989, with the great majority of his career being spent in the American League for the White Sox and Yankees. Yesterday's chart showed us that during the years 1965, 1970, 1975, 1980 and 1985, the average 3rd-5th best starters threw 265, 291, 300, 276 and 262 innings, or an average of 279 innings per season. John threw 4710 innings, or 16.9 seasons (based on 279 innings per season).
With Pete Alexander, whose career spanned 1911-1930 (with only 21 innings in 1930), I would use the 1910, 1915, 1920 and 1925 figures to generate an average of 294 innings. Alexander threw 5190 innings or 17.7 seasons. When I list pitcher's statistics for comparison purposes in the Hall Watch, there will be both innings and seasons listed.
Next week, we'll deal with quality measurements-ERA+, fielding independent pitching, and contributions to team wins above and beyond run prevention. Merry Christmas.
