This is an attempt to measure how much a select group of MLB managers worked their starting pitchers in the years 1960 thru 1988. 1960 was the year prior to baseball's first expansion and 1988 was the first year in which pitch counts were officially tracked.
To come up with this measure every boxscore of every major league regular season game, including suspended games for the years 1960 thru 1988 amounting to a total of 107,939 games, was downloaded in order to obtain each starting pitcher's pitching line for said games. Pitching lines were used to estimate pitch-counts. Because actual pitch counts weren't available for all years tracked except 1988, starting pitcher pitch-counts were estimated by using Tom Tango's basic pitch count estimator, including for the year 1988. Tango’s basic pitch count estimator seems to be quite accurate. The formula is:
3.3 PA + 1.5 SO + 2.2 BB, where PA = 3 IP + H + BB.
With estimated pitch counts in hand, five categories were then created in order to measure how much a manager had worked his starting pitchers. Below are the five categories that were used:
1. Pitcher Work Points (PWP)/162 games or (PWP/162)
2. Percentage of a manager's top 5 PWP pitchers of total PWP assigned to said manager
3. Percentage of Slow-Hooks (SH) of total games managed
4. Percentage of Slow-Hooked (SH) pitchers of total number of starting pitchers managed
5. Ratio of Total Slow-Hooks to number of Slow Hooked pitchers
The Pitcher Work Point (PWP) statistic is an off-shoot of Pitcher Abuse Points (PAP) which was created by the Baseball Prospectus in 1998. The PAP statistic is a measure as to how much "abuse a pitcher" was subject to based on his pitch count. The higher the PAP's, the more at risk the pitcher was to injury. PAP's were calculated by first creating pitching tiers and then assigning points for each pitch within that tier. For example, if a pitcher threw between 1 and 100 pitches each pitch in that tier was worth zero points. If a pitcher threw between 101 and 110 pitches each pitch in that tier was worth one PAP. For pitches 111-120, each pitch in that tier would be worth two points. The tiers of 10's climb up to tier 5 which contains pitches 141-150 and are worth five PAP's for each pitch. For pitch 151 or greater, PAP’s/pitch are worth six points.
Here is an example- a pitcher throws 115 pitches. Since zero PAP points are given to pitch counts below 100, zero points are assigned to the starter for his first 100 pitches. However, for the 15 pitches thrown above 100, PAP's are assigned in the following manner- one point for each pitch in the 101-110 pitch tier plus two points for each pitch in the 111-120 pitch tier. Therefore, a pitcher who throws 115 pitches in an outing accumulates 20 total PAP's.
Several years later the Baseball Prospectus PAP formula was altered. The new formula was as follows: (Number of pitches – 100) raised to the power of 3. With this new formula a pitcher would be assigned 3,375 PAP's for the same 115-pitch outing or (115-100) ^3.
Instead of using either of the Baseball Prospectus PAP formulas described above, a hybrid of the two formulas was created and used to calculate Pitcher Work Points or PWP. In the PWP version 100 is subtracted from the number of pitches thrown. The result is then multiplied by the tier number. For example, if the pitch-count of 115 is used yet again, the total number of PWP's is calculated as follows: (115-100) X 3 = 45 PWP’s. The number 3 represents the tier in which the 115 pitch-count is located.
PWP's were assigned to managers only for the games that they had managed in. In other words, if a manager was hired or fired during the season, his PWP total would start or stop on the day he was hired or fired. A manager’s PWP total was then prorated for a full season i.e. PWP/162.
Slow-hooks (SH) are defined as starts made by a pitcher in which he had pitched more than nine innings or a start in which he allowed seven or more runs or the pitcher’s innings pitched plus his runs allowed equaled 13 or greater. With the SH data three more categories were then created: percentage of slow-hooks of total games managed, percentage of slow-hooked pitchers of total number of starting pitchers managed and the ratio of total slow- hooks to number of slow-hooked pitchers.
Z-scores were then calculated for each of the five categories. The sum of the five z-scores was then used to create an overall rank for the managers. The higher the total z-score, the more the manager was inclined to work his starting pitchers. Scores were calculated for 24 different managers. Managers were chosen by way of most games managed in the years 1960 thru 1988. If a manager managed beyond 1988, those years were not included in the manager's overall score. Below are the top 5 managers in terms of working their starting pitchers:
By this method, Billy Martin is rated the highest in terms of managerial workloads for starting pitchers. Martin finished way ahead of the field in terms of PWP/162 games (9,241) and percentage of starts that resulted in a slow-hook (SH) - 11.03% which produced z-scores of 1.72 and 2.61 respectively. The next closest manager in terms of PWP/162 was Red Schoendienst who was ranked eighth overall. In terms of PWP’s accumulated by a manager’s top 5 PWP pitchers, Martin finished sixth. Martin’s top 5 starting pitchers in terms of PWP accounted for 37.44% of his total PWP’s. Those pitchers are listed below:
After managing the Minnesota Twins for one year in 1969, Martin was hired by the Detroit Tigers in October of 1970. Martin managed the Tigers for the better part of three seasons – ’71, ’72 and ’73. In those years the pitchers Martin worked the hardest were Mickey Lolich and Joe Coleman.
In 1971, Martin’s first year in Detroit, Tiger starting pitchers accumulated a total of 12,527 PWP’s, almost 5,000 points greater than the 7,823 MLB average PWP total and about 1,400 PWP’s above the Chicago Cubs’ 11,418 PWP total which was the second highest PWP total in baseball. Lolich and Coleman alone accounted for a combined 9,928 PWP’s or 79.30% of the Tiger total PWP’s.
In 1972 though, Tiger starting pitchers totaled “only” 8,332 PWP’s, a drop of 4,195. The Tigers’ 8,332 PWP total was just less than 1,000 points greater than the MLB average of 7,398 but 4,195 points less than their 1971 point total. Below is a table comparing 1971 Tiger pitcher PWP’s with 1972 Tiger pitchers:
The 1972 MLB season began with a 13-day player’s strike. The strike cost the Tigers six games and at least one start for both Lolich and Coleman. However, the main reason for both pitchers’ PWP decline was due, deliberately or not, to a lightened workload. The difference between Lolich’s and Coleman’s 1971 combined total of 9,928 PWP’s and their 1972 total of 6,322 was 3,606 PWP’s which is 86% of the 4,195 drop in Tiger PWP’s from 1971 to 1972.
Indeed, in 1971 Martin slow-hooked Lolich 11 times and Coleman four times. Conversely, in 1972 Martin slow-hooked Lolich only five times and Coleman three times. The difference in Lolich’s slow-hooks from ’71 to ’72 was mainly due to Lolich’s lack of opportunities to pitch into extra innings in ‘72. In ’71 Lolich pitched in five extra-inning games, including two 13-inning affairs. In ’72, Lolich was only needed to go beyond nine innings just once, that being on September 28 versus the New York Yankees. The Tigers entered that game only one-half games behind the Red Sox for first place in the AL East with six games remaining on the schedule.
Therefore, the main reason for the reduction in Lolich’s PWP’s in 1972 was not due to a change in Martin’s approach toward Lolich but rather Lolich’s extra-inning chances. That point was underscored by how Martin managed that crucial September 28th game versus New York with his workhorse Mickey Lolich starting.
In the bottom of the 11th inning with the score tied at two, Martin had Lolich lead off the inning to face Yankee reliever Sparky Lyle. Lolich worked a walk. Martin then had pinch-hitter Tony Taylor successfully advance Lolich to second base with a sacrifice bunt. However, even with Lolich now in scoring position, Martin refused to lift his pitcher for a pinch-runner. That decision may have ended up costing the Tigers the game. With Lolich at second and with only one out, Lyle uncorked a wild-pitch that Lolich failed to advance to third on. Lyle eventually escaped the inning unscathed. Lolich returned to the mound in the 12th inning and surrendered a homerun to Yankee outfielder Roy White. The Tigers ended up losing the game 3-2. When asked by the press why he didn’t pinch-hit or pinch-run for Lolich in the 11th inning, Martin simply replied, “He’s my best pitcher, so I left him in there (1).” Indeed.
The main reason for Joe Coleman’s 810 point reduction in PWP’s in 1972 was due to his seven less complete games pitched that season. Coleman’s average estimated pitch count per start in 1971 was 119 (equivalent to 38 PWP/start). In 1972 that figure dropped to 112 pitches per start (24 PWP/start). Over 38 starts the difference between 119 pitches thrown per start and 112 pitches thrown per start is 532 PWP’s, calculated as follows: (38 PWP’s – 24 PWP’s) * 38 games started. In Coleman’s case, the lightened workload may have been deliberate.
On several occasions during the ’72 season, Coleman disclosed the fact that he was dealing with “tendonitis (2)” in his right shoulder that had hampered his “effectiveness.” Coleman was also experiencing discomfort in his back stemming from a “muscle strain (3).” Coleman stated that he had been dealing with the discomfort(s) since the spring. He believed the issues were the result of playing too much golf during the off-season. Martin may have taken Coleman’s shoulder and back issues into consideration and managed his young all-star’s workload accordingly. In fact, when commenting on a spring-start made by a less than 100% healthy Coleman that year, Martin cited the number of pitches Coleman had thrown, “He (Coleman) threw 76 pitches and that’s too many (4).”
In 1973 Martin worked his Tiger starters to the tune of 8,899 PWP/162. Lolich’s PWP’s dropped from 3,169 in ’72 to 2,631 in ’73. Coleman’s total PWP’s increased slightly to 2,856 from his 1972 total of 2,703. The Tigers’ 8,899 PWP/162 was just over 600 points greater than their 8,332 1972 total. This is surprising considering the large jump in PWP/162 totals amongst American League teams in ‘73, presumably due to the league’s adopting the designated hitter rule that year. The table below shows the difference in each MLB clubs’ 1972 and 1973 PWP/162 totals:
The Tigers’ increase of 310 PWP’s from 1972 to 1973 was the lowest increase amongst AL teams. Of note, Martin had managed the Tigers in 134 games in ’73 before being fired. If Martin’s pace of 8,899 PWP/162 is used instead of the actual Tiger total of 8,642 PWP/162 the difference between ’72 to ’73 PWP/162 is 567, still the lowest amongst AL clubs. Unlike most AL managers, the implementation of the DH rule had virtually no effect as to how Martin managed his starters in 1973 when compared to the previous year.
Almost immediately after being fired by the Tigers, Martin was hired by the Texas Rangers. Martin managed the Rangers in 23 games in 1973. In October of that year, the Rangers acquired Fergie Jenkins in a trade with the Chicago Cubs. By his standards Jenkins’ was coming off a subpar 1973 season with Chicago. In Jenkins the Rangers had acquired a “workhorse” for Martin similar to what Martin had in Detroit in Joe Coleman. “Jenkins gives us the stopper we needed desperately. He’s a workhorse and he’s a winner (5)” is what Martin told the press once the deal for Jenkins was announced. Indeed, Jenkins was a workhorse. In fact, Jenkins had accumulated the 9th most PWP/Start in MLB over the last five years, resembling that of Martin’s Tiger workhorse, the aforementioned Joe Coleman:
However, even with the addition of Jenkins to the Ranger starting rotation, Martin’s PWP/162 for his starting pitchers was 8,869, in line with the MLB average of 8,484. While pitching for Martin, Jenkins’ PWP/start actually decreased by 10 points from Jenkins’ previous five years mark:
If we compare Jenkins’ 1974 season under Martin to his 1971 Cy Young Award winning season, the differences are even more dramatic:
In 1974 Jenkins had thrown three more innings, faced six less batters and completed only one less game than his 1971 totals yet Jenkins’ total PWP’s dropped by 23% from 3,733 to 2,892 under Martin. The eight less pitches thrown per start played a significant role in the decrease in PWP’s.
Apparently the decrease in pitches per start was by design. Martin had stressed to Jenkins the need to attack hitters. Jenkins’ aggressiveness in pitching within the strike-zone under Martin’s tutelage was the main reason for Jenkins’ 1974 turn-around. According to Martin, “He (Jenkins) was throwing a lot of pitches. When he stopped trying to pitch so fine he just started pitching strikes and now he’s getting ahead of the hitters (6).” Jenkins won a career high 25 games in 1974 and led the American League in complete games and strikeouts to walk ratio. He finished second in Cy Young Award voting to Oakland’s Catfish Hunter.
Of course Martin’s act didn’t last long in Texas. He was fired after 95 games in 1975 but was once again immediately hired, this time by the New York Yankees. Martin managed the Yankees for 56 games in ‘75. In 1976, Martin’s first full year in New York, Yankee starting pitchers’ PWP/162 was 10,409, over 3,000 points greater than the MLB average of 7,284 but almost identical to the 10,216 PWP’s Yankees’ starters accumulated in 1975 with Bill Virdon managing the team for 104 games. In 1977 Yankee starting pitchers’ PWP/162 dropped to 9,198. The MLB average also dropped but not at the same rate as the Yankee drop.
In 1978 Martin managed the Yankees in 94 games before being fired. At that point, Martin’s starters were on pace for a 6,437 PWP/162 which would have been just below the MLB average. With the exception of 1969 and 1988, Martin’s first and last years managing, 1978 was the only year in Martin’s managerial career in which his PWP/162 was lower than the MLB average. Below are Martin’s PWP/162 compared to the MLB average in the years ‘76 thru ‘78:
One of the main reasons for the Yankees’ drastic drop in PWP/162 under Martin in 1978 was the health of Jim “Catfish” Hunter. In 1977 Hunter began the season on the disabled list due to shoulder tendonitis and a bruised instep thanks to a line-drive hit off his foot. He ended the regular season back on the DL with a hernia. Hunter was limited to 143 innings pitched in 1977 and 118 innings in ’78. As a result, Hunter’s PWP’s declined from 3,940 in ’76 to zero in ’78. Below is a breakdown of the Yankees’ starting four pitchers in the years ’76 thru ’78 including Catfish Hunter:
A steady decline in PWP’s and estimated average pitch count for the Yankees starting four occurred from ’76 to ’78. In 1976 and 1977, Martin had slow-hooked his starters 25 and 21 times respectively. However, in 1978 Martin slow-hooked his starters only five times in 94 games- Figueroa (3), Gullett (1), Guidry (1). Prior to the 1978 season, the Yankees signed reliever Rich “Goose” Gossage to join 1977 Cy Young Award winner Sparky Lyle in the bullpen. The signing of Gossage impacted Martin’s managing of his starting pitching but not in the way one would expect. Gossage wasn’t setting up Lyle to end games or vice versa. Instead, Gossage did become the Yankees closer by way of pitching in the most high-leveraged and medium-leverage situations coming out of the pen. Sparky Lyle though saw his high-leveraged appearances decrease and his low-leveraged situations increase. The Yankees’ breakdown of bullpen appearances relative to leveraged situations in the years ’76 thru ’78 is contained in the table below:
The addition of Gossage in ’78 did not change Martin’s managerial approach in high or medium-leveraged situations with respect to his starters and his bullpen; however, Martin did utilize his bullpen more often in low-leveraged situations which could be the reason as to why Martin’s total quick-hooks of his starters exceeded slow hooks by a 4:1 margin (20 QH, 5 SH) that year. A quick-hook is defined as a start in which the pitcher has failed to throw at least six innings but has given up three runs or less.
Clearly though, the steady decline in Yankee starters’ PWP/162 during Martin’s first go-around in New York was mainly due to Catfish Hunter’s physical ailments catching up with him and the Yankees’ lack of a true workhorse to replace him.
In 1980 Martin was hired to manage the Oakland A’s. That year Oakland starting pitchers’ total PWP’s was a staggering 16,591. That figure was the second highest in the majors in the years 1960 thru 1988. Only the 1973 California Angels’ 18,531 PWP total was higher than Martin’s 1980 Oakland A’s mark. In ’80 Martin slow-hooked his starters a total of 35 times, far exceeding the MLB team average of 10. Below is a breakdown of the 1980 Oakland A’s rotation by way of PWP’s, slow hooks and estimated average pitch-count per start:
Rick Langford and Mike Norris finished first and second in PWP’s amongst all major league starters in 1980. Matt Keough and Steve McCatty finished fourth and ninth respectively. That year, the average PWP total per start for starting pitchers was approximately 36 points. Therefore a starting pitcher who started 30 games should have generated, on average, approximately 1,080 PWP’s. Every one of Oakland’s starting five far exceeded that figure.
During the 1980 season, as the complete games were piling up for his starting pitchers, the media began to focus on the workload Martin was placing on them. “I want to see how their arms are next year (7)” was what former Yankee player turned baseball broadcaster Tony Kubek said at the time regarding the A’s starters. Martin justified the use of his starting pitchers with his implementing a five-man rotation in 1980. “People forget we have a five-man rotation, not a four-man rotation (8),” Martin reminded the press late in the 1980 season. Once the season concluded Martin reiterated his five-man rotation claim, “People worried that all the complete games would wear them out but the dummies forgot our starters were pitching every fifth day, not fourth (9).” A’s starter Matt Keough concurred with his manager, “People haven’t examined our situation closely…we’re in a five-man rotation. How many starts will any of us get? How many innings will any of us throw (10)?”
While true, Martin had utilized a five-man rotation for most of the 1980 season, it was also true that the overwhelming majority of MLB teams had done the same. Oakland starting pitchers made starts in a total of 139 games with at least four days’ rest, 12th most in MLB. The MLB team average was just under 133 starts with at least four days’ rest. In 1979 Martin’s predecessor in Oakland, Jim Marshall, had A’s pitchers start 136 games with at least four days’ rest. Therefore, in terms of rest for his starters, what Martin had done in 1980 wasn’t unique yet he worked his starters at a historic rate.
Martin’s other reason for the workload placed on his starting pitchers in 1980 was due to his weak bullpen. “If the bullpen comes thru, we’ll use it more (11)” was Martin’s reply when he was once again asked about his starters’ complete game totals. In this regard, Martin had somewhat of a point. In November of 2012, Hardball Times writer Dan Lependorf came up with a way of grading bullpens based on the statistic Wins Probability Added. Specifically, Lependorf graded bullpens by way of Wins Probability Added less Win Probability Added/Leverage Index or WPA – WPA/LI. If we grade the 1980 Oakland A’s bullpen using Lependorf’s WPA - WPA/LI method, the -1.14 total generated by Oakland relievers ranked 14th in the majors that year; not very good but not historically bad either.
Some evidence does exist that perhaps Martin did not work his 1980 Oakland starting pitching staff as hard as has been perceived over the years and that Martin was fully aware of his starters’ pitch counts. In fact Martin had elaborated on pitch-counts with respect to his starters:
“It’s not the number of innings that’s important- it’s the number of pitches he throws. I remember when Nolan Ryan threw 300 pitches against the Yankees and then after couldn’t wipe his ass; after that he lost six games in a row….We knocked down the number of warmup pitches they threw in the bullpen. And many times they didn’t throw that many pitches in a game; Langford and Norris were usually in the 95-100 pitch range (11).”
When it comes to Langford, Martin’s above claim regarding his actual pitch-count may have some validity. Below is a list of Langford starts in which his actual pitch-count was reported by the media:
With the exception of his July 20th 14-inning effort in which he threw 164 pitches versus Cleveland, Langford was in the “95-100 pitch range” as Martin had claimed. Moreover, in the August 27, 1980 edition of the Sacramento Bee, the paper reported that Langford, “in his last seven starts had not thrown more than 100 pitches in a game (12).” That means that if the Bee was correct in its reporting, Langford hadn’t exceeded 100 pitches in all his starts after the 164-pitch Cleveland start including his starts on July 25th and August 10th, 15th and 21st which are not contained in Table 11. Taking the claim even one step further was San Francisco Examiner beat writer and card-carrying member of the BBWAA, Stephanie Salter. In her September 7 column titled “Langford’s 21st Full Game in Row,” which appeared in the paper after Langford had beaten the Baltimore Orioles the day prior, Salter wrote the following, “He (Langford) threw 104 pitches, not an undue number for a major league pitcher but more than he has tossed all year. The 28 year-old Langford had not thrown 100 pitches in any other outing (13).”
Given that Langford did indeed exceed 100 pitches in his July 20th 14-inning effort versus the Indians, Salter was mistaken in her claim. However, Salter may have meant to exclude Langford’s only extra-inning effort of the 1980 season. Of course Salter may also have been completely wrong. Either way, there is evidence that given Langford’s efficiency in terms of pitch-counts as per the news reports of various games, Martin may have not worked Langford as hard as has been perceived over the years.
Mike Norris though may be another matter. Martin had made the claim that Norris was also usually in the 95-100 pitch range. Norris’ actual pitch-counts were reported in four of his 1980 starts listed below:
Of those four starts, only the June 11th start against Baltimore did Norris throw for more than nine innings. In that start Norris went 14 innings, one of five starts in which Norris had pitched into extra-innings that season. In his September 5th start against the Orioles, Norris threw 113 pitches in seven innings. In the other two reported starts versus Toronto and Boston, Norris threw 134 and 145 pitches respectively in his nine innings of work. Therefore, the fact that Norris exceeded nine innings of work a total of five times in 1980 and that in his three reported nine-inning games, Norris far exceeded 100 pitches makes Martin’s claim that Norris was usually in the 95 to 100 pitch range rather dubious.
In 1981 the MLB season was interrupted by another player’s strike. The strike began on June 12 and ended on July 31. As a result, the A’s played a 109-game schedule as opposed to 162 games. The workload placed on Oakland starters by Martin in ’81 was still significantly higher than the MLB average but about 23% less than 1980:
Below is a comparison of PWP’s for A’s starters in 1980 and 1981. Figures for 1981 are pro-rated for a full-season:
In 1981 Martin did not work Langford at the historic rate that he had in 1980. However, Langford once again led the majors in complete games with 18. Steve McCatty was second in the majors in complete games with 16 as he clearly took over the number two spot in the rotation from Mike Norris. Norris’ very ordinary post-strike numbers (4W – 6L, 4.32 ERA) resulted in less work placed upon him by Martin. Prior to the strike, Norris averaged 104 PWP’s per start. After the strike, that figure was cut exactly in half to 52 PWP’s/start. Brian Kingman pitched mainly out of the bullpen in the second-half of ’81 due to Martin electing to go with a four-man rotation once the strike had ended.
Martin had actually implemented a six-man rotation in late May of ‘81. Martin’s implementation of the six-man rotation coincided with Oakland’s acquisition of left-handed pitcher Tom Underwood from the Yankees just one week prior. Martin’s reasoning for the six-man rotation: “This makes Tom Underwood one of our starters and gives us more versatility in the bullpen (14).” Martin also believed that the six-man rotation would make his starters, “all that much stronger (15).” Martin’s plan was to have his starters pitch out of the bullpen on their regularly scheduled throwing days. Martin made it clear that his plan to have starters pitching out of the bullpen “was not a reflection of my bullpen (16).” In fact Martin stated that his ‘81 Oakland bullpen was “the best in baseball (17)” that year.
Martin definitely had a point in that regard. Again, if we measure the Oakland bullpen by way of WPA-WPA/LI, the A’s 3.19 mark was indeed the best in baseball. As a result, Martin utilized his bullpen more often than he had in 1980. However, even though Martin had the “best bullpen in baseball” his PWP/162 was still significantly greater than the MLB average. The table below is a comparison between Martin’s bullpen usage in 1980 and 1981. *1981 figures are pro-rated to a 162-game season:
In terms of high leverage and medium leverage situations, the total number of times Martin had called on his bullpen in 1981 was 69, just nine times less than he had in 1980 in 53 less games. If the 1981 figures are pro-rated to a 162-game schedule, Martin’s total bullpen usage in ’81 in high and medium leverage situations is 103, 25 times more than his bullpen usage in 1980.
In 1982, as Oakland’s starting pitching went south so had Martin’s PWP/162. Martin’s 8,003 PWP/162 was still approximately 2,630 higher than the MLB average of 5,373 but for the first time since his arrival in Oakland; Martin did not lead the majors in the category. That distinction went to Milwaukee’s managerial combination of Buck Rodgers and Harvey Kuehn with 9,421 PWP/162 followed by Detroit’s Sparky Anderson with 9,112.
In 1983 Martin returned to the Yankees. The bulk of Martin’s 8,050 PWP/162 stemmed from his working of Ron Guidry. Guidry’s 3,063 PWP’s was just over 500 PWP’s greater than the next highest Yankee starter, Shane Rawley and over 1,300 more than Dave Righetti’s total of 1,707. Below is a comparison of the three Yankee left-handers:
Surprisingly, despite Righetti having a season equivalent to, if not better, than those of both Ron Guidry and Shane Rawley, Martin did not work Righetti as hard as he had Guidry and Rawley.
Righetti won the AL Rookie of the Year Award in 1981. He was 8-4 with a 2.05 ERA in 105 innings pitched. Righetti led the AL in ERA+ (174), FIP (2.12) and SO/9 (7.6). In 1982 Righetti regressed. His ERA jumped to 3.79 and he led the AL in walks with 108. For a better part of a month that season, Righetti had been relegated to the bullpen to straighten out issues with his control. Nevertheless Righetti was signed to a four-year $2.9 million contract prior to the 1983 season. During spring training, Martin made it clear that the Yankees needed a big season from Righetti to compete. “This staff has the potential to be the best in the league but we need that big year from Righetti. That’s the key, (18)” is what Martin told the press just prior to the opening of the season.
So despite the fact that Righetti had succeeded previously at the major league level, had just signed a new contract and according to Martin, was a key member of the rotation, Martin worked Righetti less than Shane Rawley. Righetti’s 1983 1,707 PWP total was still the 14th highest in baseball that year but by Martin’s standards it was considered low given how much Martin had worked young pitching stars like Joe Coleman (when completely healthy) and Mike Norris in the past.
Perhaps the main reason for Martin’s seemingly lightened workload for Righetti was the fact that Righetti had a tender shoulder during the spring and as a result had missed two weeks of spring training in order to rest it. Righetti was eventually able to make his first start of the season on April 7 versus the Seattle Mariners. In what Martin called an “overpowering” performance, Righetti pitched 5.2 innings, allowed three hits and struck out five. His estimated pitch count was 85. In his next start Righetti tossed 6.2 innings in a win versus the Tigers in which he allowed seven hits, struck out seven and walked one. His estimated pitch count for that start was 105.
Gradually Martin allowed Righetti to pitch deeper into games. By the end of June, Righetti was 9-3 with a 3.53 ERA in just under 110 innings pitched. In his first start in July which occurred on Independence Day, Righetti no-hit the Boston Red Sox. Righetti’s no-no was the first no-hitter thrown by a Yankee since 1956. After the game Martin was quoted as saying, “I put Dave (Righetti) in the same category as Catfish Hunter (19).” However, Martin did not work Righetti to the extent that Hunter was typically worked during his career or for that matter, any other of his top young starters such as the aforementioned Joe Coleman and Mike Norris. Below is a month by month breakdown of Righetti’s workload under Martin in 1983:
Clearly Martin began to work Righetti harder after his July 4 no-hitter. However, in September Righetti began to tire. He was ineffective in his last two starts of the season. He was unable to throw at least three innings in either start and was eventually shelved for the rest of the season due to a tired arm. His last start was made on September 18. The Yankees opted to convert Righetti to a closer prior to the 1984 season for several reasons which included the preservation of Righetti’s arm and to replace Goose Gossage who had left the Yankees after the season via free agency. The idea to move Righetti to the bullpen was opposed by Martin. Martin though was not returning to the Yankees in ’84. Righetti remained in the bullpen for the rest of his career.
Though Martin did not work Righetti to the extent as one would expect, Martin still worked the young left-hander more than his counterparts Bob Lemon and Gene Michael had but less than Clyde King. Righetti’s workload as a starter by each of his managers in the years 1981 thru 1983 is contained in the table below:
Both Gene McMichael and Bob Lemon managed Righetti in his rookie season in 1981 and again in 1982. Clyde King took over the Yankees from Michael on August 4 of ’82.
Like Righetti, Ron Guidry pitched for multiple managers during his career with the Yankees. Under Martin, Guidry accumulated a total of 10,163 PWP’s or 81 PWP/Start and an average estimated pitch count per start of 120. Below is a comparison of how each of Guidry’s managers worked him in the years 1977, his first year as a full-time starter, thru 1985 which was Guidry’s last 20-win season:
Martin is responsible for the three highest PWP/Start totals in Guidry’s career in ’79, ’83 and ’77 respectively. Martin is also responsible for Guidry’s four highest estimated average pitch-counts per start which fell in a narrow range of 121 to 123 pitches. Where Martin far exceeded his fellow Yankee managers was in the number of times Martin slow-hooked Guidry.
Guidry made a total of 124 starts under Martin. He was slow-hooked 23 times or in approximately 18.5% of his games started. Conversely, Guidry made 144 starts for all other managers other than Martin and was slow-hooked only six times or in 4.1% of his starts- four times by Yogi Berra and two times by Dick Howser. Yankee managers Bob Lemon, Gene Michael and Clyde King did not slow-hook Guidry.
Of the 22 times Martin did slow-hook Guidry, only once was the slow-hook the result of Guidry pitching into extra-innings. Martin slow-hooked Guidry 21 times in starts in which Guidry had given up five or more runs. The games in which Martin slow-hooked Guidry are contained in the table below:
During a two-month stretch in ’83 which began on July 13th with a 6-1 loss to the Twins, Martin slow-hooked Guidry a total of seven times. Prior to the July 13th start versus the Twins, Guidry had missed approximately three weeks due to back spasms. Two weeks later, Martin had Guidry complete another game in which the Yankees were on the losing end of a lop-sided affair, this time against the White Sox. Guidry threw eight innings and gave up seven runs (five earned) on eleven hits. He struck out nine and walked one.
In Guidry’s next start versus the Blue Jays, the left-hander gave up a then career-high 14 hits in a 6-2 losing effort. The estimated pitch count for Guidry’s consecutive starts against Chicago and Toronto were 136 and 135 for a total of 271. According to media reports though, Guidry’s actual combined pitch-count for the two starts was 294 pitches (20). When asked by the press why he elected to have Guidry pitch the entire game against the Blue Jays, Martin snapped back:
“I ain’t gonna burn up a bullpen when we’re down four or five runs. I’m not about to use pitchers when we’re down. Not when a pitcher’s out there pitching his heart out. They were just hitting bloopers and toppers. The score should have been 2-2 going into the ninth (21).”
At the time Martin claimed his bullpen had been over-worked and was down to just three men- George Frazier, Dale Murray and Bob Shirley. Right-hander Jay Howell had a strained knee and Goose Gossage had worked 1.2 innings the night before.
Martin brought up his over-worked bullpen again after he had Guidry toss a complete-game win versus Oakland on August 30th. When asked again as to why he stuck with Guidry for the entire game, this time Martin was calmer in his reply, “Very simple, I was a little strapped in the bullpen….I had Goose up and I was thinking about using him but Goose couldn’t get loose, so I forgot about it (22).” Martin then elaborated further, “And besides, I don’t concern myself with innings pitched so much as I do the number of pitches made. That’s what determines if a pitcher is overworked or not….some other guys I might have yanked but not Guidry (23).” Guidry’s estimated pitch-count for this start was 159.
Guidry was appreciative of Martin leaving him in the game to earn the win: “I feel lucky to have a manager like Billy who lets you pitch. You can’t win games on the bench (24).”
1983 was Guidry’s second 20-win season; his first was in 1978. In 1985 Guidry became a 20-game winner for the third time. Martin managed Guidry in all three of Guidry’s 20-win seasons. Martin managed a total of eleven 20-game winners in his career: Dick Boswell (’69), Jim Perry (’69), Mickey Lolich (’71, ’72), Joe Coleman (’73), Fergie Jenkins (’74), Ron Guidry (’78, ’83, ’85), Tommy John (’79) and Mike Norris (’80). Of the eight pitchers who won 20 games under Martin, six were first-time 20 game winners. Only Fergie Jenkins and Tommy John were 20-game winners prior to being managed by Martin.
Martin mentioned the importance of winning 20 games after he had Guidry go six innings despite allowing seven runs to the Detroit Tigers in a mid-September 1985 start at Tiger Stadium. “He (Guidry) was going for his 20th win,” Martin said, “and I was hoping we’d get him some runs and help him back into it (25).” Guidry went on to win his 20th game in his next start in Baltimore versus the Orioles. Nevertheless Martin slow-hooked Guidry in his following start and once again it was against the Orioles. In that game Guidry gave up five runs in nine innings pitched, walked four and struck out five. The Yankees though were able to get Guidry his 21st win of the season by knocking in two runs in the bottom of the ninth. Guidry’s estimated pitch-count for that game was 159. According to press reports however, Guidry’s actual pitch-count was 145 (26).
Once again Martin provided his reasoning as to why he left Guidry in the game despite Guidry giving up the five runs and four walks: “I had two things in mind; the win and the chance at the Cy Young Award (27).” Martin then added, “I almost took him out three times (28).” At the time the eventual 1985 Cy Young Award winner, Brett Saberhagen, had won 19. Saberhagen won his 20th game one day later.
1985 was one of the few years Martin’s PWP/162 was somewhat in line with the MLB average. That year Martin had accumulated a 5,772 PWP/162 mark, about 800 more than the MLB average of 4,922. Forty-six year-old knuckle-baller Phil Neikro and Ron Guidry totaled 2,156 and 1,494 PWP’s respectively under Martin. Martin had taken over the Yankees from Yogi Berra early in the 1985 campaign but was dismissed after the season.
In 1988 Martin returned to the Yankees one last time. Martin managed the Yankees in 68 games that year. His PWP/162 was 2,599, approximately 2,220 less than the MLB average. Saddled with a staff made up of pitchers past their prime such as the 45 year-old Tommy John, 37 year-old Ron Guidry, 35 year-old Rick Rhoden and 34 year-old John Candelaria, the Yankees finished third last in the AL in pitching. Martin ended up having to quick-hook his starters in 13 of the 68 games he managed that season. The estimated average pitch-count for Yankee starters under Martin in 1988 was 91. 1988 was Martin’s last year managing.
It is clear that Martin worked his starters harder than any other of the 23 managers reviewed. Martin was way ahead of the field in terms of PWP/162 as well as the percentage of starts in which he slow-hooked his starter. Martin is fourth in average estimated pitch-count per start with a mark of 105. In only three seasons in which Martin managed, excluding the 23 games he managed for the Texas Rangers in ’73, was Martin’s PWP/162 less than the MLB average. Martin was also responsible for the historic 1980 Oakland A’s rotation’s workload which accumulated the second highest PWP total in the years 1960 thru 1988. The loyalty Martin had shown toward his workhorses such as Mickey Lolich, Rick Langford, Mike Norris and Ron Guidry by allowing them to finish games definitely impacted Martin’s overall rank. And even when Martin had very good bullpens, his PWP/162 was still at or greater than the MLB average.
However, there is evidence that Martin exercised restraint when it came to working pitchers who health wise, may have been less than 100% such as in the cases of Joe Coleman, Catfish Hunter and Dave Righetti. Martin wasn’t indiscriminate when it came to working his starting pitchers, he worked his top pitchers harder; however, when those pitchers’ production slipped, as was the case with Mike Norris in 1981, Martin adjusted accordingly.
And finally, Martin was well aware of pitch-counts. He too was of the belief that the number of pitches a pitcher throws determines how hard the pitcher has worked. As was the case with Fergie Jenkins, Martin had stressed the minimizing the number of pitches thrown by attacking the hitters in the strike zone. The same could be said of Rick Langford who threw 28 complete games for Martin in 1980. In fact Langford may have been so efficient in his pitching that he rarely threw over 100 pitches during that 1980 season. Billy Martin was according to some a baseball genius, so it isn’t surprising that there was much more involved as to how and why he worked his starters as mush as he did during his very unique managerial career.
Sources:
Times Herald: September 29, 1972 (1)
Detroit Free Press: May 28, 1972 (2)
Windsor Star: August 24, 1972 (3)
Times Herald: March 13, 1972 (4)
Poughkeepsie Journal: October 26, 1973 (5)
Kilgore News: August 7, 1974 (6)
Boston Globe: August 24, 1980 (7)
Atlanta Constitution: September 21, 1980 (8, 10)
San Francisco Examiner: February 22, 1981 (9, 11)
Sacramento Bee: August 27, 1980 (12)
San Francisco Examiner: September 7, 1980 (13)
Odessa American: May 27, 1981 (14, 15)
St. Louis Dispatch: May 31 (16, 17)
Daily News: March 28, 1983 (18)
Daily Record: July 6, 1983 (19)
Harford Courant: August 4, (20, 21)
Press Democrat: August 31, 1983 (22)
The Record: August 31, 1983 (23, 24)
Daily News: September 18, 1985 (25)
Daily News: September 29, 1985 (26)
Press and Sun-Bulletin: September 29, 1985 (27, 28)