Thursday, May 29, 2014

Back to the '64 Phillies: Pitching Problems on the Horizon

Fifty years ago, the 1964 Phillies ended the month of May with a 25-15 record. They were in first place, only half-a-game up on the Giants and 2-1/2 ahead of the third-place Cardinals. The Reds--the fourth team to figure in the drama to come--were in sixth, 6 games out. As predicted by many pre-season analysts, including in Sports Illustrated, pitching was a Phillies' strength: through the first 48 days of the season, they had given up the fewest runs in the league and their team 3.15 ERA was second-lowest in the league after San Francisco's 2.92. Before too much longer, however, a healthy starting rotation was to become very problematic for manager Gene Mauch, which would have significant ramifications in the final weeks of September. This is the third installment in this blog's continuing series on what happened with the 1964 Philadelphia Phillies.

The '64 Phillies: Pitching Problems on the Horizon

Gene Mauch began the season with a four-man rotation featuring right-handers Jim Bunning, Art Mahaffey and Ray Culp and southpaw Dennis Bennett. Already by mid-May, as happens all the time in baseball, an arm injury was disrupting the Phillies' rotation--specifically, the sore elbow bothering Culp. The 22-year old had had an excellent rookie season in 1963, making 30 starts, going 14-11 and making the All-Star team, but pitched poorly in his first six starts of the season. Having pitched as many as five innings only twice, Culp was 1-4 with a 5.81 ERA when Mauch replaced him in the rotation with lefty Chris Short, who would prove to be the second most-valuable pitcher on the staff after Bunning.

Meanwhile, Dennis Bennett--after failing to last five innings in his opening day start--was one of the best pitchers in baseball as the season turned to June.  He led the Phillies with 10 starts, had 5 complete games (one of them a loss), had thrown a shutout, and was 6-3 with a 2.54 ERA. (Bunning was 5-2 with a 2.14 ERA in nine starts through the end of May.) Perhaps of some consequence in light of his 1964 season arc, on May 23 Bennett pitched 13 innings, giving up only two runs to beat the Dodgers but throwing 159 pitches to 51 batters, according to his game log on the website,

Bennett pitched well in his first two starts in June--giving up five runs in 15 innings against the Giants and Dodgers--but the rest of the month did not go so well. Although he was 2-1 in six starts that month, his earned run average in June was a far-less than elite 6.07. It is quite likely that the left shoulder problems that ended up diminishing his effectiveness and depleting Mauch's corps of reliable starting pitchers (Culp also being damaged goods) began surfacing this month. The genesis of Bennett's pained pitching shoulder stemmed from a car accident that sent him crashing through the windshield before the start of the 1963 season--his second year in the majors--in which he nonetheless went 9-5.

Both Bennett and Culp pitched through pain in July. Bennett started six games, relieved in three others and had a 3.98 EA in July, but lost four of his five decisions. The pitcher who was confident he would win 20 games in 1964 ended July with a 9-8 record and shoulder pain serious enough that he pitched only 24 innings in August, making only three starts with four stints out of the bullpen to an ERA of 4.88. Bennett lost two of his three starts, lost another game in relief, and twice went six days on ice as his shoulder made him unavailable to Mauch. His last three appearances in August were particularly concerning: Bennett faced 41 batters in 7.1 innings of work, meaning 19 (46 percent) got on base against him; his ERA in those three games--12.27.

As for Ray Culp, his sore elbow limited him to only 18 innings in the month after Mauch took him out of the starting rotation after his May 16 loss to Houston. The only game Culp started during this stretch was the second game of a doubleheader on June 9, which he lost 4-0 to the Pirates in less than five innings of work. The lowly New York Mets--the worst team in baseball--were the cure Culp needed to try to salvage his season. After pitching five innings of one-run relief against them in the first game of a twinbill on June 14 and getting his second win of the season, Culp started the second game of another doubleheader against them on June 19 and pitched a complete game for a 7-2 win. His next start, four days later, was also the back end of a doubleheader, this time against the not-very-good Cubs, in which Culp pitched a shutout and allowed only two to reach base, giving up a walk in the first and a single in the sixth.

Culp was back in the rotation on a regular basis all through July, pitching well with a 4-1 record and 2.42 ERA in six starts and two relief appearances. After another strong outing in his first start in August--six innings, giving up one run against Houston--Culp's next two starts were less than successful as his elbow pain became increasingly debilitating. Culp came out in the second inning of a start against the Mets on August 15, failing to get an out, and that was that. Culp did not make another start in 1964, appearing in only five more games as a reliever, just one during the September stretch, during which he gave up 19 hits, walked 6, and allowed 18 runs (12 earned) in only 9.2 pain-wracked innings.

Hurtful as it was, Ray Culp's elbow put him in his manager's doghouse; Gene Mauch thought Culp was unwilling to pitch through pain. Dennis Bennett, meanwhile, persevered. After his pained August, Bennett returned to the rotation at the beginning of September, making six straight starts on the norm of every four days, and his final start on five days' rest. By this time--when, as it turned out, the Phillies had become desperate for wins--shoulder pain made Bennett unable to deliver.  He gave up nine runs in his last nine innings of work as the Phillies' seemingly sure-thing pennant slipped away.

The physical travails of Culp and Bennett forced Mauch to improvise in his starting rotation, although Bennett did make 32 starts in 1964--the second-most on the team after Bunning, who was the undisputed ace and had a 19-8 record. Short, who at first replaced Culp, secured his place in the rotation with a 17-9 record, and his ERA of 2.20 turned out to be even better than Bunning's 2.63. Mahaffey, with a 12-9 record in 29 starts, was the fourth man in the rotation and a mostly-reliable pitcher until Mauch lost confidence in him at the worst possible time in the desperate final weeks of the season--about which this blog will be almost exclusively devoted to come September. Until then, this series will continue with period updates on the Philadelphia Phillies, 50 years ago.

Thursday, May 22, 2014

The DH Impact on Managing Starting Pitching in the Two Leagues: A Brief History

The futility of Mets pitchers at the plate--who set a new record with their collective 0-for-64 to start the season before one of them finally got a base hit--helps to reinforce the opinion of probably at least half of all baseball fans that nobody wants to see pitchers hit and it is way past time for the National League, too, to adopt the Designated Hitter. On the flip side of offense, this Insight looks at the relative difference between the two leagues in the impact on starting pitching--particularly complete games--since the DH was introduced in the American League.

The DH Impact on Managing Starting Pitching in the Two Leagues: A Brief History

Coming into play in 1973, the DH was the American League's response to batters having spent most of the previous ten years greatly overmatched by major league pitchers. Reaching its apex in 1968--the Year of the Pitcher, when seven qualifying pitchers had ERAs under 2.00 and seven pitchers won at least 20 games--major league baseball was emerging from a hitters' dark age, including by drastically lowering the pitcher's mound, but the AL was lagging behind the NL in the recovery. The three previous seasons, 1970 to 1972, AL teams averaged 7 percent fewer runs, hit .245 compared to the NL's .253, and had a league-wide slugging percentage 12 percentage points lower than the .374 mark put up by NL batters. The offensive difference between the two leagues had been most pronounced in 1972, with the National League scoring 13 percent more runs and hitting for a much higher batting average (.248 to .239) and slugging percentage (.365 to .343). AL pitchers' contributions at the plate amounted to a .145 batting average and .182 slugging percentage, almost exactly the same as how NL pitchers fared at the plate in 1972.

Introducing the DH made a big AL difference in the very first year. Offensive numbers were up substantially in both leagues in that year of 1973, but much more so in the American League. With designated hitters combining for more home runs than batters at any of the fielding positions, scoring was up in the American League by nearly 30 percent (compared in 11 percent in the NL); the league batting averaged increased by 20 percentage points in one year to .259; and the slugging percentage of AL batters was up nearly 40 points to .381. (NL batters collectively hit .254 with a .376 slugging percentage.)

The DH also had a profound and immediate effect on how managers used their pitching staffs. American League managers now had the option to keep their starting pitcher in the game for as long as he was effective, even if late in the game on the losing end of a pitchers' duel ... while in the National League, managers were still having to make the fraught decision about removing an effective pitcher--even their ace--for a pinch hitter late in a close game, especially if there were runners on base. The percentage of complete games in the American League jumped from less than 24 percent on average since 1960 to 32 percent in 1973, and stayed over 30 percent each of the next three years. Seven AL pitchers threw more than 300 innings in 1973, as would seven the year following. Even in the 'teens decade of the dead ball era, when more than half of all starts were complete games, neither league had that many pitchers throw 300 innings in a single season. In the National League, meanwhile, the percentage of complete games dropped from 28 percent the two previous years to 23 percent in 1973 and stayed at about that level each of the next three years. Only three NL pitchers threw 300 innings between 1973 and 1976, compared to 19 in the American League.

While the percentages began a steady decline in both leagues in the middle-late 1970s, the American League held a distinct advantage in complete game starts over the National League until the end of the 1980s--21 percent to 14 per cent in 1983, a decade into the DH era, and 16 percent to 10 percent in 1987, for example. Since the end of the mass disruption caused by the players' strike/owners' lockout in 1994-95, the two leagues have been in virtual sync in percentage of complete games even though, because of the DH, American League managers do not have to take into account the pitcher's spot in the batting order when contemplating a pitching change, while NL managers do. In 1996, AL pitchers completed 7 percent of their starts, compared to 6 percent in the NL; in 2000, both leagues completed 5 percent of their starts; and the American League has been between 3 and 4 percent and the National League between 2 and 3 percent since we entered the 21st century.

The unprecedented emphasis on pitch counts, specialization in the bullpen (and not just with regard to closers) and situational pitching changes has rendered the DH mostly irrelevant when it comes to starting pitchers going the distance. Except for the first ten years of the DH when the American League had a decided advantage, the percentage of complete game victories has been more or less equal, with the National League often marginally ahead of the AL in that stat.

Having a DH, however, probably has made a difference in the number and frequency of pitching changes. NL managers have consistently used more pitchers in non-complete games than their AL counterparts: NL managers made an average of 2.1 pitching changes per non-complete game start in the first ten years of the DH era, compared to 1.77 in the DH league; 2.56 compared to the AL's 2.47 in the 1990s; and 3.01 to 2.78 in the AL since 2000. National League teams have averaged more than three relievers per non-complete game (3.12, to be precise) in each of the last eight years, while the AL has reached the three-reliever threshold just once--in 2012. Perhaps somewhat surprisingly, starting pitchers averaged about the same ratio of innings-per-start most seasons in both leagues, with or without the DH.

The more significant indicator on the pitching staff-management impact of the DH is in the percentage of complete games that were losses. That number had been relatively stable at between 12 and 15 percent in both leagues since the beginning of the expansion era in 1961 and remained at that level in the National League through the 1970s. With the advantage of the starting pitcher not having to bat--so managers' decisions on pitching changes were based entirely on their starting pitchers' effectiveness and the game situation with the other team at bat--between 30 and 33 percent of complete games in the American League were losses in the first decade after the DH went into effect. The AL percentage of complete games losses has generally hovered between 25 and 30 percent since the mid-1980s.  Last year, complete games accounted for 4 percent of total victories in both leagues, but while 83 percent of NL complete games were in the "win column" (to quote Orioles radio play-by-play announcer Joe Angel), that figure was only 72 percent in the American League--meaning that having the DH allows AL managers to let starters pitching very effectively stay in the game to the very end, even if on the wrong side of the score, in hopes that a late rally would reward his efforts at keeping the game close with a victory.

The NL manager without the DH?  Well, much as he might prefer to keep an effective starting pitcher still within his pitch count in a tight game, the time inevitably comes in the late innings when he has to decide whether to pinch hit for his offensively-deficient pitcher in hopes of scoring the runs needed to tie or for victory.  And that, many baseball fans will argue, is reason enough for MLB not to go the designated hitter route in the National League.

Wednesday, May 14, 2014

The Offensive Efficiency Paradox of the Hitless (Punchless) Wonders

Until their offensive rampage against the Yankees in their two games in the Bronx this week, the Mets had been struggling to score runs even when getting runners on base.  A major reason why: the second-lowest total bases in major league baseball.  Brings to mind the hidden reality of the team best known in history as the "Hitless Wonders"--the 1906 Chicago White Sox--who might better have been called the "punchless wonders."  

The Offensive Efficiency Paradox: of the  Hitless (Punchless) Wonders

Once upon a time--oh, more than a hundred years ago to be precise, back when there be spitters and dentists in the professional game--the Chicago White Sox shocked the baseball world by winning a relatively close three-team pennant race despite finishing dead last in batting in the American League with a .230 average and hitting only seven home runs, and then in the World Series, by golly, taking out the overwhelmingly favored 116-win cross-town Cubs, who had the most formidable offense in baseball and the stingiest pitching besides. The '06 White Sox were in the middle of a five-year stretch-- 1904 to 1908--when they not only had the best overall record in the American League that averaged to 91-63, or five games better than second-best Cleveland's 86-68 average season during those years, but also the league's best record over any five-year period for the entire decade, by one game over 1901-05 Boston and two over 1906-10 Detroit.  Unlike their pennant-race rivals in any one of those seasons, the White Sox were competitive every year from 1904 to 1908, twice finishing within two games of the top. But they won only the one pennant.

They were blessed, of course, by terrific pitching that included a Hall of Fame spitballer (Big Ed Walsh) and the aforementioned dentist (Doc White, who had off-season patients), but were below the league average in batting every year.  After winning the pennant despite their league-low batting average in 1906, the "hitless wonders" lived up to that appellation each of the next two years by being legitimately competitive with next-to-lowest batting averages of .238 in 1907 and .224 in 1908.  From '06 to '08, Chicago averaged only 7.4 hits per game, compared to 8.7 per game by the six other teams that finished first, second or third.

Well, maybe the ChiSox weren't quite the offensive lightweights their historical moniker makes them out to be.  They were fairly prolific at scoring runs: 3rd in 1904, 2nd in 1905, 3rd in 1906, and 3rd in 1907. Alas, they were fifth in scoring in 1908, undoubtedly undermined by their near-subterranean .224 batting average, which was certainly not helpful in a three-way pennant chase they did not lose until the final day of the season.  Their ratio of one run scored for every two hits from 1904 to 1908 was substantially better than the league-average ratio of 2.24 hits per run and better than the 2.17 hits for each run of the ten other teams that finished first, second or third in the standings during those seasons.  Or put another way, the 1904-08 White Sox scored 95 percent as many runs as those other ten teams averaged on 11 percent fewer hits.

How'd they do it?  They were selective and disciplined at the plate! leading the league in walks every year from '04 to '08. They sacrificed for the cause! leading the league in sacrifice bunts every year from '04 to '07, and probably in '08 too--a year the statistic was muddled because sacrifice flies were included in the same category as sac bunts--since they had only seven fewer sacrifice hits than Cleveland, which had a far more imposing offense. They stole if they had to! leading the league in 1904 and stealing the second or third-most bases each of the other years. So, yes, this was clearly an efficient team when it came to scoring.

Ah, but there is a paradox. Based on the proposition that base runners are currency and runs are profit, the White Sox were far less efficient in capitalizing on their number of runners on base. Their ratio of runs scored to runners earning their way on base by hit,walk or hit batter was not as good as those of their fellow contenders--the teams finishing first, second or third--particularly from 1906 to 1908; the ChiSox required 2.9 base runners for each run they scored, 7 percent more than the 2.7 base runners per run for the six other contending teams those years.  This was at the same time they scored 89 percent as many runs as their rivals from '06 to '08 on 14 percent fewer hits.

The reason for this paradoxical discrepancy between two indicators of offensive efficiency--runs scored to hits versus runs scored to total base runners--is simple really, and can be summarized by the words "total bases."  A whopping 83 percent of the White Sox' hits from 1906 to 1908 were singles. The 17 percent of their total hits that went for extra bases was appreciably below the 21 percent extra base hits averaged by the AL's six other contending teams during those years.  The White Sox averaged 149 doubles, 42 triples and only 5 home runs those three years, substantially fewer in all categories than the 199 doubles, 69 triples and 16.5 home runs annual average of their pennant-race rivals.

Extra-base hits, after all,  have much bigger impact than singles because they both set up and score more runs. Chicago's rivals for the pennant scored more of their total base runners because they averaged 24 percent more total bases on their hits.  It is perhaps remarkable that the White Sox scored as many runs as they did--without which they would not have been in the pennant hunt every year even with their exceptional pitching--for as relatively few total bases they had.  The White Sox averaged 11 percent fewer total bases for each run they scored than the AL's six other first, second and third place teams from 1906 to 1908.

But the broader point is that even in the dead ball era, before Babe Ruth introduced the home run as its own offensive strategy, extra-base hits provided the most solid foundation for scoring runs.  Without that foundation, the White Sox' "no extra-base offense" had to work harder to score runs through singles and walks and sacrifice bunts and stolen bases and well-timed hits with runners in scoring position, requiring more runners on base (ultimately) for each run scored than teams that had more (doubles and triples) power in their line-up. They were not so much the "hitless wonders" as the "punchless wonders."

Or, to paraphrase Earl Weaver from a much later generation, White Sox' player-manager Fielder Jones would have loved the three-run double or triple. With more punch in their line-up, the White Sox--who finished six games back in 1904, two off the pace in 1905, third and 5-1/2 back in 1907 and 1-1/2 behind in their third-place ending to the 1908 season--might have had more than just their 1906 championship to show for having the best record in the American League from 1904 to 1908 and, for that matter, the entire first decade of the junior circuit's history.