This article was written by Timothy Mulligan
This article was published in the 1986 Baseball Research Journal
Teams that blow a 3-1 lead in the World Series or LCS are twice as likely to lose as to win. Star pitchers frequently are less effective in Game 7 than in their first two starts.
There can be little doubt that the outstanding feature of the 1985 post-season playoffs was the Kansas City Royals’ ability to overcome 3-1 deficits in both the American League Championship Series and the World Series. Considering that only four other teams have achieved that feat in World Series history, the Royals’ achievement appears monumental. But was it?
Our perspective changes if we consider the question: What is the final record for teams that lose a 3-1 advantage in the World Series or LCS and are forced to play a seventh game? Including the 1986 playoffs, ten teams have been in that position. The following went on to win the final game:
- 1912 Boston Red Sox
- 1967 St. Louis Cardinals
- 1972 Oakland A’s
The remainder lost the final game and the series:
- 1925 Washington Senators
- 1958 Milwaukee Braves
- 1968 St Louis Cardinals
- 1979 Baltimore Orioles
- 1985 Toronto Blue Jays (ALCS)
- 1985 St. Louis Cardinals (WS)
- 1986 California Angels (ALCS)
The record shows, then, that a championship-caliber team that allowed its opponents to tie the series by winning the fifth and sixth games was twice as likely to lose the seventh game as to win it. This appears more remarkable than the Royals’ twin comebacks and certainly is worthy of a closer look.
Many would attribute this phenomenon to a “shift of momentum” from one team to the other. Psychological factors may indeed play a role because players who see a championship slipping out of their grasp begin pressing while their opponents, considered all but eliminated already, seem to gain renewed confidence. Such factors cannot, of course, be broken down and analyzed. But one factor of at least equivalent significance can be studied: pitching.
A maxim of post-season pitching strategy has always held that a team’s best starter should open the series so that he will be in position to start three of a possible seven games. The application of that theory, however, did not save five of the seven teams whose 3-1 advantage ended with their best pitcher losing the seventh game. Toronto’s Dave Stieb and St. Louis’ John Tudor are the most recent examples; they joined the distinguished company of Bob Gibson (1968 St. Louis), Lew Burdette (1958 Milwaukee Braves), and Walter Johnson (1925 Washington).
Even the 1912 Series, when a tie led to an eight-game set, was determined in a similar fashion. Boston’s ace, Smokey Joe Wood, started Games 1, 4 and 7 and won his first two appearances impressively; in Game 7, however, he yielded seven hits and six runs in the first inning alone. But because an eighth game was required, Wood returned the next day in relief of Hugh Bedient to record the decisive win with only three innings. The losing pitcher in that game was the New York Giants’ Christy Mathewson, making his third start of the series.
We might hypothesize, then, that three effective starts from even your best pitcher in a seven-game series is too much to expect. Ten series hardly provide enough data to establish a relationship.
What we can examine are the pitching records for all three-game starters in seven-game post-season series. The data for those series should reveal whether any patterns of pitching effectiveness emerge.
First, let us compare pitching performances in the first two starts against the third start. Using data for the 39 instances when a pitcher started three times in a seven-game World Series and American League Championship Series, we find the following:
Two Starts Third Start
Innings Pitched 576⅔ 251½
Avg. 1P/Start 7⅓ 6⅓
Runs 177 116
Earned Runs 151 102
ERA 2.36 3.65
Won-Lost 49-22 16-16
Won-Lost Pct. .690 .500
Without question, pitchers’ effectiveness declined sharply after two starts. Three starts in eight or nine days – a strain under any condition and added to the toll taken by the innings worked in the regular season and to the pressure of pitching against championship opposition – would seem to produce tired pitchers. Fatigued arms result in reduced velocity and control, while opposing batters gain the advantage of increased familiarity with a pitcher’s strengths and weaknesses. What some see as the ultimate in sports “pressure” may be in reality the ultimate in sports endurance.
But are there circumstances that affect this performance? Can we further analyze the third starts to see when pitchers were more effective, and what might have made them so? One possible key would be relative effectiveness in the first two starts as a factor in third-start performance: If the pitcher was ineffective in one or both of his first two starts, would he be more or less effective in his third start than those who had pitched well in both of their previous games?
If we define an “ineffective” start as one in which the pitcher had an ERA over 3.00 and did not win the decision, we find that 11 of the 39 pitchers experienced disappointment in at least one of their first two games. How these pitchers fared in their third starts is indicated by the following cumulative data:
Innings Pitched 63⅔
Avg. IP/Start 5⅔
Earned Runs 22
Won-Lost Pct. .667
Twenty-eight of the 39 pitchers in our study demonstrated their abilities by being effective in both of their first two starts. Their record in the third start, however, reveals the difficulty in maintaining that level:
Innings Pitched 187⅔
Avg. IP/Start 6¼
Earned Runs 81
Won-Lost Pct. .435
The evidence therefore indicates that a pitcher with at least one ineffective outing in his first two games was more likely to pitch well in the third start, though ironically he would be yanked more quickly than his counterpart who had finally faltered. Of the seven teams that lost the seventh game after building a 3-1 edge in the series, four lost with their best pitcher after he had already pitched well in two previous starts (Walter Johnson, Bob Gibson, Dave Stieb and John Tudor). A fifth, the 1958 Milwaukee Braves, lost when their ace, Lew Burdette, experienced poor outings in both Games 5 and 7.
It is interesting to note the trend over the past 20 years toward deeper starting staffs among the top teams. In the 1958, 1962 and 1965 World Series, three pitchers started three games apiece; the 1964, 1967 and 1968 Series each featured two pitchers starting three games. The rules changes and practices introduced after 1968 to reduce the pitchers’ advantage (lowered mound, reduced strike zone, umpires’ increasing tendency to give the inside part of the plate to the batter) undoubtedly restored offensive punch to baseball, but probably also contributed (especially with the introduction of the designated hitter in the American League) to deeper starting staffs and more rest between starts. Of the six seven-game World Series played between 1971 and 1982, only two included three starts by one pitcher. The 1985 Kansas City Royals continued this trend by using four starters to defeat Toronto and St. Louis, both of whom attempted to rely on a weary ace to win a third start. As we have seen, the odds were not with them.