Series Sweeps in Competitive Situations:
The Line between Can and Can’t
Brian J. O’Leary, Holly A. Vande Walle,
and Bart L. Weathington
The University of Tennessee at Chattanooga
A review of the results of seven-game playoff series in Major League Baseball (MLB), the National Basketball Association (NBA) and the National Hockey League (NHL), indicates that the outcome is all but determined when one of the teams loses the first three games. It also appears that teams that lose the first three games are likely to lose game four, resulting in a “series sweep,” at a rate much higher than would be expected by chance. The following study examined archival data on 222 seven-game finals series in MLB, NBA and NHL held between 1905 and 2005.The data were collected from a variety of sources, including www.mlb.com, www.nba.com, www.nhl.com and www.whowins.com. The results of our analysis supported the hypothesized statistical anomaly. We propose theoretical bases for this effect and provide suggestions for future research aimed at developing individual- and team-level interventions to mitigate the negative outcome.
With the Houston Astros down 3-0 to the Chicago White Sox in the 2005 World Series, sportscasters announced that, of the 21 previous teams with such a deficit in the World Series, 18 had been swept in four games, the remaining three losing in five. The Astros subsequently succumbed to the odds by losing in game four. In the 2006 NBA Eastern Conference finals, the Detroit Pistons were down 3-1 to the Miami Heat. ESPN reported that 40 of the previous 43 teams who found themselves in the same situation as Detroit had ultimately lost the playoff series. Detroit continued this trend by succumbing in the next game. The same thing occurred to another Detroit team, the Tigers, in the 2006 World Series.
Why are teams unable to come back from a 3-0 deficit? The objective is to win four out of seven games; not an unrealistic goal if the team is in a championship series. During the year, it is likely that the team’s winning percentage was significantly higher than that required to win the series (.571 in a seven game series). And the records indicate that many teams have come back from a 1-0 or 2-0 deficit to win the World Series (www.mlb.com) or championship series in the NBA (www.nba.com) and the NHL (www.nhl.com).
For most professional sports teams entering a best of seven playoff series, winning four out of seven games is a reasonable goal. Relatively little changes if a team loses its first game. Winning four out of six games generally appears to be a very reasonable goal. With each successive loss, however, the apparent difficulty of the goal increases, escalating the possibility of failure, possibly taxing cognitive resources and creating the potential for increased conflict within the team. Psychology provides a number of theories, at both the individual and team levels of analysis, that could help illuminate this pattern. Among the most obvious would be theories of motivation such as goal-setting theory (Locke & Latham, 2002), control theory (Campion & Lord, 1982), expectancy theory (Vroom, 1964) and social cognitive theory (Bandura, 1986). In the next section, we provide a brief review of the relevant motivational theories to support our belief that motivation is significantly altered, both in form and intensity, when a professional sports team goes down three games to none in a seven game playoff series.
Theories of Motivation and Series Performance
The theories of motivation identified above suggest that self-efficacy (Bandura, 1986), expectancy (Vroom, 1964), and goal setting (Locke & Latham, 2002, 2004; Meyer, Becker & Vandenberghe, 2004) may play a role in the psychology of the series sweep. Self-efficacy is an individual’s belief in his/her ability to perform in a given situation through mobilization of motivation, implementation of appropriate strategies and summoning required cognitive resources. Efficacy beliefs are positively correlated with performance feedback: negative feedback, such as losing a game, can reduce the level of subsequent goals and decrease persistence (Wood & Bandura, 1989). According to expectancy theory (see Vroom, 1964), individuals put forth effort if they expect resulting behaviors to lead to the instrumental outcomes that are required to achieve desired outcomes. For example, winning a game (the instrumental outcome) is necessary in order to obtain the fame, fortune, and satisfaction associated with winning the series. Expectancies about desired outcomes should change, either positively or negatively, with the results of the instrumental outcomes.
As proposed in control theory (Campion & Lord, 1982; Carver & Scheier, 1998), team members have the ability to detect the discrepancy between their present state (e.g., entering the World Series) and desired state (winning four out of seven games) through a continuous self-regulating feedback system. Similar to expectancy theory, control theory accounts for the different types of goals that teams have, which Campion and Lord labeled superordinate and subordinate goals. The superordinate goal for teams entering a championship series is to win the Series. Subordinate goals are those goals set for each individual game. In a best of seven competition, achieving four subordinate goals before four unsuccessful attempts results in the achievement of the superordinate goal. From game to game, however, the team’s immediate and distal past performance impacts future goals. Consequently, the winning team members perceive no goal-performance discrepancy and continue their drive toward the superordinate goal. Conversely, members of the losing team recognize a discrepancy that indicates a need to adapt their behavior to ensure that subsequent subordinate goals can be successfully attained in support of the superordinate goal of winning the Series.
Carver, Sutton, and Scheier (2000) extended this self-regulative perspective by integrating ideas from self-discrepancy theory (Higgins, 1987, 1996) and the literature on affectivity (e.g., Watson, Clark & Tellegen, 1988). In addition to the discrepancy-reducing loops identified in earlier conceptualizations of control theory, Carver et al. proposed the existence of discrepancy-increasing loops. Rather than focusing on attaining the goal, as occurs in a discrepancy-reducing loop, a discrepancy-increasing loop redirects the focus away from what they termed an “anti-goal.” This transition from a discrepancy-reducing to a discrepancy-increasing loop reflects a change from an approach orientation to an avoidance orientation, each of which has a different affective component. An approach orientation is generally related to positive affectivity while an avoidance orientation tends toward negative affectivity. The differences in orientation are best exemplified by the affective reactions to positive outcomes: success in an approach orientation (e.g., winning) results in feelings of elation and accomplishment; whereas, an avoidance orientation, where success is defined in terms of not failing (e.g., losing), is accompanied by feelings of relief. In the context of a playoff series, it is possible that goal orientation can shift from approach to avoidance as uncertainty about attaining the superordinate goal of winning the series increases with successive failures to achieve the subordinate goal of winning an individual game.
According to goal setting theory, commitment to a specific, difficult goal increases performance (Locke & Latham, 1990). Goal commitment has traditionally been treated as a monolithic construct that either exists or not (Klein, Wesson, Hollenbeck, & Alge, 1999). However, researchers in the area of organizational commitment have found evidence that there are various types of commitment, and that these different types of commitment have different antecedents and different consequences (Meyer & Allen, 1991; Meyer, Allen, & Smith, 1993). Meyer et al. (2004) recently applied this multidimensional conceptualization to goal commitment. Affective goal commitment reflects an individual’s affective attachment to or the intrinsic value of the goal. Normative goal commitment captures a sense of obligation to perform, while continuance goal commitment focuses on the costs associated with failure to achieve the goal. In the context of the current study then, affective goal commitment could promote a desire to win the series for personal satisfaction, while normative commitment may shift the focus onto winning the series to meet the expectations of others, and continuance commitment might encourage individuals to focus on avoiding losses.
Wicker, Turner, Reed, McCann, and Do (2004) identified a temporal basis for a decrease in “unrealistic optimism,” suggesting that individuals predict their competence less optimistically, and possibly more realistically, as they approach points where they will be held more accountable for their productivity or achievements (Tetlock, 1992). Specifically, Wicker et al. hypothesized and found that predictions of performance were less optimistic as events approached, resulting in a decrease in event-related motivation due to an increased awareness of the potential for failure. For example, at the beginning of a playoff series, team members may have high expectancies and possibly unrealistic optimism about winning four out of seven games, even possibly sweeping the series. However, negative feedback, such as losing a game, increases perceptions of accountability and reduces expectations that the superordinate goal can be achieved. While losses in games one and two do not appear to significantly reduce expectations of success, after a loss in game three the cumulative effect of these losses on expectations of winning may reach a critical mass that most individuals and teams cannot overcome.
Tenenbaum, Lidor, Lavyan, Morrow, Tonnel, and Gershgoren (2005) studied the effects of self-efficacy on effort in physically demanding tasks. Prior researchers have indicated that the level of perseverance an individual exhibited in achieving a goal was determined by a combination of goal orientation (i.e., task- or ego-oriented) and task specific factors such as competence, determination and commitment. Tenenbaum et al. found that task-oriented individuals increased their competence by exerting more effort and adapting to the outcomes of past games (i.e., past subordinate goals). These individuals remained focused on superordinate goals regardless of aversive environmental feedback, such as a losing streak and negative publicity. Ego-oriented individuals, on the other hand, were highly sensitive to situations that might hurt their egos. Therefore, when faced with a losing streak, individuals on teams that shared this mentality reduced their efforts toward subordinate goals because they wished to protect their feelings of self-competence.
McClelland (1980) studied goal commitment and acceptance, and the results are somewhat parallel to the goal-orientation literature. Those high in need for achievement set only moderately difficult goals for themselves because they dislike failure (similar to ego-oriented individuals). The motivational value of the goal is that it is achievable with the exertion of a reasonable amount of effort. On the basis of the expectancy value theories of motivation (e.g., Tolman, 1955; Vroom, 1964), it appears that, if expectancies decrease as the likelihood of failure increases, then motivation would also decrease as losses mount. To keep goals attainable and within the realistic range of achievement, individuals high in need for achievement must reduce their goal levels when their expectancies decline (Wicker et al., 2004). This process may also affect goal commitment (Locke & Latham, 1990) or possibly cause goal commitment to change from affective through normative to continuance commitment (Meyer & Allen, 1991; Meyer et al., 2004).
Applying the Theories
On the basis of the theories and concepts reviewed above, it appears that it is reasonable to believe that there are identifiable and measurable factors that could be used to address the statistical anomaly that the seven game series sweep represents. For example, motivation to win could be reduced by lowered self-efficacy and expectations of winning (e.g., expectancy theory), by the perception that the goal of winning the series has become too difficult (e.g., need for achievement), or because the goal has become less intrinsically satisfying and more burdensome (e.g., cognitive evaluation theory). As the first step in this process, however, we must determine that there is a statistical difference in the actual series outcomes associated with losing only the first and second games versus losing the first three games. The odds of winning or losing the first three games in a series should be equal. However, we hypothesize that professional sports teams who lose the first three games in a row will be more likely to lose the fourth game than would be expected due to chance.
The current study used archival data for three professional sports leagues using a seven game playoff format: Major League Baseball (MLB), the National Basketball Association (NBA), and the National Hockey League (NHL). These data were collected from a variety of online sources, including www.mlb.com, www.nba.com, www.nhl.com and www.whowins.com. The data covered all years in which each league used a seven game playoff format. We summarized the results of every playoff series for each of the three leagues in an database that included the names of the teams involved in each series by year, and the results of each game of the series (won or lost). To take into account skill differences between teams due to their seeding in the earlier rounds (e.g., a 1 seed playing an 8 seed), only data from the final round of playoff games were included in the analyses. For consistency, five game series, as used for the Stanley Cup in the NHL prior to 1939, as well as the nine game championship series used sporadically by MLB in the early 1900s, were excluded from the analysis. There was a possibility that the data taken from multiple sources for a given sport would disagree, so we cross-checked between sources to verify their accuracy and found no discrepancies. In total, 222 best of seven-game playoff finals were played between 1905 and 2005 (MLB = 97, NBA = 59, and NHL = 66). We performed a series of ratio analyses to identify the winning percentages for the losing teams in each of the subsequent games, and then performed a chi-square analysis to determine whether these percentages changed significantly from game to game.
Testing the hypothesis depends on demonstrating that the proportion of teams that won or lost the next game in a best of seven series differs for teams that lost the first three games in a row versus those that only lost the first one or two games.
All 222 MLB, NBA, and NHL series included in the analysis resulted in a clear winner or loser – no tie games. Of the 222 teams that lost game one, 106 won game two (106/222 = 47.7%). One hundred and sixteen teams lost games one and two. Of these, 58 teams won game three (58/116 = 50%). Of the 58 teams that lost the first three playoff games, 12 won game four (12/58 = 20.7%). One of these teams managed to come back and win the series (1/58 = 1.7%).
Supporting the hypothesis, a chi-square test including only results from games one and two provided a non-significant outcome, χ2 (1, N = 338) = 0.15, p >.05; see Table 1).
The inclusion of the outcome of game three with the previous results revealed a significant difference between the odds of winning the next game with three versus one or two losses, χ2 (2, N = 396) = 15.69, p < .01 (see Table 2).
Our results indicate that something occurs that changes the dynamics of the series as losses mount, particularly between losses in the first two games and losses in the first three. There are several possible explanations and opportunities for research. One explanation may be the occurrence of a strategic shift between games two and three wherein a manager or coach pulls out all the stops to avoid the situation where the team is losing the series three games to none. If the third games is then lost, any rhythm that the team may have developed in the earlier rounds of the playoffs could also be lost. Further, resources that would have been held in reserve under less onerous circumstances may be expended and, as a result, be unavailable or rendered less effective for future use. For example, in baseball the best pitcher on the team may be used on two, rather than three days rest, which suggests he may not be well rested and therefore prone to injury or mistakes. This move also changes the pitching rotation for the rest of the series, making it more likely that a lower quality pitcher will be available for game four. In basketball or hockey, starters may play longer in game three and be less rested for game four. Conversely, teams may fail to adapt due to what Staw (1981) referred to as escalation of commitment to a course of action. Athletes and coaches may fail to make necessary adjustments due to their belief that they have chosen the proper course, despite indications that their current approach to the series is ineffective.
A review of current textbooks dealing with sport psychology (Cox, 2007; Gill, 2000; Horn, 2002; LeUnes & Nation, 2002; Weinberg & Gould, 2003; Williams, 2006) reveals that the importance of motivation on individual and group performance is well recognized. However, what is not understood is why some teams come back from what on the surface appears to be an insurmountable deficit (e.g., the Toronto Maple Leafs in 1942 and New York Islanders in 1975 both came back from 0-3 deficits to win the series). The theories discussed in this paper in conjunction with issues of resource utilization and allocation can help to explain this phenomenon. For example, goal setting theory (Locke & Latham, 1990) is arguably the most heavily researched and utilized theory of motivation in applied sport psychology and it is recognized that specific, moderately difficult goals have the most impact on performance. It may be that, when down three games in a seven game series, the ability to set goals of this nature is severely limited. It may become almost impossible to remove thoughts of the overall goal (win the series) from immediate objectives (score points and win the game).
There may also be contextual variables, such as the type of interdependence required (Thompson, 1967), that influence the nature of the required intervention. While all three of the team sports discussed in this paper require what Thompson described as “reciprocal” interdependence, that is, team members must coordinate their actions based on no predetermined sequence to maximize performance, baseball lacks the time pressure and continuous action that characterize basketball and hockey. As a result, organizational consultants to baseball teams may need to focus interventions more on the individual, whereas, interventions designed for hockey and basketball may need to be more team-oriented.
As to future research needs to focus on both the individual and team levels. At the individual level, a simple game could be employed to set up a series of competitions, either between individuals or against a computer, in which a seven game series could be completed. Questions related to self-efficacy, competence, expectations of winning, both in terms of magnitude and strength (e.g., Locke, Frederick, Lee, & Bobko, 1984) and goal level could be administered between games. Measures of individual level personality variables, such as extraversion, openness to experience and locus of control, may also provide insight as moderators of the relationship between mounting losses and subsequent performance.
At a group level, in addition to laboratory investigations, there are a variety of opportunities for longitudinal research on playoff series of various lengths at both a professional and amateur level. In addition to addressing the effects of the individual differences identified above, researchers could also examine team-level constructs such as collective or team efficacy (Bandura, 1997; Knight, Durham, & Locke, 2001), team ability (Bandura, 1986) and team cohesion (Dion, 2000). A sports team is more than just a collection of individual athletes and a continued awareness of the interaction between teams and individuals is essential in producing meaningful research that can be applied to sport and athletic performance in the real world.
Implications for Practice
Having demonstrated that an effect does indeed exist, it is now necessary to identify those controllable factors that can be used to design and implement interventions that can blur the line between can and can’t. We have suggested some possible contributors to the “sweep” phenomenon, such as self-efficacy, expectancy, and accountability. However, the practical implications of our current findings are contingent upon an empirical examination of the correlations between the variables we have proposed and actual game four performance. These future studies can be performed in both laboratory and field settings for both individual and team competitions. For example, an individual level laboratory study could include a short series of games in which we would measure the variables of interest (e.g., self-efficacy, expectancies of winning, form of motivation) before and after each game and correlate any changes in these variables with game outcomes. In the field, team members involved in a series competition (e.g., volleyball, fencing) could be measured on the variables of interest before and after each game or match which would then be correlated with the results of subsequent games. Validating these causal relationships will assist the practitioner in developing interventions that mitigate the negative effects of mounting losses on individual and team performance.
While rare, teams have come back from the abyss; furthering our understanding of the psychological factors that impact the losers, and the winners, at both the individual and team levels may make it possible to train coaches and athletes to push the “broom” aside more frequently.
Bandura A. (1986). Social foundations of thought and action: A social cognitive theory. Englewood Cliffs, NJ: Prentice Hall.
Bandura, A. (1997). Self-efficacy: The exercise of control. New York: W. H. Freeman.
Bandura, A., & Wood, R. (1989). Effect of perceived controllability and performance standards on self-regulation of complex decision making. Journal of Personality and Social Psychology, 56, 805-814.
Campion, M. A., & Lord, R. G. (1982). A control systems conceptualization of the goal setting and changing process. Organizational Behavior and Human Performance, 30, 265-267.
Carver, C. S., & Scheier, M. F. (1998). On the self-regulation of behavior. New York: Cambridge University Press.
Carver, C. S., Sutton, S. K., & Scheier, M. F. (2000). Action, emotion, and personality: Emerging conceptual integration. Personality and Social Psychology Bulletin, 26, 741-751.
Cox, R. H. (2007). Sport psychology: Concepts and applications (6th ed.). New York: McGraw-Hill.
Dion, K. L. (2000). Group cohesion: From “field of forces” to multidimensional construct. Group Dynamics: Theory, Research, and Practice, 4, 7-26.
Gill, D. L. (2000). Psychological dynamics of sport and exercise (2nd ed.). Champaign, IL: Human Kinetics.
Higgins, E. T. (1987). Self discrepancy: A theory relating self and affect. Psychological Review, 94, 319-340.
Higgins E. T. (1996). Ideals, oughts, and regulatory focus: Affect and motivation from distinct pains and pleasures. In P. M. Gollwitzer & J. A. Bargh (Eds.), The psychology of action: Linking cognition and motivation to behavior (pp. 91-114). New York: Guilford Press.
Horn, T. S. (2002). Advances in sport psychology (2nd ed., pp.101-181). Champaign, IL: Human Kinetics.
Klein, H. J., Wesson, M. J., Hollenbeck, J. R., & Alge, B. J. (1999). Goal commitment and the goal-setting process: Conceptual clarification and empirical synthesis. Journal of Applied Psychology, 84, 885-896.
Knight, D., Durham, C. C., & Locke, E. A. (2001). The relationship of team goals, incentives, and efficacy to strategic risk, tactical implementation, and performance. Academy of Management Journal, 44, 326-338.
LeUnes, A., & Nation, J. R. (2002). Sport Psychology (3rd ed.). Pacific Grove: Wadsworth.
Locke, E. A., Frederick, E, Lee, C., & Bobko, P. (1984). Effect of self-efficacy, goals, and task strategies on task performance. Journal of Applied Psychology, 69, 241-251.
Locke, E. A., & Latham, G. P. (1990). A theory of goal setting and task performance. Englewood Cliffs, NJ: Prentice Hall.
Locke, E. A., & Latham, G. P. (2002). Building a practically useful theory of goal setting and task motivation: A 35-year odyssey. American Psychologist, 57, 705-717.
Locke, E. A., & Latham, G. P. (2004). What should we do about motivation theory? Six recommendations for the twenty-first century. Academy of Management Review, 29, 388–403.
McClelland, D. C. (1980). Motive dispositions: The merits of operant and respondent measures. In L. Wheeler (Ed.), Review of personality and social psychology (Vol. 1). Beverly Hills, CA: Sage Publications.
Meyer, J. P., & Allen, N. J. (1991). A three-component conceptualization of organizational commitment. Human Resource Management Review, 1, 61–89.
Meyer, J. P., Allen, N. J., & Smith, C. A. (1993). Commitment to organizations and occupations: Extension and test of a three-component conceptualization. Journal of Applied Psychology, 78, 538-551.
Meyer, J. P., Becker, T. E., & Vandenberghe, C. (2004). Employee commitment and motivation: A conceptual analysis and integrative model. Journal of Applied Psychology, 89, 991–1007.
Staw, B. M. (1981). Escalation of commitment to a course of action. Academy of Management Review, 6, 577-587.
Tenenbaum, G., Lidor, R., Lavyan, N., Morrow, K., Tonel, S., & Gershgoren, A. (2005). Dispositional and task-specific social-cognitive determinants of physical effort perseverance. The Journal of Psychology, 139, 139-157.
Tetlock, P. E. (1992). The impact of accountability on judgment and choice: Toward a social contingency model. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 25, pp. 331-376). San Diego, CA: Academic Press.
Thompson, J. D. (1967). Organizations in action. New York: McGraw-Hill.
Tolman, E. C. (1955). Principles of performance. Psychological Review, 62, 315-326.
Vroom, V. H. (1964). Work and motivation. New York: Wiley.
Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures of positive and negative affect: The PANAS scales. Personality and Social Psychology Bulletin, 54, 1063-1070.
Weinberg, R. S., & Gould, D. (2003). Foundations of sport and exercise psychology (3rd ed.). Champaign, IL: Human Kinetics.
Wicker, F. W., Turner, J. E., Reed, J. H., McCann, E. J., & Do, S. L. (2004). Motivation when optimism declines: Data on temporal dynamics. The Journal of Psychology: Interdisciplinary and Applied, 138, 421-432.
Williams, J. M. (Ed.) (2006). Applied sport psychology: Personal growth and peak performance (5th ed., pp. 43-62). New York: McGraw-Hill.
Wood, R., & Bandura, A. (1989). Impact of Conceptions of Ability on Self-Regulatory Mechanisms and Complex Decision Making Journal of Personality and Social Psychology, 56, 407-415.