Content uploaded by Scott Madey
Author content
All content in this area was uploaded by Scott Madey
Content may be subject to copyright.
Journal of Personality and Social Psychology
Copyright 1995 by the American Psychological Association, Inc.
Volume 69(4) October 1995 p 603–610
When Less Is More: Counterfactual Thinking and Satisfaction Among
Olympic Medalists
[Attitudes and Social Cognition]
Medvec, Victoria Husted1; Madey, Scott F.2; Gilovich, Thomas1,3
1Department of Psychology, Cornell University
2Department of Psychology, University of Toledo.
3Correspondence concerning this article should be addressed to Thomas Gilovich, Department of Psychology, Uris
Hall, Cornell University, Ithaca, New York 14853-7601.
This research was supported in part by grants from the National Institute of Mental Health (MH45531) and the
National Science Foundation (SBR9319558).
We would like to thank Todd Bickford, Theresa Buckley, Nancy De Hart, Deborah Fidler, Nina Hattiangadi, Allison
Himmelfarb, Elena Jeffries, Danielle Kaplan, Talia Korenbrot, Renae Murphy, Sara Sirlin, and Shane Steele for their
help in editing the videotapes and collecting data.
Received Date: January 10, 1994; Revised Date: May 8, 1995; Accepted Date: May 10, 1995
Outline
zAbstract
zStudy 1
zMethod
zParticipants.
zStimulus materials.
zProcedure.
zResults
zDiscussion
zStudy 2
zMethod
zParticipants.
zStimulus materials.
zProcedure.
zResults
zDiscussion
zStudy 3
zMethod
zParticipants.
zProcedure.
zResults
zDiscussion
zGeneral Discussion
zReferences
Gra
p
hics
Pa
g
e 1 of 15Ovid: Medvec: J Pers Soc Ps
y
chol, Volume 69
(
4
)
.October 1995.603–610
1/8/2004htt
p
://
g
atewa
y
1.ma.ovid.com/ovidweb.c
g
i
zFigure 1
Abstract
Research on counterfactual thinking has shown that people's emotional responses to events are
influenced by their thoughts about “what might have been.” The authors extend these findings by
documenting a familiar occasion in which those who are objectively better off nonetheless feel
worse. In particular, an analysis of the emotional reactions of bronze and silver medalists at the
1992 Summer Olympics—both at the conclusion of their events and on the medal stand—
indicates that bronze medalists tend to be happier than silver medalists. The authors attribute
these results to the fact that the most compelling counterfactual alternative for the silver medalist
is winning the gold, whereas for the bronze medalist it is finishing without a medal. Support for
this interpretation was obtained from the 1992 Olympics and the 1994 Empire State Games. The
discussion focuses on the implications of endowment and contrast for well being.
So we have the paradox of a man shamed to death because he is only the second pugilist or the
second oarsman in the world. That he is able to beat the whole population of the globe minus one
is nothing; he has “pitted” himself to beat that one; and as long as he doesn't do that nothing else
counts. (James, 1892, p. 186)
James's (1892) observation represents an early statement of a fundamental principle of
psychology: A person's objective achievements often matter less than how those
accomplishments are subjectively construed. Being one of the best in the world can mean little if
it is coded not as a triumph over many, but as a loss to one. Being second best may not be as
gratifying as perhaps it should.
Since James's time, of course, this idea has been both theoretically enriched and extensively
documented. Social psychologists have shown that people's satisfaction with their objective
circumstances is greatly affected by how their own circumstances compare with those of relevant
others (Festinger, 1954; Suls & Miller, 1977; Taylor & Lobel, 1989). A 5% merit raise can be quite
exhilarating until one learns that the person down the hall received an 8% increase. Psychologists
have also demonstrated that satisfaction with an outcome likewise depends on how it compares
with a person's original expectations (Atkinson, 1964; Feather, 1967, 1969). Someone who receives a
5% raise might be happier than someone who receives an 8% increase if the former expected less
than the latter. Often it is the difference between the actual outcome and the expected outcome,
or the actual outcome and the outcomes of others, that is decisive (Crosby, 1976; Olson, Herman, &
Zanna, 1986).
More recently, psychologists have discovered a third way in which the determinants of
satisfaction are relative. In particular, people seem to be greatly affected by how their objective
outcomes compare to imagined outcomes that “might have been” (Kahneman & Miller, 1986;
Kahneman & Tversky, 1982b; Markman, Gavanski, Sherman, & McMullen, 1993; Miller, Turnbull, & McFarland,
1990; Roese, 1994; Roese & Olson, in press). The intensity of people's reactions to events appears to be
proportional to how easy it is to conjure up greater or lesser outcomes that “almost happened.”
An 8% return on one's investment might exceed expectations and yet be disappointing if one is
reminded of an alternative investment one “almost” made that yielded a substantially higher
return. The critical com
p
arison in this case is a
p
ostcom
p
uted res
p
onse to what has occurred,
Pa
g
e 2 of 15Ovid: Medvec: J Pers Soc Ps
y
chol, Volume 69
(
4
)
.October 1995.603–610
1/8/2004htt
p
://
g
atewa
y
1.ma.ovid.com/ovidweb.c
g
i
rather than a precomputed representation of what seems likely, ex ante, to occur (Kahneman &
Miller, 1986).
Most of the research on counterfactual thinking has held outcome constant and examined the
reactions of people contemplating different counterfactual alternatives. For example, Kahneman and
Tversky (1982b) asked their participants to imagine the reactions of two travellers who both missed
their scheduled flights, one by 5 minutes and the other by 30 minutes. The outcome is the
same—both must wait for the next flight—but it is easier to imagine a counterfactual world in
which the first traveller arrives on time. Studies such as this have repeatedly shown that the same
outcome can produce strikingly different reactions as a function of the ease of generating various
counterfactual alternatives (Johnson, 1986; Kahneman & Miller, 1986; Kahneman & Tversky, 1982a, 1982b;
Miller & McFarland, 1986; Miller et al., 1990; Turnbull, 1981; Wells & Gavanski, 1989).
We wished to take this a step further. We were interested in whether the effects of different
counterfactual comparisons are sufficiently strong to cause people who are objectively worse off
to sometimes feel better than those in a superior state. Moreover, we were interested not just in
documenting isolated episodes in which this might happen, but in identifying a specific situation
in which it occurs with regularity and predictability. The domain we chose to investigate was
athletic competition.
We chose this domain of investigation because in athletic competition outcomes are typically
defined with unusual precision. Someone finishes first, second, or third, for example, thereby
earning a gold, silver, or bronze medal. With all else equal, one would expect the athletes' levels
of satisfaction to mirror this objective order. We suspected, however, that all else is not equal—
that the nature of athletes' counterfactual thoughts might cause their levels of satisfaction to
depart from this simple, linear order.
Consider the counterfactual thoughts of bronze and silver medalists. What might their most
compelling counterfactual thoughts be? One would certainly expect the silver medalist to focus
on almost winning the gold because there is a qualitative difference between coming in first and
any other outcome. Each event has only one winner, and to that victor belongs the considerable
spoils that the modern commercial-athletic world bestows (R. H. Frank & Cook, 1995). Moreover, for
the silver medalist, this exalted status was only one step away. To be sure, the silver medalist
also finished only one step from winning a bronze, but such a downward social comparison does
not involve much of a change in status (i.e., neither the bronze nor silver medalist won the event,
but both won medals), and thus does not constitute as much of a counterfactual temptation.
In contrast, bronze medalists are likely to focus their counterfactual thoughts downward. Like
the qualitative jump between silver and gold, there is a categorical difference between finishing
third and finishing fourth. Third place merits a medal whereas the fourth-place finisher is just
one of the field. This type of categorical difference does not exist in the upward comparison
between second and third place.
Because of this asymmetry in the direction of counterfactual comparison, the person who is
objectively worse off (the bronze medalist) might nonetheless feel more gratified than the person
who is objectively better off (the silver medalist). Like William James's (1892) pugilist, silver
medalists may torment themselves with counterfactual thoughts of “if only …” or “why didn't I
just ….” Bronze medalists, in contrast, may be soothed by the thought that “at least I won a
medal.” The net result is that with respect to athletic competition, there may be times when less
is more.
Pa
g
e 3 of 15Ovid: Medvec: J Pers Soc Ps
y
chol, Volume 69
(
4
)
.October 1995.603–610
1/8/2004htt
p
://
g
atewa
y
1.ma.ovid.com/ovidweb.c
g
i
We conducted three studies to examine this question. First, we analyzed the affective reactions
of bronze and silver medalists as they won their medals in the 1992 Olympic games in
Barcelona, Spain. Second, we had participants evaluate the Olympians' postcompetition
interviews to see whether silver medalists seemed to be focused on the medal they almost won
whereas third-place finishers appeared to relish the pleasure simply of being medalists. In the
third study, we asked athletes themselves about the nature of their counterfactual thoughts.
Study 1
We videotaped all of the National Broadcasting Company (NBC) coverage of the 1992
Summer Olympic games in Barcelona, Spain. From this footage, two master tapes were
constructed. The first showed the immediate reactions of all bronze and silver medalists that
NBC chose to televise at the time the athletes learned how they had finished. Thus, the tape
shows Janet Evans as she touched the wall of the pool and discovered she had come in second,
and Jackie Joyner-Kersey after she completed her last long jump and earned a bronze medal. The
second tape consisted of all bronze and silver medalists whom NBC showed on the medal stand
during the award ceremony. For example, this tape shows Matt Biondi receiving a silver medal
for his performance in the 50-m freestyle, and the Lithuanian men's basketball team (in uniforms
designed by The Grateful Dead) after they received their bronze medals.
Each tape was shown to a separate group of participants who were asked to rate the expressed
emotion of each athlete. Because of the asymmetry in the likely counterfactual comparisons of
the bronze and silver medalists, we expected those who finished third to be demonstrably happier
than those who finished second.
Method
Participants.
Twenty Cornell University undergraduates served as participants. Only people who indicated
they were uninterested in and uninformed about sports were recruited. This ensured that their
ratings would not be affected by any preexisting knowledge about the athletes or their
performance in the Olympic games.
Stimulus materials.
The tape of the athletes' immediate reactions included shots of 23 silver and 18 bronze
medalists. Not surprisingly, given NBC's main audience, most of these shots (25) were of
Americans. To create the master tape, we simply copied all footage of the finish and immediate
aftermath of all silver and bronze medal winners. These shots tended to be rather brief (M = 14.4
s; SD = 8.3 s), and we stayed with the scene for as long as NBC did. Because the issue of what
footage to include involved minimal judgment, we did the editing ourselves.
This was not the case for the medal stand videotape. Here there were too many editing
decisions to be made. Should a shot of the athlete leaving the medal stand be included? Should a
certain “head and shoulders” shot on the medal stand be included or not? To eliminate the
possibility of our expectations guiding these editorial decisions, we turned the job over to
someone unaware of our hypothesis. We identified all medal stand shots of second- and third-
place finishers in NBC's coverage, and asked our editor to copy those moments that best
captured the emotion that the athletes appeared to be feeling. This resulted in a master tape of 20
silver and 15 bronze medal winners. The avera
g
e len
g
th of each shot was 14.7 s, with an SD of
Pa
g
e 4 of 15Ovid: Medvec: J Pers Soc Ps
y
chol, Volume 69
(
4
)
.October 1995.603–610
1/8/2004htt
p
://
g
atewa
y
1.ma.ovid.com/ovidweb.c
g
i
13.8 s. In this case fewer than half of the shots (15) were of American athletes.1
Two versions of each tape were created, with the order of presentation of the athletes varied
across versions. Blank spaces were inserted between shots of the different athletes to provide
participants with time to complete their ratings.
Procedure.
Participants arrived at the laboratory in groups and were told that they would be watching a
videotape of athletes from the 1992 Olympic games. They were informed that they were to rate
the expressed emotions of each athlete on a 10-point “agony to ecstasy” scale. The participants
were first asked to watch a few shots of athletes without making any ratings in order to give them
an idea of the range of emotions shown on the tapes. After participants were familiar with the
format of the videotape, the rating session commenced.
Five participants rated each version of each of the two videotapes. The tapes were shown
without sound to eliminate the chance that commentators' remarks might affect their evaluations
of the athletes' expressed emotions. A 1.5-inch (3.8 cm) strip of paper was affixed to the bottom
of the video screen to occlude various graphics used by NBC to indicate the athlete's order of
finish.
Results
Participants' ratings were highly reliable, for both the immediate-reactions videotape
(Spearman-Brown index = .97) and the medal stand tape (Spearman-Brown index = .96). Thus,
the ratings of all participants viewing the same tape were averaged to create an index of the
happiness of each of the athletes. Preliminary analyses revealed no effect of order of
presentation, so the data were collapsed across the two versions of each tape.
The mean happiness ratings are presented in Figure 1. As predicted, bronze medalists appeared
happier on average than their counterparts who won silver medals. When assessing the athletes'
immediate reactions, participants assigned the bronze medalists a mean happiness rating of 7.1
(SD = 2.0) but the silver medalists a mean rating of only 4.8 (SD = 1.9). When examining the
athletes on the medal stand, participants assigned the bronze medalists a mean rating of 5.7 (SD
= 1.7) and silver medalists a mean rating of only 4.3 (SD = 1.8). These data were analyzed with a
2 (type of medal: bronze vs. silver) × 2 (tape: immediate vs. medal stand) analysis of variance
(ANOVA). This analysis revealed two significant main effects, but no interaction. The main
effect of tape, F(1, 72) = 4.78, p < .05, indicates that the athletes on the whole looked happier
immediately after their performances than when they were on the medal stand. More important,
the main effect of type of medal, F(1, 72) = 18.98, p < .001, indicates that the athletes who
finished third looked significantly happier than those who finished second.
Pa
g
e 5 of 15Ovid: Medvec: J Pers Soc Ps
y
chol, Volume 69
(
4
)
.October 1995.603–610
1/8/2004htt
p
://
g
atewa
y
1.ma.ovid.com/ovidweb.c
g
i
Figure 1. Mean happiness ratings.
There is a potential artifactual explanation of these results, however. In certain Olympic
events, the competition is structured such that bronze medalists have just won a match or a game
whereas silver medalists have just lost. A bronze medalist in wrestling, for example, would have
just defeated the fourth place finisher, and the silver medalist would have just lost to the gold
medal winner. We were concerned that being in the immediate aftermath of victory or defeat
might have contaminated our comparison of bronze and silver medalists. Fortunately, most
Olympic events (such as those in track, swimming, and gymnastics) are not structured in this
way. In these events the athletes simply finish first, second, and third depending on how well
they do.
To eliminate this “just won”–“just lost” artifact, we reanalyzed the data excluding all athletes
involved in sports with this structure. This reduced our pool of 23 silver and 18 bronze medalists
in the immediate-reactions videotape to 20 and 15, respectively. Similarly, it reduced our pool of
20 silver and 15 bronze medalists in the medal-stand tape to 14 and 13, respectively. A 2 × 2
ANOVA of these data yielded the same significant main effect of type of medal as before, F(1,
58) = 6.70, p < .02. Bronze medalists appeared happier both immediately after their events (M =
6.7) and on the medal stand (M = 5.6) than their counterparts who had won silver medals (Ms =
5.0 and 4.7). Consistent with our thesis, impartial judges viewed bronze medalists as being
happier than silver medalists, and this effect was not limited to those few events in which bronze
and silver medalists were in the immediate aftermath of a victory or a defeat, respectively.2
Is there any other alternative interpretation of these data? Might these results be due to
differences in the ex ante expectations of bronze and silver medalists rather than—as we
propose—their ex post thoughts about what might have been? We think not. First of all, there is
no reason to believe that bronze medalists as a whole tended to exceed their expectations or that
silver medalists on average tended to fall short of theirs. To be sure, our sample of silver
medalists probably entered the Olympics with higher expectations on average than our sample of
bronze medalists, but they also performed better as well. There is certainly no compelling reason
to believe that one group over- or under-performed relative to their initial expectations.
This alternative interpretation can also be dismissed on empirical grounds. We obtained an
unbiased measure of the athletes' likely expectations prior to the Olympics and then used a
regression analysis to examine the effect of medal won (bronze or silver) after initial
expectations were controlled statistically. The athletes' likely expectations were derived from
Sports Illustrated's Olympic preview (Verschoth, 1992). Sports Illustrated predicted the likely
bronze, silver, and gold medal winners of every Olympic event the week before the games
b
e
g
an. Athletes who were ex
p
ected to win
g
old, silver, or bronze medals were assi
g
ned
Pa
g
e 6 of 15Ovid: Medvec: J Pers Soc Ps
y
chol, Volume 69
(
4
)
.October 1995.603–610
1/8/2004htt
p
://
g
atewa
y
1.ma.ovid.com/ovidweb.c
g
i
expectation scores of 1, 2, and 3, respectively. Those not predicted to win a medal were
assigned an expectation score of 4. As anticipated, the athletes in our samples who won silver
medals were originally expected to do better (M = 2.8) than those who won bronze medals (M =
3.0), although not significantly so, t < 1.0. More important, however, is that a comparison of
actual and anticipated performance argues against the claim that our results are due to differences
in initial expectations of bronze and silver medalists. Silver medalists as a whole did better than
anticipated (actual = 2.0; anticipated = 2.8), and therefore should have been relatively happy.
Bronze medalists, on the other hand, performed on average exactly as expected (actual and
anticipated = 3.0).
More formally, we entered the expected finish of each athlete into a regression equation that
predicted the agony–ecstasy ratings from the medal won (silver or bronze), the medal predicted
(gold, silver, bronze, or none), and the type of videotape segment (immediate reactions or medal
stand). This analysis revealed that the effect of medal won remained significant when
expectations were statistically controlled, t(72) = 4.3, p < .0001.3 Silver medalists looked less
satisfied with their performances than did bronze medalists, and they did so for reasons unrelated
to how well they were expected to perform.
Discussion
Our first study highlights a reliable context—Olympic competition involving bronze and silver
medal winners—in which those who perform better nonetheless feel worse. On the surface this
result is surprising because an underlying premise of all serious athletic competition is that
athletes should strive as hard as they can, and that the higher they finish the better they feel.
When examined with an eye toward the athletes' counterfactual thoughts, however, our findings
seem less surprising. To the silver medalist, the most vivid counterfactual thoughts are often
focused on nearly winning the gold. Second place is only one step away from the cherished gold
medal and all of its attendant social and financial rewards. Thus, whatever joy the silver medalist
may feel is often tempered by tortuous thoughts of what might have been had she only
lengthened her stride, adjusted her breathing, pointed her toes, and so on. For the bronze
medalist, in contrast, the most compelling counterfactual alternative is often coming in fourth
place and being in the showers instead of on the medal stand.
But can we confidently attribute these results to the athletes' counterfactual thoughts?
Although the data from Study 1 are consistent with this claim, it is important to examine directly
the proposed asymmetry in the athletes' counterfactual comparisons. The following two studies
were designed to do exactly that. Do silver medalists tend to think about how they almost won
the gold? Do bronze medalists focus on how close they came to missing out on a medal
altogether? What exactly do athletes think about after they learn their medal standing?
Study 2
To examine the nature of Olympic medalists' counterfactual thoughts, we turned once again to
NBC's coverage of the 1992 Summer Olympic games. NBC's sportscasters interviewed
numerous medal winners immediately following their events, and from this footage we
developed a master tape of all of NBC's interviews of bronze and silver medalists. Participants
were shown the tape and asked to assess the extent to which the athletes seemed preoccupied
with thoughts of how they did perform versus how they almost performed.
Method
Pa
g
e 7 of 15Ovid: Medvec: J Pers Soc Ps
y
chol, Volume 69
(
4
)
.October 1995.603–610
1/8/2004htt
p
://
g
atewa
y
1.ma.ovid.com/ovidweb.c
g
i
Participants.
Ten Cornell University students served as participants. As in the first study, we recruited
students who considered themselves to be nonsports fans because we did not want any prior
knowledge about the athletes to affect their ratings.
Stimulus materials.
NBC interviewed 13 silver medalists and 9 bronze medalists immediately after their events,
and these 22 interviews comprised the stimulus tape for this study. Two versions of the tape were
created, with the order of presentation of the athletes varied across the versions. The average
length of each interview clip was 27 s (SD = 14 s). Blank spaces were inserted between the
interviews to allow participants time to complete their ratings.
Procedure.
Participants arrived at the laboratory in groups and were told that they would be watching a
videotape of athletes from the 1992 Olympic games. They were asked to watch and listen to each
interview carefully and to rate each athlete's comments in two ways. First, they rated the
apparent content of each athlete's thoughts on a 10-point scale ranging from “at least I …” (1) to
“I almost …” (10). To clarify the meaning of this scale, participants were given an example of
how a student who receives a B in a course could have various thoughts ranging from “at least I
didn't get a C” to “I almost got an A.”
Second, participants were asked to assess the extent to which the apparent content of the
athlete's thoughts fell into three categories: (a) “Athlete seems focused on how he/she could have
done worse; makes a comparison with one or more competitors who finished behind;” (b)
“Athlete seems focused on how he/she could have done better; makes a comparison with one or
more competitors who finished ahead;” (c) “Athlete seems focused on what he/she
accomplished; no comparison to competitors.” Participants were asked to indicate the percentage
of the athlete's thoughts that seemed focused on each of the three categories. They could assign
any number from 0 to 100% to each of the three categories, but the percentages they assigned
had to add up to 100% for each athlete. The participants were asked to watch a number of clips
without making any ratings so that they were aware of the types of comments they would be
evaluating. Once participants were familiar with the format of the videotape and the rating
scales, the rating session began.
Five participants rated each of the two versions of the videotape. As in the first study, a 1.5-
inch (3.8 cm) strip of paper was affixed to the bottom of the video screen to occlude various
graphics depicting the athlete's order of finish.
Results
The interrater reliability of participants' ratings was acceptably high (Spearman-Brown index
= .74 and .93 for the first and second measures, respectively 4), and so the ratings were averaged
for each scale to create indices of the apparent thoughts of each athlete. Preliminary analyses of
these data revealed no effect of order of presentation, so the data were collapsed across the two
versions of the tape.
As
p
redicted, silver medalists' thou
g
hts were rated as bein
g
more focused on “I almost” than
Pa
g
e 8 of 15Ovid: Medvec: J Pers Soc Ps
y
chol, Volume 69
(
4
)
.October 1995.603–610
1/8/2004htt
p
://
g
atewa
y
1.ma.ovid.com/ovidweb.c
g
i
were those of bronze medalists. On the 10-point “At least I” to “I almost” scale, participants
assigned silver medalists' thoughts an average rating of 5.7 (SD = 1.5) and bronze medalists'
thoughts an average rating of only 4.4 (SD = 0.7), t(20) = 2.37, p < .03.
The data from the second measure were less clear cut. First, participants thought that only a
small percentage of the athletes' thoughts were focused downward on those they beat. The
average assigned to this category was only 7.5% and did not differ between bronze and silver
medalists. The percentages assigned to the other two categories conformed more closely to our
predictions. Participants rated silver medalists as being more focused on upward comparisons (M
= 38%) than bronze medalists (M = 20%), whereas bronze medalists were judged to be more
focused on their own performance (M = 73%) than silver medalists (M = 54%). Because these
data are not independent (the percentage assigned to all categories must equal 100%), our test of
significance was based on an index that combined the last two categories. Specifically, the
percentage assigned to the category “looking upward” was subtracted from the percentage
assigned to the category “focusing on one's own performance.” As predicted, this index was
higher for bronze medalists (M = 53%) than for silver medalists (M = 16%), although the
difference was only marginally significant, t(20) = 1.57, p < .15.
Discussion
The results of the second study provide support for the hypothesized difference in the
counterfactual thoughts of the bronze and silver medalists. Silver medalists seem to be focused
on the gold medal they “almost” won, while bronze medalists seem content with the thought that
“at least I did this well.” This asymmetry can thus explain the observed differences in the
athletes' expressed emotions in Study 1. This can be seen most clearly through an analysis that
combines the data from Studies 1 and 2. Fifteen of the 22 athletes whose counterfactual thoughts
were assessed in Study 2 were on the immediate-reactions videotape in Study 1 and thus were
also rated on the agony–ecstasy scale. As we predicted, the two ratings correlated significantly:
The more focused the athletes were on almost finishing higher, the less happy they seemed (r = -
.56, p < .05).5 This relationship was also observed when the data for silver (r = -.51; n = 10) and
bronze (r = -.34; n = 5) medalists were considered separately, although the sample sizes were
then too small to yield statistical significance. Thus, by focusing on what they achieved, bronze
medalists are rather happy; in contrast, a concern with what they failed to achieve makes the
silver medalists seem less so.
In this study we did not have direct access to the athletes' thoughts; we had participants infer
them on the basis of the athletes' comments. It is certainly possible, of course, that the athletes
had various thoughts they did not verbalize. To overcome this limitation, we conducted a third
study that examined bronze and silver medalists' own reports of their thoughts following an
athletic competition.
Study 3
In designing Study 3, we sought an athletic forum with significant stakes where we could gain
access to bronze and silver medalists immediately after their events. The 1994 Empire State
Games provided such a forum. The Empire State Games have been a prominent amateur athletic
event in New York State for the last 17 years. Athletes from across the state compete on a
regional basis to qualify for the Empire State Games. Notable participants have included such
athletes as Olympic gold medalists Dianne Roffe-Steinrotter and Jeff Blatnick and NBA
b
asketball stars
(
and “Dream Team” members
)
Christian Laettner and Chris Mullin. In 1994,
Pa
g
e 9 of 15Ovid: Medvec: J Pers Soc Ps
y
chol, Volume 69
(
4
)
.October 1995.603–610
1/8/2004htt
p
://
g
atewa
y
1.ma.ovid.com/ovidweb.c
g
i
more than 5,000 athletes from across New York State competed in the 4-day event.
Method
Participants.
One hundred fifteen Empire State Game medalists participated in this study. All of the
participants won bronze (n = 55) or silver (n = 60) medals in swimming or track events. The
athletes competed in either the Scholastic Division (composed exclusively of students up to 12th
grade; n = 31 males and 34 females) or the Open Division (consisting mainly of college students;
n = 25 males and 25 females).
Procedure.
The athletes were approached individually following their events and asked to rate their
thoughts about their performance on the same 10-point scale used in Study 2. Specifically, they
were asked to rate the extent to which they were concerned with thoughts of “At least I …” (1)
versus “I almost” (10). Special effort was made to ensure that the athletes understood the scale
before making their ratings. This was accomplished by mentioning how athletes might have
different thoughts following an athletic competition, ranging from “I almost did better” to “at
least I did this well.”6
Results
As predicted, silver medalists' thoughts following the competition were more focused on “I
almost” than were bronze medalists'. Silver medalists described their thoughts with a mean rating
of 6.8 (SD = 2.2), whereas bronze medalists assigned their thoughts an average rating of 5.7 (SD
= 2.7), t(113) = 2.4, p < .02.
Discussion
The data from this study are consistent with the findings from Study 2: Following a
competition, silver medalists tend to focus more on what they failed to achieve than do bronze
medalists. This asymmetry in counterfactual comparisons explains why bronze medalists tend to
be happier than silver medalists. While bronze medalists can find contentment in thinking “at
least I won a medal,” silver medalists are often confronted with an imagined outcome that almost
occurred—a preferred outcome in which they are the winner and have the gold medal hanging
around their neck. Imagining what might have been can lead those who do better to feel worse
than those they outperform.
General Discussion
The purpose of this research was to examine whether there are reliable situations in which
those who are objectively better off nonetheless feel worse than those in an inferior position.
Athletics offered an ideal context in which to test this question for the same reason that it offers a
useful context for investigating many psychological hypotheses—the availability of data of
unusual objectivity and precision (Baumeister & Steinhilber, 1984; M. G. Frank & Gilovich, 1988; Gilovich,
Vallone, & Tversky, 1985; Lau & Russell, 1980). In addition, athletics was chosen as the domain of
investigation in this case because performance in athletics often yields a clearly defined rank
order: Someone enters the record books as the first-, secon
d
-, or thir
d
-
p
lace finisher.
Pa
g
e 10 of 15Ovid: Medvec: J Pers Soc Ps
y
chol, Volume 69
(
4
)
.October 1995.603–610
1/8/2004htt
p
://
g
atewa
y
1.ma.ovid.com/ovidweb.c
g
i
It should be clear, however, that the significance of the present results extends far beyond the
playing field or the medal stand. There are many other situations in which the same processes
documented here can likewise cause those who are better off to feel worse. A student who misses
out on an A- by one point and earns a B+ could easily feel less satisfied with the grade than
someone who is in the middle of the pack of Bs. Or consider a person who correctly guesses all
but one number in a lottery. Such an individual misses out on the jackpot, but usually wins a
modest sum for coming close. The prize no doubt provides some enjoyment, but the knowledge
of having just missed the jackpot is bound to come up from time to time and ruin otherwise
blissful moments. More generally, as our opening quote from William James suggests, being one
of the best may not be as satisfying as it might seem. The existence of a rival “best” can turn a
gratifying appreciation of what one is into a disquieting focus on what one is not.
The hedonic impact of such a rival “best” raises the question of the extent to which social
comparison processes rather than (or in addition to) counterfactual thoughts may have been
responsible for our findings. We believe that our results are best situated in the research on
counterfactual thinking for two reasons. First, we obtained evidence for the hypothesized
asymmetry in the direction of counterfactual comparisons in Studies 2 and 3, but as yet no such
evidence exists to support an asymmetry in the direction of social comparisons. Second, there is
nothing in social comparison theory per se that would predict upward comparisons on the part of
silver medalists and downward comparisons on the part of bronze medalists. Although such a
pattern could certainly be made to fit with social comparison theory, it requires extratheoretical
elements to do so. In contrast, the present pattern of results was originally derived from the work
on counterfactual thinking and the psychology of “coming close” (Kahneman & Varey, 1990; Miller et
al., 1990).
This does not mean, of course, that social comparison processes are never activated in the
immediate aftermath of Olympic competition, or that such processes contributed nothing to the
present findings. Social comparison processes and counterfactual thoughts are doubtless
frequently intertwined. Social comparisons can be a source of counterfactual thoughts about
“possible worlds” that one would not have otherwise, and counterfactual thoughts can make
salient particular social comparisons that would otherwise remain hidden. Unfortunately, it is
presently unclear how much of the asymmetry in counterfactual thinking we documented in this
context was intertwined in this way with significant social comparisons.
Although the predicted findings were originally derived from previous research on
counterfactual thinking, they also extend the work in this area in two important respects. First, as
we stated at the outset, past research has held outcome constant and shown that the same
outcome can give rise to very different reactions as a function of the counterfactual thoughts that
are generated. Our results take this a step further: There are contexts in which people's
counterfactual thoughts are sufficiently powerful to lead those who are objectively worse off to
be reliably happier than those in a better position.
Our results also extend previous findings in this area by emphasizing the “automatic” or
“imposed” nature of many counterfactual thoughts. Much of the recent work on counterfactual
thinking has emphasized a person's ability to choose the most strategic counterfactual
comparisons (Markman et al., 1993; Roese, 1994). “Counterfactual generation has functional value, and
people tend to generate those counterfactuals that hold the greatest psychological value for them
in a given situation” (Markman et al., 1993, p. 103). Downward comparisons (i.e., thinking about a
worse outcome) are thought to provide comfort, whereas upward comparisons (i.e., thinking
about a better outcome) are thought to improve future performance. Indeed, it has been shown
that
p
eo
p
le who ex
p
ect to
p
erform a
g
ain in the future are more likel
y
to
g
enerate u
p
ward
Pa
g
e 11 of 15Ovid: Medvec: J Pers Soc Ps
y
chol, Volume 69
(
4
)
.October 1995.603–610
1/8/2004htt
p
://
g
atewa
y
1.ma.ovid.com/ovidweb.c
g
i
counterfactuals than those who expect to move on (Markman et al., 1993).
Although many counterfactual thoughts are doubtless strategically chosen in this way, such
motivational considerations cannot account for the present findings. On the whole, the silver and
bronze medalists at the Barcelona Olympics were at the peak of their athletic careers and
therefore likely to continue to engage in similar high-level competitions in the future. From a
motivational perspective, then, both groups should have made upward counterfactual
comparisons in order to prepare for future contests. The asymmetry in counterfactual
comparisons that we observed implies that many counterfactuals are imposed by the nature of the
events experienced.
Indeed, Kahneman (in press) outlined a continuum of counterfactual thinking that ranges from
“automatic” to “elaborative.” Elaborative counterfactual processing is partly brought on through
the exercise of choice, and its direction and intensity is influenced by the individual's motives
and intentions. Automatic counterfactual thinking, in contrast, is “initiated by the occurrence of
an event and … [is] … explainable largely in cognitive terms” (Kahneman, in press). The
counterfactual thoughts that distinguish silver and bronze medalists shade toward the latter end
of this continuum. Coming close to winning the gold, for example, appears to automatically
activate frustrating images of having almost won it all.
We are not suggesting, of course, that finishing second or coming close to a cherished
outcome always leads to less satisfaction than a slightly more modest performance. Finishing
second is truly a mixed blessing. Performing that well provides a number of direct benefits that
increase our well being—recognition from others, boosts to self-esteem, and so on. At the same
time, it can indirectly lower satisfaction by the unfortunate contrast with what might have been.
Thus, the inconsistent effect of finishing second is analogous to the “endowment” and “contrast”
polarity that Tversky and Griffin (1991) claimed affects the hedonic significance of all experienced
events. According to their analysis, any experience has a direct effect on well-being by what it
brings to one's endowment—that is, the pleasure or pain derived from the event itself. But a
person's experiences also have an indirect effect on well being by altering the adaptation level
against which future experiences are contrasted. Their contrast (in which the event itself
establishes a new standard against which future events are compared) is different than the one at
work here (in which the events' proximity to a better outcome causes one to lose sight of what is
and focus on what might have been). The core idea is the same, however. In both cases, the
direct effect of the event itself is offset by a comparison process with the opposite effect, be it a
comparison of future outcomes to the present, or the present outcome to a counterfactual
alternative that was almost attained.
Tversky and Griffin (1991) have delineated some of the general rules that govern the relative
weighting of endowment and contrast, and thus whether the net effect of a given event enhances
or diminishes well being. They acknowledged, however, that the degree to which a given event
evokes endowment and contrast can be highly idiosyncratic. As a consequence, when applied to
a problem such as ours it can be difficult to predict exactly when those who are better off will
nonetheless feel worse than those who are less fortunate.
Another unresolved issue, this one more tractable, concerns the duration of the effects we have
documented here. We have established that bronze medalists are happier than silver medalists in
the short run, but does this effect hold up over time? As yet there are no data to answer this
question. Nevertheless, one of the most noteworthy features of life's near misses seems to be
their durability. Consider the account of finishing second that Nicholson Baker (1991) provides his
wife:
Pa
g
e 12 of 15Ovid: Medvec: J Pers Soc Ps
y
chol, Volume 69
(
4
)
.October 1995.603–610
1/8/2004htt
p
://
g
atewa
y
1.ma.ovid.com/ovidweb.c
g
i
[I] told her my terrible story of coming in second in the spelling bee in second grade by
spelling keep ‘c-e-e-p’ after successfully tossing off microphone, and how for two or three years
afterward I was pained every time a yellow garbage truck drove by on Highland Avenue and I
saw the capitals printed on it, ‘Help Keep Our City Clean,’ with that impossible irrational K that
had made me lose so humiliatingly …
Or consider the case of Abel Kiviat, the 1,500 m silver medalist in the 1912 Olympics in
Stockholm. Kiviat had the race won until Britain's Arnold Jackson “came from nowhere” to beat
him by one-tenth of a second. “I wake up sometimes and say, ‘What the heck happened to me?’
It's like a nightmare.” Kiviat was 91 years old when he said this in an interview with the Los
Angeles Times (cited in Tait & Silver, 1989, p. 351). It appears that thoughts about what might have
been may plague us for a very long time.
References
Atkinson, J. W. (1964). An introduction to motivation. Princeton, NJ: Van Nostrand. [Context Link]
Baker, N. (1991). Room temperature. New York: Vintage. [Context Link]
Baumeister, R. F., & Steinhilber, A. (1984). Paradoxical effects of supportive audiences on performance under pressure:
The home field disadvantage in sports championships. Journal of Personality and Social Psychology, 47, 85–93. [Context
Link]
Crosby, F. (1976). A model of egoistical relative deprivation. Psychological Review, 83, 85–113. Bibliographic Links
[Context Link]
Feather, N. T. (1967). Valence of outcome and expectation of success in relation to task difficulty and perceived locus of
control. Journal of Personality and Social Psychology, 7, 372–386. [Context Link]
Feather, N. T. (1969). Attribution of responsibility and valence of success and failure in relation to initial confidence and
task performance. Journal of Personality and Social Psychology, 13, 129–144. [Context Link]
Festinger, L. (1954). A theory of social comparison processes. Human Relations, 7, 117–140. Bibliographic Links [Context
Link]
Frank, M. G., & Gilovich, T. (1988). The dark side of self- and social perception: Black uniforms and aggression in
professional sports. Journal of Personality and Social Psychology, 54, 74–85. [Context Link]
Frank, R. H., & Cook, P. (1995). The winner-take-all society. New York: Free Press. [Context Link]
Gilovich, T., Vallone, R. P., & Tversky, A. (1985). The hot hand in basketball: On the misperception of random
sequences. Cognitive Psychology, 17, 295–314. Bibliographic Links [Context Link]
James, W. (1892). Psychology. New York: Holt. [Context Link]
Johnson, J. T. (1986). The knowledge of what might have been: Affective and attributional consequences of near
outcomes. Personality and Social Psychology Bulletin, 12, 51–62. [Context Link]
Kahneman, D. (in press). Varieties of counterfactual thinking. In N. Roese & J. Olson (Eds.), What might have been: The
social psychology of counterfactual thinking. Hillsdale, NJ: Erlbaum. [Context Link]
Kahneman, D., & Miller, D. T. (1986). Norm theory: Comparing reality to its alternatives. Psychological Review, 93,
136–153. Bibliographic Links [Context Link]
Kahneman, D., & Tversky, A. (1982a). The psychology of preferences. Scientific American, 2, 160
–
173. [Context Link]
Pa
g
e 13 of 15Ovid: Medvec: J Pers Soc Ps
y
chol, Volume 69
(
4
)
.October 1995.603–610
1/8/2004htt
p
://
g
atewa
y
1.ma.ovid.com/ovidweb.c
g
i
Kahneman, D., & Tversky, A. (1982b). The simulation heuristic. In D. Kahneman, P. Slovic, & A. Tversky (Eds.),
Judgment under uncertainty: Heuristics and biases (pp. 201–208). New York: Cambridge University Press. [Context Link]
Kahneman, D., & Varey, C. A. (1990). Propensities and counterfactuals: The loser that almost won. Journal of
Personality and Social Psychology, 59, 1101–1110. [Context Link]
Lau, R. R., & Russell, D. (1980). Attributions in the sports pages: A field test of some current hypotheses about
attribution research. Journal of Personality and Social Psychology, 39, 29–38. [Context Link]
Markman, K. D., Gavanski, I., Sherman, S. J., & McMullen, M. N. (1993). The mental simulation of better and worse
possible worlds. Journal of Experimental Social Psychology, 29, 87–109. Bibliographic Links [Context Link]
Miller, D. T., & McFarland, C. (1986). Counterfactual thinking and victim compensation: A test of norm theory.
Personality and Social Psychology Bulletin, 12, 513–519. [Context Link]
Miller, D., Turnbull, W., & McFarland, C. (1990). Counterfactual thinking and social perception: Thinking about what
might have been. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 22, pp. 305–331). San Diego,
CA: Academic Press. [Context Link]
Olson, J. M., Herman, P., & Zanna, M. P. (1986). Relative deprivation and social comparison: The Ontario symposium
(Vol. 4). Hillsdale, NJ: Erlbaum. [Context Link]
Roese, N. J. (1994). The functional basis of counterfactual thinking. Journal of Personality and Social Psychology, 66,
805–818. [Context Link]
Roese, N. J., & Olson, J. M. (in press). What might have been: The social psychology of counterfactual thinking.
Hillsdale, NJ: Erlbaum. [Context Link]
Suls, J. M., & Miller, R. L. (1977). Social comparison processes: Theoretical and empirical perspectives. New York:
Wiley. [Context Link]
Tait, R., & Silver, R. C. (1989). Coming to terms with major negative life events. In J. S. Uleman & J. A. Bargh (Eds.),
Unintended thought (pp. 351–382). New York: Guilford Press. [Context Link]
Taylor, S. E., & Lobel, M. (1989). Social comparison activity under threat: Downward evaluation and upward contacts.
Psychological Review, 96, 569–575. [Context Link]
Turnbull, W. (1981). Naive conceptions of free will and the deterministic paradox. Canadian Journal of Behavioural
Science, 13, 1–13. Bibliographic Links [Context Link]
Tversky, A., & Griffin, D. (1991). Endowment and contrast in judgments of well-being. In R. J. Zeckhauser (Ed.),
Strategy and choice (pp. 297–318). Cambridge, MA: MIT Press. [Context Link]
Verschoth, A. (1992, July 22). Who will win what. Sports Illustrated. [Context Link]
Wells, G. L., & Gavanski, I. (1989). Mental simulation of causality. Journal of Personality and Social Psychology, 56,
161–169. [Context Link]
1The reason that Americans were not as overrepresented in this tape as they were in the other
is that for many of our medal stand segments, it is the gold medal winner who is being featured
by NBC and the silver and bronze medalists who are pictured incidentally. Thus, there were
many instances in which NBC was focused on an American gold medalist, and we were able to
capitalize on their ancillary coverage of a silver or bronze medalist from another country. [Context
Link]
2One other as
p
ect of the data should be noted. Because seven of the athletes
p
ictured in the
Pa
g
e 14 of 15Ovid: Medvec: J Pers Soc Ps
y
chol, Volume 69
(
4
)
.October 1995.603–610
1/8/2004htt
p
://
g
atewa
y
1.ma.ovid.com/ovidweb.c
g
i
immediate-reactions videotape are also shown on the medal stand tape, the data are not all
strictly independent and the ANOVA reported is not completely accurate. However, the results
are not changed when this overlap is deleted from the analysis. We did this two ways: first, by
keeping these athletes' immediate-reactions data and deleting their medal stand data; second, by
deleting their immediate reactions and keeping their data from the medal stand. Regardless of
which analysis we used to eliminate the redundancy, the bronze medalists were still rated
significantly happier than the silver medalists. [Context Link]
3This effect remains significant when the athletes from the just won–just lost events are
excluded and when the redundancy created by the 7 athletes who appear on both tapes is
removed. [Context Link]
4The .93 is the average Spearman–Brown index for the last two components of the second
dependent measure, which, as we discuss in the main text, are the focus of our analysis. [Context
Link]
5We used the data from the immediate-reactions videotape in Study 1 rather than those from
the medal stand tape simply because there was more overlap with the athletes interviewed in
Study 2 for the former (15) than for the latter (7). Furthermore, 5 of the overlapping 7 from the
medal stand tape also appeared on the immediate-reactions tape. [Context Link]
6We had hoped to include a question similar to the second measure used in Study 2 in which
participants divided the athletes' thoughts into three categories by assigning the appropriate
percentage to each. We thought this might be difficult for the athletes to do the moment they
emerged from the heat of competition, however. We therefore tried to simplify matters by
presenting the task spatially: The athletes were shown a plexiglass board in which we had carved
a triple-pronged “pitchfork.” The athletes were to distribute 10 metal tokens contained in the
handle of the pitchfork into the categories represented by each prong. The three prongs were
labeled “Who I Beat,” “No Comparisons,” and “Who Beat Me,” and the athletes were told to
apportion the tokens so as to represent the extent to which their thoughts were focused on each.
Unfortunately, the measure proved to be exquisitely ineffective. Some athletes, particularly a
number of shivering swimmers who had just emerged from the pool, seemed unable to
comprehend it; others managed to dislodge the tokens; and the responses of still otherathletes
were contaminated by the comments of onlookers who found the device fascinating and offered
unsolicited advice. [Context Link]
Accession Number: 00005205-199510000-00002
Copyright (c) 2000-2003 Ovid Technologies, Inc.
Version: rel9.0.0, SourceID 1.8300.1.161
Pa
g
e 15 of 15Ovid: Medvec: J Pers Soc Ps
y
chol, Volume 69
(
4
)
.October 1995.603–610
1/8/2004htt
p
://
g
atewa
y
1.ma.ovid.com/ovidweb.c
g
i
A preview of this full-text is provided by American Psychological Association.
Content available from Journal of Personality and Social Psychology
This content is subject to copyright. Terms and conditions apply.
