Abstract
Achieving an ambitious goal frequently requires succeeding in a sequence of intermediate tasks, some being critical for the final outcome, and others not. However, individuals are not always able to provide a level of effort sufficient to guarantee success in all such intermediate tasks. The ability to manage effort throughout the sequence of tasks is, therefore, critical when resources are limited. In this paper, we propose a criterion of importance that is person- and context-specific, as it is based on how an individual should optimally allocate a limited stock of exhaustible efforts over tasks. We test this importance criterion in a laboratory experiment that reproduces the main features of a tennis match. We show that our importance criterion is able to predict the individuals’ performance and it outperforms the Morris-importance criterion that defines the importance of a point in terms of its impact on the probability of achieving the final outcome.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
Note that, as in Morris (1977), we define what importance is and measure how relevant our definition is to explain behavior, but we are not stating a theory to explain behavior. Let us illustrate this point by considering an individual that has to complete with success three intermediary steps to reach her final goal. A failure at any intermediary step means a global failure. Her “base” probability to succeed at any step is 1/2. Using the Morris-importance criterion, the first intermediary step has a measure of importance of 1/4. Indeed, if successful at the first step, the individual still has a 1/4 chance to succeed at the remaining two steps. If not successful at the first step, the individual has a 0 probability to have a global success by assumption. The importance is then 1/4–0 = 1/4. If successful and reaching the second intermediary step, the individual is now facing a step with importance 1/2. Indeed, if successful at the second step, the individual still has a 1/2 chance to succeed at the remaining last step. If not successful at the second step, the individual has a 0 probability to have a global success. The importance is then 1/2–0 = 1/2. If we want the notion of importance to have any positive and relevant content, we need to assume that the individual’s behavior will be modified and the a priori probability to succeed at the second step differs from the a priori probability to succeed at the first step, because importance is no longer the same. But then, the computation that we made for the probabilities to succeed at the global level to compute importance itself at each step is flawed: we should have taken into account the way behavior is modified by importance in subsequent steps in the very computation of importance. At best, we have a fixed-point problem, at worse a circular definition. The only way to escape this reasoning is to consider that the definition of importance is based on a model whose purpose is not to model the behavior of a more or less rational and forward-looking individual, but only to derive a definition whose relevance is to be tested empirically and to be used in predictive models. This methodological viewpoint is standard in the machine learning literature and has been defended by Friedman (1953) or Koopmans (1957) in economics.
HNV stands for Houy-Nicolaï-Villeval.
Note that there is a literature in economics on the effects of goal setting on performance (e.g., Koch and Nafziger 2015; Goerg and Kube 2012; Clark et al. 2017). In contrast, we are not interested in measuring the impact of goal setting on performance, but we examine how individuals allocate their effort depending on the succession of stages leading to a final outcome when the importance of these stages varies.
The idea of choking under pressure was first developed in psychology to demonstrate that the importance of an issue may generate pressure that will have negative effects on performance. In particular, Baumeister (1984) and Baumeister and Showers (1986) showed that pressure increases with the existence of competition between agents (see also, e.g., Beilock and Carr 2001; DeCaro et al. 2011; Porcelli et al. 2012; Yu 2015). Sanders and Walia (2012) insist on the fact that higher performance-contingent compensation may also have a negative impact on labor input if counterproductive processes decrease the marginal effectiveness of effort. Ariely et al. (2009) further showed in a series of field and lab experiments that very high monetary stakes tend to impair performance, because they generate supra motivation and excess stress. Using tennis data, Cohen-Zada et al. (2017) found that males choke consistently under pressure and to a larger extent than females. Contradictory evidence has been shown in the analysis of penalty shootouts in football competitions (Apesteguia and Palacios-Huerta 2010; Kocher et al. 2012; Dohmen 2008). Recent evidence has also shown that people may choke even under the pressure of non-monetary incentives (Kali et al. 2017).
Note that, as explained in footnote 1, we do not consider our model with the use of efforts as a model to be interpreted literally, but as a model giving predictions that we want to test. Hence, from that perspective, we do not consider the difference between the decision making setting of a human playing against a robot and the game-theoretical setting of two humans playing against each other [see Klumpp et al. (2019) for instance]. Solving the latter game is a different question than the one which we tackle here and it should be addressed in another unrelated piece of work.
If P is a minimal solution and \(P'\) is any solution, then \(\forall s \in S^*, P(s) \le P'(s)\). For more details, see Norris (1998). Notice that since P is the minimal non-negative solution, \(P(\{L\})=0\).
For the sake of simplicity, we consider the function \(s^+\) defined onto \(S^*\), whereas it should more rigorously be set-valued.
This concept is also used in Magnus and Klaassen (1998), Paserman (2007) and González-Díaz et al. (2012). However, González-Díaz et al. (2012) further use conditional probabilities, which allows them to break down the Morris-importance into the constituent probabilities of winning at the various hierarchical levels of the match.
The stock of efforts is exogenous, but it could be possible to make it endogenous. Indeed, the agent could decide the stock of efforts before the start of the game. The stock of efforts would depend on a number of anticipated tasks and the cost of creating an exhaustible effort stock.
Of course, this is a simplification, as, in reality, individuals may choose finer levels of effort.
Proof of existence and uniqueness is similar to the one for function P and is, therefore, omitted.
01 stands for either no or one effort.
We tested whether the average individual performance in these practice rounds differs across treatments. Performance in the practice shots is similar in the Human treatment (where people could possibly behave strategically to misrepresent their true ability; mean = 9.72) and in the Robot treatment (where they have no reason to behave strategically, as explained below, mean = 8.07) (two-tailed Mann-Whitney, z = 0.261, p = 0.794). Nevertheless, the lack of significant difference might hide the fact that in the Human treatment, some subjects do their best, while others provide strategically little effort. We reject this, as a Kolmogorov–Smirnov test shows that there is no significant difference in the distribution of performance in the practice shots across treatments (exact p = 0.914).
We assume that participants do not consider their stock of effort when performing the practice rounds, in contrast to the competition game, because there are only ten practice shots and each shot is independent from the others (there is no notion of set or match during the practice). Thus, the measure of relative ability in the practice shots is assumed to be comparable for building our two measures of importance. The Morris-importance criterion should not be more sensitive to the practice round information than the HNV-importance criterion.
Indeed, in most actual sport or occupational tournaments, when they start competing, competitors are aware of their relative ability compared to that of their opponents, based on the previous records of performance.
If the participant shoots in less than 5 s after the target appears on the screen, the window of analysis includes a minimum of 5 s. If the participant shoots between 5 and 8 s, the window analysis is the time between the display of the target and the timing of the shot.
The signal was low-pass filtered at 0.5 Hz offline, using a 5th-order Butterworth low-pass digital filter. The onset and peak of the SCR were detected when the first derivative of the filtered signal changed sign thanks to a routine written in Matlab (The MathWorks Inc., USA). Onsets were identified by a negative to positive zero crossing. Peaks were identified by a positive to negative zero crossing. The SCR amplitude was calculated as the difference between the signal amplitude at the peak and the onset times. It was threshold at 0.02 \(\upmu\)S. The whole signal was visually inspected prior to further analysis and ectopic response was removed. The analysis assigns zero amplitude to the subjects without a measurable response (Dawson et al. 2007).
We do not need to use tobit, since only 0.41% of the data are censored at 0.
Every 50 ms, we draw two integer numbers, lx and ly in the interval [− 5; 5] with uniform probability. The player’s pointer is modified by lx pixels on the x-axis and ly pixels on the y-axis. Let \(Lx=(lx_i)_{i \in [1;5]}\) be the sequence of the five values of lx drawn before the player shot and \(\textit{Ly}=(\textit{ly}_\textit{i})_{i \in [1;5]}\) be the sequence of the five values of ly drawn before the player shot. The value which we report is \(S=\sqrt{(}(\sum _{i=1}^{5}lx_i)^2+(\sum _{i=1}^{5}ly_i)^2)\). Then, S is the distance between the positions of the pointer 0.25 s before the player’s shot and at the time of the shot that is generated by the noise alone.
We also tested the potential influence on the current distance of the emotional arousal experienced during the previous shot at the time of the feedback on one’s score and on the trial’s outcome. This requires dropping the observations relative to the first shot in the game. Since these two SCR measures were never significant, we do not include these variables in the regressions reported here and keep the data of all periods. Note also that in all regressions, the SCR magnitude is used as a control variable and not as an outcome variable. To test whether a higher effort could generate a higher SCR level, we have also estimated a variety of regression models with clustering at the individual level with the SCR magnitude as the dependent variable and performance as an independent variable. Regardless of the specification used, effort never correlates significantly with the SCR magnitude and it is far from reaching standard significance levels (lowest p value = 0.18).
We also estimated models 4 and 5 in Table 5 and all models of Table 6 without including the Distance Practice and Relative Ability variables among the independent variables. This does not change our results (see Table 10 in Appendix). Note also that all our regressions have been conducted with cluster adjusted standard errors at the individual level instead of using GLS models with fixed effects. This is usually considered as more efficient and this accounts for within-cluster correlation and heteroskedasticity; we included a few observable individual characteristics as controls to test whether ability, gender, or age affects performance specifically. However, we also ran GLS models with fixed effects to control for both observable and unobservable time-invariant individual characteristics. The Morris-importance criterion does not become more significant. The HNV-importance criterion in the Human treatment is less precisely estimated than in the original regressions (it is never significant in the Robot treatment, like before). It remains significant for stocks of effort of 20, 30, and 50 (at the 5%, 1%, and 10%, respectively); the p value of the coefficients for stocks of effort of 10 and 40 does not reach standard levels of significance (p = 0.121 and 0.111, respectively). These estimations are available upon request.
Note that we also tested the same model considering the assignment of more than 50 units of effort. The HNV allocation variable remains significant at the 1% level. To save space, we omit these regressions here.
For some individuals, we have to reject the five regressions and assign to the player a missing HNV allocation variable. This occurs when the HNV allocation variable receives a positive coefficient (indicating that performance decreases when the point is more important in the sense of the HNV model). This corresponds to 7.14% of the players in the Human treatment and to 9.09% of the players in the Robot treatment (the difference is not statistically significant according to a proportion test). These observations are kept in the regressions and we include a dummy variable (“Missing HNV allocation”) to the other independent variables.
References
Apesteguia, J., & Palacios-Huerta, I. (2010). Psychological pressure in competitive environments: Evidence from a randomized natural experiment. American Economic Review, 100(5), 2548–2564.
Ariely, D., Gneezy, U., Loewenstein, G., & Mazar, N. (2009). Large stakes and big mistakes. Review of Economic Studies, 76(2), 451–469.
Baumeister, R. F. (1984). Choking under pressure: self-consciousness and paradoxical effects of incentives on skillful performance. Journal of Personality and Social Psychology, 46(3), 610.
Baumeister, R. F., & Showers, C. J. (1986). A review of paradoxical performance effects: Choking under pressure in sports and mental tests. European Journal of Social Psychology, 16(4), 361–383.
Beilock, S. L., & Carr, T. H. (2001). On the fragility of skilled performance: What governs choking under pressure? Journal of Experimental Psychology General, 130(4), 701–725.
Blascovich, J., Seery, M. D., Mugridge, C. A., Norris, R. K., & Weisbuch, M. (2004). Predicting athletic performance from cardiovascular indexes of challenge and threat. Journal of Experimental Social Psychology, 40(5), 683–688.
Buckert, M., Schwieren, C., Kudielka, B. M., & Fiebach, C. J. (2015). How stressful are economic competitions in the lab? an investigation with physiological measures. Technical report, Discussion Paper Series, University of Heidelberg, Department of Economics.
Clark, D., Gill, D., Prowse, V., & Rush, M. (2017). Using goals to motivate college students: Theory and evidence from field experiments. NBER Working Paper 23638, NBER.
Cohen-Zada, D., Krumer, A., Rosenboim, M., & Shapir, O. M. (2017). Choking under pressure and gender: Evidence from professional tennis. Journal of Economic Psychology, 61(August), 176–190.
Dawson, M. E., Schell, A. M., & Filion, D. L. (2007). The electrodermal system. In J. T. Cacioppo, L. G. Tassinary, & G. G. Berntson (Eds.), Handbook of psychophysiology (pp. 159–181). Cambridge University Press.
DeCaro, M. S., Carr, T. H., Albert, N. B., & Beilock, S. L. (2011). Choking under pressure: Multiple routes to skill failure. Journal of Experimental Psychology. General, 140(1), 390–406.
Dohmen, T. J. (2008). Do professionals choke under pressure? Journal of Economic Behavior & Organization, 65(3), 636–653.
Friedman, M. (1953). The methodology of positive economics. In Essays in Positive Economics (pp 3–43). University of Chicago Press.
Gauriot, R., & Page, L. (2019). Does success breed success? A quasi-experiment on strategic momentum in dynamic contests. The Economic Journal, 129(624), 3107–3136.
Goerg, S., & Kube, S. (2012). Goals (th)at Work- Goals, Monetary Incentives, and Workers’ Performance. Max Planck Institute for Research on Collective Goods 19, Max Planck Institute.
González-Díaz, J., Gossner, O., & Rogers, B. W. (2012). Performing best when it matters most: Evidence from professional tennis. Journal of Economic Behavior & Organization, 84(3), 767–781.
Greiner, B. (2015). Subject pool recruitment procedures: organizing experiments with orsee. Journal of the Economic Science Association, 1(1), 114–125.
Hancock, P. A., & Desmond, P. A. (2001). Stress, Workload, and Fatigue. Lawrence Erlbaum Associates Publishers.
Hockey, R. (2013). The Psychology of Fatigue: Work, Effort and Control. Cambridge University Press.
Kali, R., Pastoriza, D., & Plante, J.-F. (2017). The burden of glory: Competing for nonmonetary incentives in rank-order tournaments. Journal of Economic & Management Strategy, 27, 102–118.
Klumpp, T., Konrad, K. A., & Solomon, A. (2019). The dynamics of majoritarian Blotto games. Games and Economic Behavior, 117, 402–419.
Koch, A. K., & Nafziger, J. (2015). Self-regulation through goal setting. Scandinavian Journal of Economics, 113(1), 212–227.
Kocher, M. G., Lenz, M. V., & Sutter, M. (2012). Psychological pressure in competitive environments: New evidence from randomized natural experiments. Management Science, 58(8), 1585–1591.
Koopmans, T. (1957). Three Essays on the State of Economic Science. McGraw-Hill.
Magnus, J., & Klaassen, F. (1998). On the existence of “big points” in tennis: four years at wimbledon. CentER for Economic Research: Tilburg University.
Moore, L. J., Vine, S. J., Wilson, M. R., & Freeman, P. (2012). The effect of challenge and threat states on performance: An examination of potential mechanisms. Psychophysiology, 49(10), 1417–1425.
Morris, C. (1977). The most important points in tennis. Optimal Strategies in Sports, 5, 131–140.
Norris, J. (1998). Markov Chains (Vol. 2). Cambridge University Press
Paserman, M. D. (2007). Gender differences in performance in competitive environments: Evidence from professional tennis players. CEPR Discussion Papers 6335, C.E.P.R. Discussion Papers.
Porcelli, A. J., Lewis, A. H., & Delgado, M. R. (2012). Acute stress influences neural circuits of reward processing. Frontiers in Neuroscience, 6, 157.
Sanders, S., & Walia, B. (2012). Shirking and “choking” under incentive-based pressure: a behavioral economic theory of performance production. Economics Letters, 116(3), 363–366.
Yu, R. (2015). Choking under pressure: The neuropsychological mechanisms of incentive-induced performance decrements. Frontiers in Behavioral Neuroscience, 9, 19.
Acknowledgements
We are grateful to M. Joffily for assistance in the analysis of skin-conductance responses. We thank S. Chowdhury for comments on a previous version of this paper. This research has been supported by a grant of the LABEX CORTEX (ANR-11-LABX-0042), within the program Investissements d’Avenir (ANR-11-IDEX-007) operated by Agence Nationale de la Recherche, and by the IDEXLyon from Université de Lyon (project INDEPTH) within the Programme Investissements d’Avenir (ANR-16-IDEX-0005).
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendices
Appendices
1.1 Illustration of Morris-importance and HNV(k)-importance for tennis scores
Consider a tennis match between player 1 and player 2. Obviously, they both have the objective to win the match. We do not consider service and receive.
1.1.1 Illustration of Morris-importance and HNV(k)-importance for a single game
We first focus on a single game and we assume that the player who wins the game is declared victorious. Note that a margin of two points is required. We assume that the probability to win one point is the same for both players (p = 0.5).
Illustration of a single game The X axis (the Y axis, respectively) counts the gains of player 1 (player 2, respectively). Each square in Fig. 2 represents a potential point. The first square in the lower part of the figure on the left illustrates the starting point in the game (0,0). The square on the right (above) depicts the score if player 1 (player 2) has won the point.
Illustration of a single game
Morris-importance and HNV(1)-importance We calculate, for each point, focusing on player 1’s point-of-view, both Morris-importance and HNV(1)-importance. We illustrate these results in Fig. 3. The following color codes are used: green means high importance, light blue means somewhat high, dark blue means medium importance, purple means somewhat low importance, and red means low importance.
Morris-importance (left) and HNV(1)-importance (right) for all the scores of a single game for p = 0.5
Note that the two panels in Fig. 3 are symmetric. Indeed, the probability of winning one point is the same for both players. First of all, Morris-importance is the highest for the scores (30, 30), (30, 40), and (40, 30). Morris-importance is the lowest when the score difference is equal to three points, i.e., for the scores (0, 40) and (40,0). The closer to the end of the game players are, the less likely will it be for the one who loses this point to win the game. In the case of a single effort left, there are only two possibilities: the player has incentives to make an effort or not. Player 1 has positive incentives to exert effort only for (30, 40) and (40, 30). If player 1 wins at (30, 40), the score will be a draw (30, 30); otherwise, he will lose the match. In the case of (40, 30), either he wins the match or there is draw.
We consider five different cases:
-
For (30, 40) and (40, 30), the two measures are similar and high.
-
For (30, 30), the Morris-importance is high, while player 1 has no incentive to make an effort. The intuition is the following. This point is not irreversible. If player 1 loses this point, he can always make an effort subsequently and comes back to draw.
-
For (15, 15), (30, 15), and (15, 30), player 1 has no incentives to make an effort. Winning these points modifies the probability of winning the game, but these points are not irreversible: it is better to keep the effort and use it latter.
-
For scores (0, 0), (0, 15), and (15, 0), Morris-importance is medium, while HNV(1)-importance is low.
-
For the other scores, the two measures are similar and insignificant.
To conclude, there are only two possibilities: the player has incentives to make an effort or not. The closer to the end of the match players are, the less likely will it be for the player who loses this point to win the game. In other words, the closer to the end of the game the score is, the higher the HNV(1)-importance will be. We also show that the two measures are similar only at the end of a tight game or when the game is fairly one-sided and far from the end.
Two exhaustible efforts and HNV(2)-importance We also analyze the situation in which the player has two efforts left. Figure 4 depicts HNV(2)-importance for a single game. Note that when player 1 has two efforts left, the points for which he has incentives to use the first effort are clearly different from those for which he has incentives to use the second. The intuition is the following: with two remaining efforts, the goal is to use the first effort to avoid losing the game with a close score and use the last effort to win the game.
HNV(2)-importance for p = 0.5
1.1.2 Morris-importance and HNV(1)-importance for all the scores of a whole match
We consider now a whole match and calculate the Morris-importance and the HNV(1)-importance for all the potential points, for the probability p = 0.5. The results for a whole match extend the observations for a single game. Figures 5, 6 illustrate Morris-importance and HNV(1)-importance for all the scores of a whole match with p = 0.5 and where the first player to win two sets is victorious.
-
The X axis (the Y axis, respectively) displays the gains of player 1 (player 2, respectively).
-
Each small square depicts a point. A game is modeled as previously.
-
The same rules as in the experiment are used. The first of the two competitors who wins at least 4 points with at least 2 points ahead of the opponent wins the round. The first of the two competitors who wins 6 rounds wins the set. The first of the two competitors who wins 2 sets wins the match.
-
We use the same color codes as previously: high importance (green), somewhat high importance (light blue), middle importance (dark blue), somewhat low importance (purple), and low importance (red).
The figures show that the two measures are similar only at the end of a tight match and when the match is fairly one-sided and far from the end. The notion of irreversibility is crucial. When the match is tight, the individual should make an effort on the tasks which lead to an irreversibility: if this point is lost, the victory is almost impossible.
Morris-importance for all the scores of a whole game for p = 0.5
HNV(1)-importance for all the scores of a whole game for p = 0.5
1.2 Instructions for the experiment
These instructions, translated from French, are both for the Human treatment and the Robot treatment. When the instructions differ between the two treatments, those specific to the Robot treatment are put in italics and between brackets.
We thank you for participating in this experiment in economics in which you can earn a certain amount of money. You will also receive a 5 Euros participation fee for showing up on time. You will be paid in cash and in private at the end of the session.
Throughout the session, we ask you not to communicate with other participants. All information entered into the computer will remain anonymous.
1.3 Description
This experiment consists of one part in which you will be paired with one other person in the room. The identity of this person will remain unknown. Similarly, the other person with whom you will be paired will never be informed of your identity. This person will be called your “opponent” in the remainder of these instructions.
[Previous paragraph replaced by the following in the Robot treatment: This experience consists of one part in which you will be paired with one other player. This player is not a person, \({\underline{it\,is\,a\,computer}}\). This player will be called your “opponent” in the remainder of these instructions.]
Your earnings in this part depend on your performance and on that of your opponent during the game. In addition to the participation fee, one of the two co-participants will earn 20 Euros and the other will earn 0 Euro, according to the rules detailed below.
[Previous sentence replaced by the following in the Robot treatment: In addition to the participation fee, you will earn either 20 Euros or 0 Euro, according to the rules detailed below.]
This part consists of several stages, each of which is divided into several rounds. During each round, we ask you to perform the task described below.
The task The task involves using your mouse pointer to reach the center of a target that appears on your computer screen. You have to shoot as accurately as possible in the center of the target. The level of difficulty differs according to a perturbation of the position of the pointer which is random on each shot. The difficulty is the same for each participant.
Each participant has 8 s maximum to shoot. If the shot was not performed before the 8 s have elapsed, the distance to the target center is arbitrarily set to 1000.
During each round, you will have to shoot several times.
[Only in the Robot treatment:
How does your computer-opponent play?
For each shot from your robot, the program randomly selects a shot among the shots of the human participants in a previous session. During this previous session, only human participants were involved and they were paired together randomly. The “robot shot” is, therefore, not random: it is the shot of a human participant in a previous session aimed at the center of the target.]
Performance and number of points Your performance in this task is indicated by the distance between the impact of your shot and the center of the target. The shorter this distance, the better your performance.
Once the two co-participants have shot, the performances of the two co-participants are compared. The co-participant with the best performance (that is to say, the lowest distance to the center of the target) scores one point.
In case of a tie between the two co-participants, a random draw determines which of the two participants is awarded the point.
The first of the two co-participants who wins at least 4 points with at least 2 points ahead of the opponent wins the round.
The first of the two co-participants who wins 6 rounds wins the set.
The first of the two co-participants who wins 2 sets wins the game.
Determination of payoffs Within the pair, if you win the game you earn 20 Euros in addition to your participation fee. If you do not win the game, you earn nothing; you only receive your participation fee.
Information As shown in the screenshot below (Fig. 7), the screen displays a target. Below the target you can see a time counter that indicates the remaining time to shoot. After each shot, you are informed on your screen of your performance and that of your opponent. Finally, you are informed as to whether you or your opponent scores the current point. The bottom of the screen permanently shows:
-
the number of sets won so far;
-
the number of rounds won in the current set;
-
the number of points earned in the current round, both for yourself and for your opponent;
your screen also informs you whether the next shot is a shot that decides the winner of the round, the set or game.
Screenshot of the task
Beginning of the session Before the game starts, you have to make 5 test shots and 10 practice shots. The five test shots are intended to familiarize you with the task. After each shot, you are informed of your own performance (your distance to the center of the target).
The 10 practice shots are used to calculate your ability at this task compared to that of your opponent. After each of these shots, you are informed of your own performance.
After the 10 practice shots, the computer program compares your performance in each of the 10 practice shots to the performance of your opponent in each of those 10 shots, taking into account all the possible combinations of these shots. The shots that have not been done in the allotted time are not included in this calculation. This calculation will tell you what is the theoretical number of chances out of 100 that you should win the point on your opponent for a given shot.
Your performance during these test and practice shots will not be taken into account in calculating the points in the actual game.
End of the session At the end of the game, you will see on your screen a final questionnaire. Once you have completed this questionnaire, we ask you to wait quietly until the end of the session without communicating with other participants. The session will end once all session participants have finished their games. To wait, you can read the magazines that are available on your desk or any document that you brought with you. You can also enjoy all the activities that will be available on your computer screen.
To sum up:
-
The task involves shooting at a target.
-
The winner of the game earns 20 Euros and the loser gets 0.
-
The winner of the game is the one of the two co-participants who won two sets.
-
To win a set, one must be the first co-participant to win 6 rounds.*
-
To win a round, one must be the first co-participant to win at least 4 points with at least 2 points ahead of his opponent.
Please read these instructions again and answer the questionnaire that was distributed. If you have any question, please raise your hand and we will answer your questions in private. Once you have completed the questionnaire, raise your hand, so we can come and check your answers (see Tables 9, 10, and 11).
1.4 Questionnaire
Thank you for answering these questions to familiarize yourself with the rules of the experiment.
Propositions | Yes | No |
|---|---|---|
[Robot treatment: You play against another person present in the room.] | O | O |
In a round, imagine the current score: | ||
3 points for participant 1 and 2 points for participant 2 | ||
Does participant 1 win the round by winning the next point? | O | O |
Does participant 2 win the round by winning the next point? | O | O |
Does a participant win the set as soon as the participant | ||
has won one round more than the opponent? | O | O |
The game ends as soon as a participant has won two sets | O | O |
The earnings of each set are 20 Euros | O | O |
There may be three sets in the game | O | O |
There may be four sets in the game | O | O |
Rights and permissions
About this article
Cite this article
Houy, N., Nicolaï, JP. & Villeval, M. Always doing your best? Effort and performance in dynamic settings. Theory Decis 89, 249–286 (2020). https://doi.org/10.1007/s11238-020-09752-6
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11238-020-09752-6