oTree is a framework based on Python and Django that lets you build:
- Multiplayer strategy games, like the prisoner's dilemma, public goods game, and auctions
- Controlled behavioral experiments in economics, psychology, and related fields
- Surveys and quizzes
Rather than cloning this repo directly, run these commands:
pip3 install -U otree-core
otree startproject oTree
otree resetdb
otree runserver
Below is a full implementation of the Guess 2/3 of the average game, where everyone guesses a number, and the winner is the person closest to 2/3 of the average. The game is repeated for 3 rounds. You can play the below game here.
from otree.api import (
models, widgets, BaseConstants, BaseSubsession, BaseGroup, BasePlayer,
Currency
)
class Constants(BaseConstants):
players_per_group = 3
num_rounds = 3
name_in_url = 'guess_two_thirds'
jackpot = Currency(100)
guess_max = 100
class Subsession(BaseSubsession):
pass
class Group(BaseGroup):
two_thirds_avg = models.FloatField()
best_guess = models.PositiveIntegerField()
num_winners = models.PositiveIntegerField()
def set_payoffs(self):
players = self.get_players()
guesses = [p.guess for p in players]
two_thirds_avg = (2 / 3) * sum(guesses) / len(players)
self.two_thirds_avg = round(two_thirds_avg, 2)
self.best_guess = min(guesses,
key=lambda guess: abs(guess - self.two_thirds_avg))
winners = [p for p in players if p.guess == self.best_guess]
self.num_winners = len(winners)
for p in winners:
p.is_winner = True
p.payoff = Constants.jackpot / self.num_winners
def two_thirds_avg_history(self):
return [g.two_thirds_avg for g in self.in_previous_rounds()]
class Player(BasePlayer):
guess = models.PositiveIntegerField(max=Constants.guess_max)
is_winner = models.BooleanField(initial=False)from . import models
from otree.api import Page, WaitPage
class Introduction(Page):
def is_displayed(self):
return self.round_number == 1
class Guess(Page):
form_model = models.Player
form_fields = ['guess']
class ResultsWaitPage(WaitPage):
def after_all_players_arrive(self):
self.group.set_payoffs()
class Results(Page):
def vars_for_template(self):
sorted_guesses = sorted(p.guess for p in self.group.get_players())
return {'sorted_guesses': sorted_guesses}
page_sequence = [Introduction,
Guess,
ResultsWaitPage,
Results]Instructions.html Introduction.html Guess.html Results.html
Test bots for multiplayer games run in parallel, and can run either from the command line, or in the browser, which you can try here.
from otree.api import Bot, SubmissionMustFail
from . import views
from .models import Constants
class PlayerBot(Bot):
cases = ['p1_wins', 'p1_and_p2_win']
def play_round(self):
if self.subsession.round_number == 1:
yield (views.Introduction)
if self.case == 'p1_wins':
if self.player.id_in_group == 1:
for invalid_guess in [-1, 101]:
yield SubmissionMustFail(views.Guess, {"guess": invalid_guess})
yield (views.Guess, {"guess": 9})
assert self.player.payoff == Constants.jackpot
assert 'you win' in self.html
else:
yield (views.Guess, {"guess": 10})
assert self.player.payoff == 0
assert 'you did not win' in self.html
else:
if self.player.id_in_group in [1, 2]:
yield (views.Guess, {"guess": 9})
assert self.player.payoff == Constants.jackpot / 2
assert 'you are one of the 2 winners' in self.html
else:
yield (views.Guess, {"guess": 10})
assert self.player.payoff == 0
assert 'you did not win' in self.html
yield (views.Results)See docs on bots.
- Extensive admin interface for launching games & surveys, managing participants, monitoring data, etc.
- Flexible API, e.g. for group re-matching
- Publish your games to Amazon Mechanical Turk
Help & discussion mailing list
Contact chris@otree.org with bug reports.
- Gregor Muellegger (http://gremu.net/, https://github.com/gregmuellegger)
- Juan B. Cabral (http://jbcabral.org/, https://github.com/leliel12)
- Bertrand Bordage (https://github.com/BertrandBordage)
- Alexander Schepanovski (https://github.com/Suor/)
- Alexander Sandukovskiy
- Som Datye
The oTree core libraries are here.