SUBMISSIONS
> SUBMISSION PORTAL & DOCUMENTATION
[ 01 ]
SUBMISSION FORM
-> [ 02 ]STRATEGY TEMPLATE
-> [ 03 ]PAYOFF MATRICES
-> [ 04 ]REFERENCE STRATEGIES
->Submission Rights and Consent Notice
By submitting through the official form, you confirm that the strategy is your original work and that you have the right to publish it.
- You retain ownership of your submitted strategy and related content.
- You grant Turing Club/core organizers a non-exclusive license to host, evaluate, archive, and publish your submission for tournament operations and public research access.
- Published submission data will be shared under CC BY-NC 4.0 for research and educational reuse with attribution to the creator and Iteron.
- Similar strategies may be independently created; where independent creation is reasonably established, credit may be granted to all independent creators.
- Plagiarism or copyright claims must include sufficient supporting proof for organizer review.
Read the full Terms of Use, Privacy Policy, and Takedown Policy before submitting.
OFFICIAL RULEBOOK
OVERVIEW
General Information
A round-robin computational tournament designed to evaluate strategies across a variety of strategic environments using cumulative payoffs. Inspired by Robert Axelrod's Iterated Prisoner's Dilemma Tournament, this event tests the robustness of strategies across varying environments rather than a single static payoff matrix.
- Format: Round-robin computational tournament, conducted online.
- Eligibility: Open to all students and faculty affiliated with any recognized university.
- Deadline: Submission deadline is May 11th 2026 11:59 PM IST.
SECTION_01
Introduction
Welcome to the Iteron. Participants are required to submit code-based strategies to compete against one another. Unlike the conventional tournament, this event tests the robustness of strategies across varying environments.
SECTION_02
Strategy Originality & Structure
- Every submission must be original. Pre-made strategies and near-clones are prohibited.
- Your strategy must not be behaviourally similar to strategies from the Axelrod Strategy Index, and it must also be distinct from Tit For Tat.
- Similarity checks are behavioural: changing variable names, comments, or code structure alone does not make a strategy original.
- See the organizer-maintained reference list at /submissions/reference-strategies for quick checks before submission.
SECTION_03
Formatting & Templates
- Strategies must be submitted using the official template provided by the organizers.
- Submissions in any other format or structure will be rejected. Participants may be asked to resubmit in the correct format or face immediate disqualification.
SECTION_04
Testing
- To prevent runtime errors during the actual tournament, the organizers have released a Sample Strategy.
- You are required to test your code against this sample strategy before submission.
- Any additional information regarding the same will be communicated via mail.
SECTION_05
Technical Constraints
- Machine Learning: The use of ML within strategies is expected to be limited. Strategies relying heavily on computationally expensive models may be subject to review regarding time-complexity limits.
SECTION_06
Tournament Structure
- The tournament will be conducted in a Round Robin format.
- Every submitted strategy will play against every other opponent's strategy.
- Every submitted strategy will also play against a copy of itself.
SECTION_07
Match Duration & Iteration
- To preserve the game-theoretic essence of the "Shadow of the Future," the exact number of rounds per match will not be revealed to participants.
- Strategies should be designed to handle an indefinite time horizon.
- The entire tournament will be iterated more than once to ensure statistical significance, remove noise, and guarantee fair observations of strategy performance.
SECTION_08
The Multi-Environment System
- Unlike the conventional Axelrod's tournament which uses a static payoff matrix, this tournament will be conducted across multiple environments.
- The organizers will systematically vary the payoff matrices — the rewards for Cooperation and Defection — between tournament runs.
- Strategies must be robust enough to adapt to changes in incentives, rather than being over-fitted to a single payoff structure.
- Each submitted strategy will be evaluated in all organizer-provided payoff environments.
- During evaluation, strategies will not be told which payoff environment is currently active.
- Any attempt to branch behavior based on detecting or hard-coding a specific payoff environment is not allowed.
- Detailed payoff matrices are available at /submissions/payoff.
SECTION_09
Winner Determination & Aggregate Performance
- The final winner will be determined based on the highest average score.
- For every payoff environment, each strategy will be run multiple times; the resulting score for that environment is the average of those runs.
- The final score is then computed by averaging performance across all provided payoff environments.
- Consistency is key. Even if a strategy places 1st in k out of n specific matrix environments, it will receive no special consideration if its overall average drops due to poor performance in other environments.
- Final ranking is based solely on overall aggregated performance. No separate claims or titles will be awarded for individual sub-tournament wins.
- The ITERON team reserves the right to modify these rules to ensure the smooth operation of the tournament. Updates will be communicated via official channels.