Ibrahim (2022) introduced 'conditional defector' strategies — cheaters that selectively cooperate, either by sharing the costs of collective dispersal ('cooperate for the spread') or by paying costs to restrain or mask their greed so as to evade policing ('pay for the escape') — and reported that in two agent-based NetLogo models such cheaters routinely breach the three canonical mechanisms thought to maintain cooperation (kin selection, spatial structure, and punishment) and drive cooperators extinct. That demonstration rests on two specific educational models, particular parameter settings, and a simulation grid in which cheater greed and monitor accuracy were deliberately varied in positive tandem, and it left the cheaters' own post-invasion fate unsimulated. The open problem is to establish whether conditional defector strategies genuinely violate the canonical cooperation-maintaining mechanisms beyond this original demonstration: under what conditions can they invade cooperative populations, including regimes where greed and monitoring accuracy vary independently, what is their long-term fate after cooperator extinction, and could engineered conditional defectors overcome the cooperator defense mechanisms that currently limit cheater-based therapies against pathogens and tumors.

Background

Cooperation — paying a personal cost to produce a shared benefit, or 'public good' — is ubiquitous across life but is intrinsically vulnerable to cheaters (also called defectors or free riders) that reap the shared benefit without paying the cost. Because unfettered cheating drives the collective output down until everyone is worse off (the 'tragedy of the commons'), evolutionary biology has built its explanation of stable cooperation on a short list of supporting mechanisms: (i) kin selection, in which help is directed at relatives so that the genes underlying cooperation help copies of themselves (Hamilton's rule: help spreads when the relatedness between actor and beneficiary, times the benefit delivered, exceeds the cost paid); (ii) spatial structure (assortment), in which limited dispersal makes cooperators interact preferentially with other cooperators; and (iii) enforcement or punishment, in which policing individuals pay a cost to detect and sanction cheaters (policing itself being vulnerable to 'second-order free riders' that enjoy the policing of others without paying for it).

Ibrahim (2022) asked what happens when the cheaters are not pure defectors but 'conditional defectors': cheaters that cooperate selectively, only where doing so serves their own spread or survival. Two forms were implemented on two educational agent-based NetLogo models of S. Hanisch ('Evolution and patchy resource' and 'Evolution, resources, monitoring and punishment'; the modified code is archived at the CoMSES Computational Model Library). In the first, 'cooperate for the spread', cheaters share the otherwise prohibitive costs of dispersing between resource patches with same-type flock-mates, so that costly dispersal — normally a cooperative act cheaters are bad at — becomes affordable for them; by migrating in groups they convert a structured metapopulation into an effectively well-mixed one and defeat the spatial-structure mechanism (cheaters took over whenever the group dispersal range was 30 patches or more). In the second, 'pay for the escape', cheaters operating under policing restrict or disguise their harvest so as to stay below the detection and punishment threshold; across three punishment types (temporary harvest suspension, fines, and killing) the greedy non-punishing agents rose from an initial 1% frequency to dominate the population. Crucially, the punishment-model experiments varied cheater greed and punisher perception accuracy only as positively coupled pairs (greedy harvest 15 down to 9 as perception accuracy fell from 99% to 30%; the reported correlation coefficient r = 0.99 is the correlation between these experimenter-chosen parameter values), and the source itself presents this positive correlation as an assumption under which the 'pay for the escape' strategy operates — the cheater reduces its profit 'to escape punishment by reducing perception accuracy if there is a positive correlation between these two variables. Therefore, this model is appropriate if it can support/deny such a correlation' — so cheater dominance was never tested at uncoupled or negatively coupled combinations of greed and monitoring. The paper further proposed, but did not simulate, that after cooperators go extinct the conditional defectors themselves die out because they can no longer subsist on the public good — an invocation of the well-established 'evolutionary suicide' (adaptation-driven extinction) prediction rather than a result of the models — and suggested engineering conditional defectors as 'suicidal agents' to collapse populations of cooperating pathogens or cancer cells.

Several adjacent literatures bear on this claim without answering it. On the cooperator side, conditional strategies are well studied (for example conditional punishment in spatial public goods games, Szolnoki & Perc 2012), and zero-determinant 'extortion' strategies in the iterated prisoner's dilemma — which unilaterally set the opponent's payoff — are known to be evolutionarily unstable in adapting populations (Adami & Hintze 2013; Stewart & Plotkin 2014); Ibrahim's Discussion itself notes that a tag-based zero-determinant extortioner would count as a conditional defector. Cheaters that display a cooperative signal without paying for it ('falsebeards', subverting 'green-beard' recognition systems in which a perceptible marker, its recognition, and help directed at its bearers are genetically linked) were long hypothetical, but Lehmann et al. (2024) recently produced an empirical Dictyostelium discoideum falsebeard (cells lacking rapgapB, which keep the allorecognition signal yet avoid the altruistic stalk fate), showing that tag-based cheating exists in nature. On the therapeutic side, cheater-based 'Trojan horse' interventions have a lineage: Brown et al. (2009) proposed exploiting microbial sociality, Archetti (2013, 2021) tested engineered defector cells (growth-factor non-producers) that collapse intra-tumor cooperation in vitro with partial success, Domingo-Calap et al. (2019) showed a defector virus strain defeating wild type via a tragedy of the commons, and Mutlu et al. (2024) delivered an experimental proof of principle in which quorum-sensing cheats invade resistant Pseudomonas aeruginosa populations and restore antibiotic sensitivity in vitro and in vivo; Rast et al. (2016) showed for interfering-particle antivirals how conflicting selection pressures constrain such invaders. All of these therapeutic studies use unconditional defectors, and Ibrahim's specific assertion — that conditional defectors, by mimicking or rewarding their enforcers, could overcome precisely the cooperator defense mechanisms that limit these therapies — remains untested. Likewise, classic joint-evolution theory (Parvinen 2013 on the co-evolution of altruism and dispersal; Ridley's 2012 field evidence for coalition dispersal in a cooperative bird; Dobata et al. 2011 on a transmissible 'social cancer' lineage in ants whose costly dispersal prevents population collapse) frames dispersal–cooperation links but not cheater-financed collective dispersal.

Problem Statement

Determine whether conditional defector strategies — cheater strategies that cooperate selectively, exemplified by (a) cheaters that share dispersal costs among their own kind so as to migrate collectively and dissolve the spatial assortment protecting cooperators ('cooperate for the spread'), and (b) cheaters that pay a cost, by restraining harvest or displaying cooperative markers, to stay below or defeat the detection threshold of policing agents ('pay for the escape') — can genuinely invade cooperative populations and drive cooperators extinct, beyond the two specific agent-based NetLogo demonstrations in which this was reported. The overarching objective is to establish the generality of the claim that the canonical mechanisms maintaining cooperation (kin selection, spatial structure, and punishment) are violable by such strategies, and it comprises four linked questions: (1) Robustness and generality: does conditional-defector invasion survive alternative model structures, the broader space of conditional strategies (those in which a cheater's cooperation is conditioned on a cue bearing on its own spread or survival) of which the two published forms are instances, and — as a direct test of the correlation assumption that the source itself flags as needing support or denial — parameter regimes in which cheater greed and monitor perception accuracy vary independently rather than as positively coupled pairs? (2) Invasion conditions: under what explicit conditions (cost and benefit structure, relatedness, dispersal-cost sharing rules, monitoring accuracy, punishment severity) can conditional defectors invade a cooperative population and exclude cooperators, and when do the canonical mechanisms hold? (3) Long-term fate: after cooperators are eliminated, do conditional defector populations subsequently collapse for lack of the public good (evolutionary suicide), as asserted from external theory but never simulated, or can they persist, for example by degraded or residual cooperation? (4) Applied consequence: can engineered conditional defectors invade cooperating pathogen or tumor populations despite the target population's own defense mechanisms — the obstacle that currently limits unconditional-defector (Trojan horse) therapies? All four linked questions are jointly required for a resolution, and an answer to a proper subset does not resolve the task. (1) and (2) are not independent: a complete characterization of the invasion conditions in (2) subsumes (1)'s parameter-regime test, because that test varies the same greed and monitoring-accuracy parameters over which (2) ranges, so a single analysis may discharge both, while (1)'s tests of alternative model structures and of the broader strategy space remain distinct requirements alongside (3) and (4).

References

  1. Ahmed M. Ibrahim (2022), 'The conditional defector strategies can violate the most crucial supporting mechanisms of cooperation', Scientific Reports 12:15157. DOI: 10.1038/s41598-022-18797-2. URL: https://doi.org/10.1038/s41598-022-18797-2
  2. Ahmed M. Ibrahim (2022), 'The conditional defector strategies can violate the most crucial supporting mechanisms of cooperation', bioRxiv preprint, posted 2022-06-09. DOI: 10.1101/2022.06.07.495117. URL: https://doi.org/10.1101/2022.06.07.495117
  3. Ahmed M. Ibrahim (2022), 'The conditional defector strategy can violate the most crucial supporting mechanisms of cooperation', NetLogo model code, CoMSES Computational Model Library. URL: https://www.comses.net/codebases/8437728b-3e1f-46f3-80be-65f4f5909d81/releases/1.0.0/
  4. Eduardo V. Stock, Pablo Valverde, Juan Carlos González-Avella, José Roberto Iglesias, Sebastian Gonçalves, Roberto da Silva (2025), 'Cyclical oscillations and absorbing-state probabilities in optional public goods games: Interplay of reward and group size', Physical Review E 111:014138. DOI: 10.1103/PhysRevE.111.014138. URL: https://doi.org/10.1103/PhysRevE.111.014138
  5. Alper Mutlu, Emily J. Vanderpool, Kendra P. Rumbaugh, Stephen P. Diggle, Ashleigh S. Griffin (2024), 'Exploiting cooperative pathogen behaviour for enhanced antibiotic potency: A Trojan horse approach', Microbiology 170:001454. DOI: 10.1099/mic.0.001454. URL: https://doi.org/10.1099/mic.0.001454
  6. Peter Lehmann, Mariko Katoh-Kurasawa, Peter Kundert, Gad Shaulsky (2024), 'Going against the family: Perturbation of a greenbeard pathway leads to falsebeard cheating', iScience 27:111125. DOI: 10.1016/j.isci.2024.111125. URL: https://doi.org/10.1016/j.isci.2024.111125
  7. Marco Archetti (2021), 'Collapse of Intra-Tumor Cooperation Induced by Engineered Defector Cells', Cancers 13:3674. DOI: 10.3390/cancers13153674. URL: https://doi.org/10.3390/cancers13153674
  8. Marco Archetti (2013), 'Evolutionarily Stable Anti-Cancer Therapies by Autologous Cell Defection', Evolution, Medicine, and Public Health 2013:161-172. DOI: 10.1093/emph/eot014. URL: https://doi.org/10.1093/emph/eot014
  9. Marco Archetti, Kenneth J. Pienta (2019), 'Cooperation among cancer cells: applying game theory to cancer', Nature Reviews Cancer 19:110-117. DOI: 10.1038/s41568-018-0083-7. URL: https://doi.org/10.1038/s41568-018-0083-7
  10. Sam P. Brown, Stuart A. West, Stephen P. Diggle, Ashleigh S. Griffin (2009), 'Social evolution in micro-organisms and a Trojan horse approach to medical intervention strategies', Philosophical Transactions of the Royal Society B 364:3157-3168. DOI: 10.1098/rstb.2009.0055. URL: https://doi.org/10.1098/rstb.2009.0055
  11. Luke I. Rast, Igor M. Rouzine, Ganna Rozhnova, Lisa Bishop, Ariel D. Weinberger, Leor S. Weinberger (2016), 'Conflicting Selection Pressures Will Constrain Viral Escape from Interfering Particles: Principles for Designing Resistance-Proof Antivirals', PLOS Computational Biology 12:e1004799. DOI: 10.1371/journal.pcbi.1004799. URL: https://doi.org/10.1371/journal.pcbi.1004799
  12. Pilar Domingo-Calap, Ernesto Segredo-Otero, María Durán-Moreno, Rafael Sanjuán (2019), 'Social evolution of innate immunity evasion in a virus', Nature Microbiology 4:1006-1013. DOI: 10.1038/s41564-019-0379-8. URL: https://doi.org/10.1038/s41564-019-0379-8
  13. Christoph Adami, Arend Hintze (2013), 'Evolutionary instability of zero-determinant strategies demonstrates that winning is not everything', Nature Communications 4:2193. DOI: 10.1038/ncomms3193. URL: https://doi.org/10.1038/ncomms3193
  14. Alexander J. Stewart, Joshua B. Plotkin (2014), 'Collapse of cooperation in evolving games', Proceedings of the National Academy of Sciences 111:17558-17563. DOI: 10.1073/pnas.1408618111. URL: https://doi.org/10.1073/pnas.1408618111
  15. Attila Szolnoki, Matjaž Perc (2012), 'Conditional strategies and the evolution of cooperation in spatial public goods games', Physical Review E 85:026104. DOI: 10.1103/PhysRevE.85.026104. URL: https://doi.org/10.1103/PhysRevE.85.026104
  16. Kalle Parvinen (2013), 'Joint evolution of altruistic cooperation and dispersal in a metapopulation of small local populations', Theoretical Population Biology 85:12-19. DOI: 10.1016/j.tpb.2013.01.003. URL: https://doi.org/10.1016/j.tpb.2013.01.003
  17. Amanda R. Ridley (2012), 'Invading together: the benefits of coalition dispersal in a cooperative bird', Behavioral Ecology and Sociobiology 66:77-83. DOI: 10.1007/s00265-011-1255-6. URL: https://doi.org/10.1007/s00265-011-1255-6
  18. S. Dobata, T. Sasaki, H. Mori, E. Hasegawa, M. Shimada, K. Tsuji (2011), 'Persistence of the single lineage of transmissible "social cancer" in an asexual ant', Molecular Ecology 20:441-455. DOI: 10.1111/j.1365-294X.2010.04954.x. URL: https://doi.org/10.1111/j.1365-294X.2010.04954.x