The generality of the phenomenon is the matter of some debate and disagreement, and Zahavi's views on the scope and importance of handicaps in biology has not been accepted by the mainstream. Nevertheless, the idea has been very influential, with most researchers in the field believing that the theory explains some aspects of animal communication.
Though the handicap principle was initially controversial,—John Maynard Smith being one notable early critic of Zahavi's ideas—it has gained wider acceptance because it is supported by game theoretic models, most notably Alan Grafen's signalling game model. Grafen's model is essentially a rediscovery of Michael Spence's job market signalling model, where the signalled trait was conceived as a courting male's quality, signalled by investment in an extravagant trait—such as the peacock's tail—rather than an employee signalling their quality by way of a costly education. In both cases, it is the decreased cost to higher-quality signallers of producing increased signal that stabilizes the reliability of the signal (Fig. 2).
A series of papers by Thomas Getty shows that Grafen’s proof of the handicap principle depends on the critical simplifying assumption that signalers trade off costs for benefits in an additive fashion, the way humans invest money to increase income in the same currency. This is illustrated in the figures to the right, from Johnstone 1997. This assumption that costs and benefits trade off in an additive fashion is contested to be valid for the survival cost–reproduction benefit trade-off that is assumed to mediate the evolution of sexually selected signals. It can be reasoned that as fitness depends on the production of offspring, this is therefore a multiplicative function of reproductive success.
Further formal game theoretical signalling models demonstrated the evolutionary stability of handicapped signals in nestlings' begging calls, in predator-deterrent signals and in threat-displays. In the classic handicapped models of begging, all players are assumed to pay the same amount to produce a signal of a given level of intensity, but differ in the relative value of eliciting the desired response (donation) from the receiver (Fig. 3). The hungrier the baby bird, the more food is of value to it, and the higher the optimal signalling level (the louder its chirping).
Counter-examples to handicap models predate handicap models themselves. Models of signals (such as threat displays) without any handicapping costs show that conventional signalling may be evolutionarily stable in biological communication. Analysis of some begging models also shows that, in addition to the handicapped outcomes, non-communication strategies are not only evolutionarily stable, but lead to higher payoffs for both players. Mathematical analyses including Monte Carlo simulations suggest that costly traits used in mate choice by humans should be generally less common and more attractive to the other sex than non-costly traits.