Chico Khan-Gandapur is a managing partner at Metrica Consulting.
The Conservative Party’s current polling predicament presents one of the most analytically challenging puzzles in modern British political history.
A historically unpopular Labour government, until a few weeks ago led by the joint least popular Prime Minister on record, is presiding over a stagnant economy, a cost-of-living crisis, and the aftermath of its worst council-election performance. Yet the official Conservative Opposition is polling at barely 20% (Politico’s Poll of Polls, August 12, 2026), still below its catastrophic 2024 general election result of 23.6%, and even below the current Labour government itself, benefitting from a modest Burnham Blip.
Traditional political analysis struggles to explain this conundrum. Game Theory, however, offers a credible and strategically useful lens.
But what is game theory ? Defined simply, it is the mathematical and strategic study of decision-making in situations where the outcome for each participant depends not only on their own choices, but also on the choices of others. It analyses how rational (non-emotional) agents, the players, behave when their decisions are interdependent.
It was formally established by John von Neumann and Oskar Morgenstern in their 1944 work Theory of Games and Economic Behavior, and developed further by John Nash, who won the Nobel Prize in Economics in 1994, and whose life is depicted in the film, A Beautiful Mind.
The core idea is that, in ordinary decision-making, we choose the best option given fixed circumstances. In game theory, however, the circumstances themselves change depending on what others might choose. We must, therefore, anticipate the choices of others, while they simultaneously anticipate ours.
The classic question game theory asks is: “What is the best strategy, given that everyone else is also choosing their own best strategy ?”
It has been likened to multiple chess players playing dozens of chess games simultaneously, where the boards are connected so that moves on one board affect all the others; where some games are cooperative rather than competitive; where, on some boards, you cannot see all the pieces; where your opponents can talk to each other and form alliances between games; where your reputation from past games follows you to every new board; and where, occasionally, a sufficiently powerful player can rewrite the rules of one or more games entirely. Got it ? Simples.
Through this lens of strategic interdependence, voter behaviour becomes far more comprehensible, predictable, and, crucially, navigable. This essay applies game theory’s core frameworks to diagnose why voters remain turned off by the Conservatives, and what the Party can do to restructure the strategic environment back in its favour.
- The Credible Commitment Challenge
The most fundamental obstacle facing the Conservative Party is what game theorists call The Credible Commitment Problem.
In any repeated game, meaning one played multiple times between the same participants, a player’s future promises carry weight only if their past behaviour makes those promises believable. Credibility is not asserted; it must be earned through a track record of solid delivery.
After 14 years of very mixed government, the Conservative Party has established for itself a reputation for Defection, or non-delivery, in its repeated game with the British electorate.
Commitments on economic stability were abandoned catastrophically by Liz Truss in 2022. Promises on immigration were repeatedly made and repeatedly broken. Pledges on public services went unfulfilled. Impropriety manifested itself through Partygate, PPE-gate and Pincher-gate.
In game-theory terms, voters have rationally updated their prior beliefs about Conservative behaviour. Those beliefs now assign a very low probability to any new Conservative commitment being honoured in the way it is conveyed.
The devastating consequence for the Party is that it cannot escape this trap through better promises alone. Voters have been thoroughly conditioned, through repeated experience, rather like Pavlov’s dog, to believe that the payoff benefits of trusting Conservative promises are unfavourable. Until the underlying game structure changes, meaning until there are mechanisms that make future defection, or failure to deliver, genuinely costly for Conservatives, no amount of rhetorical renewal will shift voter beliefs.
Any remediation must therefore be structural, not merely communicative. Conservatives must create credible commitment devices: binding, publicly observable, and independently verified frameworks that structurally constrain future behaviour, similar to the fiscal rules that constrain Labour’s spending.
Published delivery scorecards, independent audits of policy pledges, and transparent accountability mechanisms change the game itself rather than merely reasserting trustworthiness within the existing game. A party that visibly binds itself to its commitments signals something qualitatively and strategically different from one that simply makes promises, even if it makes them slightly more sincerely.
- The Folk Theorem Challenge
In repeated games, The Folk Theorem establishes that cooperative outcomes can be sustained only when players adopt credible punishment strategies: withdrawing cooperation after a partner defects, and resuming it only after sufficient evidence of changed behaviour has emerged. This is not merely theoretical. It describes precisely how millions of British voters are currently behaving toward the Conservative Party.
In 2019, Johnson won the election with 13.96 million votes, a 43.6% share. Current polling at 20%, using Politico’s Poll of Polls, suggests that just over 6 million voters would support the Party now, a collapse of around 7.5 million.
The 2022 mini-budget was not simply a policy error. It was felt by voters as a further catastrophic defection in a long-running repeated game between the Conservative Party and its core supporters. Homeowners who had trusted the Party with their financial security saw mortgage rates spike almost overnight. The value of the pound collapsed, and pension funds risked severe drawdowns because of the impact on Gilt prices. The pain was direct, personal, and measurable, with money leaving bank accounts and pension values falling in real time.
This problem was compounded by Kahneman’s Loss-Aversion Principle, where losses are felt approximately twice as intensely as equivalent gains. The emotional weight of the 2022 defection was therefore even greater than its objective financial impact.
Voters have responded precisely as Folk Theorem analysis would predict. They have implemented a punishment strategy by withdrawing their cooperation, in this case their votes, and signaling these votes will not be restored until a sufficient punishment phase has elapsed, and genuine Conservative behavioural change has been demonstrated.
The critical game-theory insight is the duration of the punishment phase is determined by the severity of the original defection(s), not by the quality of the offender’s subsequent apologies.
The 2022 mini-budget was a malign defection against the Party’s most loyal supporters. The punishment phase will therefore be long. Attempts to shortcut it through aggressive campaigning or rhetorical renewal will be perceived as further defection: an effort to escape the consequences of past behaviour rather than acknowledge them. Such attempts are likely to extend, rather than shorten, the punishment period.
The strategic implication is clear. Conservatives must accept and endure the punishment phase rather than resist it. There is, however, some hope. Badenoch’s non-involvement in the 2022 Cabinet represents a genuine structural asset. In the language of game theory, she is a new player entering the repeated game, and her personal payoff history is unblemished by the original defection. She should deploy this positioning more aggressively, drawing an explicit and visible line between herself as a new player, and the Party’s historical defectors.
- The Multi-Player Coordination Problem: A Broken Binary
Historical two-party electoral politics approximated to a zero-sum game as one party’s gain was simply the other’s loss, and voters faced a simple binary choice. The strategic logic was straightforward. If you disliked the governing party, you voted for the opposition.
The emergence of Reform, the Liberal Democrats, the Greens and, to a lesser extent, Restore, alongside the nationalist parties as credible alternatives, have fundamentally transformed British politics into a multi-player, non-zero-sum game in which the old binary logic no longer applies.
In a two-player game, a voter dissatisfied with Labour has only one real rational alternative destination, the Conservatives.
But in a four-, five- or six-plus-player game, that same voter can defect to whichever alternative most precisely matches their specific interests, concerns or grievances: Reform for immigration and cultural conservatism; the Liberal Democrats for slightly leftist economic credibility; the Greens for environmental and progressive concerns; the SNP for independence, and so on.
This produces what game theorists call Fragmented Equilibria: stable distributions of voter support across multiple parties that are extremely difficult to reverse because each equilibrium is self-reinforcing. Reform voters feel their identity is validated by Reform; Liberal Democrat voters feel theirs is validated by the Liberal Democrats; and so on.
Breaking these equilibria requires a significant further disruption to the existing payoff structure, not merely a better offer within it.
The multi-player environment also creates a severe Coordination problem for Conservative recovery. Even voters who are sympathetic to returning to the Conservatives face a collective-action problem. Each individual voter’s return is worthwhile only if enough others simultaneously return to make the Conservative vote count.
A voter considering switching from Reform to Conservative in a marginal seat must estimate not only whether the Conservatives are the better party, but whether enough others will make the same switch to produce a useful outcome. This produces a coordination failure: individually rational caution leads collectively to Conservative vote fragmentation that no individual voter wants, but which all individual voters help perpetuate.
Resolving this requires the Conservatives to generate Focal Points, known in game theory as Schelling Points, which coordinate voter expectations around a credible Conservative recovery narrative. Visible momentum, credible polling trajectories, and the deliberate sequencing of coalition-building, for example Reform-leaners first and centrist defectors second, can create the expectation of mass return that makes individual return rational. Tangible examples could include regaining control of Westminster and Wandsworth councils, or winning the constituency of Aberdeen South.
- Signaling Theory: The Cost of Credibility
Michael Spence’s Signalling Theory, for which he received the Nobel Prize, addresses how players in games of asymmetric information communicate qualities which cannot be directly observed.
The Conservatives face a classic Signalling challenge. They need to communicate genuine transformation to voters who cannot directly verify whether the Party has really changed, because it is not in power and therefore cannot easily demonstrate that change nationally through government action: only local signalling is possible.
In such games, only Costly signals are seen as credible by players (voters). These are signals whose cost is high enough that only a genuinely transformed party would rationally bear them. Think of Solomon threatening to cut the baby in half to resolve a dispute between two women who each claimed to be the child’s mother. Only the real mother was willing to bear the cost of giving up the child in order to save its life.
To voters, the Conservative Party has so far generated mainly cheap signals: speeches, manifesto pledges and leadership changes. The costs of these are low enough that an unreformed party would produce them just as readily as a reformed one. Voters therefore correctly discount them.
What is required are signals whose cost is genuinely prohibitive for an unreformed party: actions so politically painful that only a party that had genuinely changed its underlying character would rationally take them.
Such signals might include publicly agreeing with Labour on a significant policy question; formally abandoning a previously central position in a way that risks alienating the base; adopting binding independent oversight of party finances and donor relationships; or elevating figures whose backgrounds and perspectives explicitly represent the communities the Party has historically failed to reach.
Each of these actions carries real political cost, which is precisely what makes them credible as signals of genuine change. A party willing to pay these costs is demonstrating something that cannot be faked cheaply.
In previous essays, I have suggested Reform of the House of Lords, which I strongly believe is warranted in any case, but which would also represent the kind of break from the past needed to bring voters back to the Party.
There are, however, tentative signs of hope. Badenoch’s improving personal ratings, from a net score of -32 to +1 over the last 10 months according to More In Common, suggest that her personal signalling is beginning to work. But the ongoing gap between her ratings and the Party’s ratings reflects the reality she has generated far more credible personal signals than the Party has generated institutional ones. Closing this gap requires the Party itself, not merely its leader, to begin producing costly signals, while ensuring that the leader does not lose the signals she has already built.
- The Prisoner’s Dilemma of Internal Discipline
The Conservative Party’s historic internal-discipline problem is structurally identical to the Prisoner’s Dilemma, arguably the most famous and consequential model in game theory.
Each individual Conservative MP faces a choice between Cooperation, meaning maintaining collective discipline and subordinating personal ambition to the Party’s recovery, or Defection, meaning pursuing their own media profile, factional interest, or individual political positioning.
Counterintuitively perhaps, the payoff structure strongly favours individual defection. An MP who breaks ranks gains personal publicity and signals ideological distinctiveness, while bearing only a fraction of the collective reputational cost their defection imposes on the Party as a whole. One can think of Robert Jenrick’s behaviour before he moved to Reform, and the subsequent improved perception of the Conservatives after he did.
The Nash Equilibrium of this game, meaning the stable outcome when every player pursues their individually rational strategy, is, perversely, persistent indiscipline. The consequence though is that every individual MP’s rational choice produces a collectively catastrophic outcome for the Party, which appears chaotic, ungovernable, and incapable of the discipline that responsible government requires: and this helps explain a great deal of Conservative behaviour between 2010 to 2024.
This is not a failure of individual character per se. It is a failure of game structure. The payoffs are arranged in a way that makes defection individually rational regardless of the collective consequences.
The only solution game theory offers to a Prisoner’s Dilemma is to change the payoff structure, making defection genuinely costly for the individual, not merely for the collective. This requires swift, visible, and consistent sanctions for public dissent, donor controversies, and factional behaviour that undermines the collective recovery project.
When the first high-profile enforcement changes an individual MP’s rational calculation, making cooperation more rewarding than defection, the equilibrium of the game can shift. Until enforcement happens, the Nash Equilibrium of persistent indiscipline will persist, regardless of how many collective calls for unity are issued. Kemi referred to this indirectly as “Spring Cleaning”. Recent examples include Gavin Bardwell having the whip removed from him, or Grant Schapps not being allowed to become a candidate MP.
- The Hawk-Dove Game: Competing With Reform
The strategic competition between the Conservatives and Reform UK for the Right side of British politics maps precisely onto the Hawk-Dove Game, one of game theory’s most powerful models of conflict between aggressive and restrained strategies.
Reform plays the Hawk: maximally aggressive, emotionally uninhibited, confrontational, and unconstrained. It must do so in order to gain initial traction. The challenge for Conservative strategists is whether to play Hawk in response, or to play Dove.
Game theory is unambiguous about the optimal strategy in Hawk-Dove interactions.
If the Conservatives play Hawk against Reform’s Hawk, they risk entering a destructive escalation contest on terrain where Reform has first-mover advantage, greater emotional authenticity, and a more established brand identity. Ironically, both parties suffer, and the primary beneficiary is Labour, which watches the Right destroy itself.
If the Conservatives play Dove, restrained, credible and governance-focused, they may concede the emotional short game, but they retain a chance of winning the strategic long game. The Nash Equilibrium of Hawk-Dove is not mutual aggression. It is one Hawk and one Dove, with the Dove ultimately inheriting the governing position when the Hawk’s aggressive strategy generates unsustainable costs.
Reform’s trust problems over Farage’s £5 million “G(r)ift” are evidence that the Hawk-Dove equilibrium is already emerging organically. Conservatives should not try to accelerate this through aggressive counter-attack, which would merely catalyse the Hawk-Hawk dynamic. Instead, they are better served by consolidating and deepening their Dove positioning, ensuring that when Reform’s Hawk strategy next generates costs, Conservatives are unambiguously the credible, stable, governing alternative. Let the left-leaning parties fight Farage for them.
- The Stag Hunt: Solving the Collective-Action Problem
The broadest structural challenge facing Conservative recovery is a Stag Hunt: a coordination game in which the best collective outcome requires simultaneous mutual commitment from multiple players who cannot directly verify one another’s intentions.
Rebuilding a Conservative majority requires voters, MPs, donors, local associations and media supporters to commit simultaneously to the recovery project. But each player is willing to commit only if they are sufficiently confident that others will also commit. In the absence of that confidence, each defaults to the individually safer option of staying with Reform, abstaining, or maintaining scepticism and inaction.
The Stag Hunt produces two equilibrium outcomes. The first is the high-level equilibrium, where everyone cooperates, the stag is caught, and the Conservative majority is rebuilt. The second is the low-level equilibrium, where insufficient confidence in others’ cooperation leads everyone to defect to safer individual options, and the Party remains trapped at 19% to 20% of the vote share.
The Party is currently locked in the low-level equilibrium. Breaking out requires generating sufficient common knowledge: not merely that each player believes in recovery, but that each player believes every other player believes in it, and also believes that every other player believes that every other player believes in it.
This recursive confidence is the precondition for coordination, and it cannot be generated by individual conviction alone.
The strategic implication is that the Conservative recovery programme must be explicitly designed as a coordination mechanism, creating visible, credible, and mutually reinforcing signals of recovery that make it individually rational for each additional player to join.
A properly resourced target-seats strategy, visible momentum in by-elections, credible polling trajectories, and the public commitment of key figures all serve this coordinating function. Paradoxically, a party that honestly acknowledges a two-election recovery horizon, rather than claiming implausible imminent success, generates more credible common knowledge because its forecast is consistent with observable reality.
Some comfort can also be taken from recent political volatility. Johnson’s 2019 victory was followed by a rapid reversal that produced Starmer’s historic win, which itself has collapsed within two years. We wait to see what a midnight blue T-shirt and eyebrows can deliver.
Conclusion: Restructuring the Game
Game-theory analysis converges on a single, powerful conclusion.
The Conservative Party is not failing simply because its policies are wrong, its leader is inadequate, or its communications are insufficiently polished.
It is failing because the strategic game structure itself is working against it.
Rational punishment strategies are being administered. Coordination failures are trapping potential returners. Cheap signals are being correctly discounted by voters. Internal Prisoner’s Dilemmas are producing collective indiscipline, albeit to a much lesser extent after Jenrick’s exit. And the Hawk-Dove dynamic with Reform, while shifting, has not yet resolved itself in the Party’s favour.
The path back does not lie in winning arguments within the existing game structure. It lies in restructuring the game itself through credible commitment devices that change voter beliefs; costly signals that communicate genuine transformation; coordination mechanisms that solve the collective-action problem; payoff restructuring that resolves the internal Prisoner’s Dilemma; and the patient acceptance of the punishment phase that the Folk Theorem demands.
As Kemi Badenoch said at her 2024 leadership launch: “In engineering, you can’t overcome resistance with rhetoric. You can’t get the machine to work with promises.”
Game theory provides the precise analytical framework that this engineering instinct requires, and its diagnosis is clear. The machine will work only when the structure of the game changes. That change begins not with a speech, but with the binding commitments above.
Chico Khan-Gandapur is a managing partner at Metrica Consulting.
The Conservative Party’s current polling predicament presents one of the most analytically challenging puzzles in modern British political history.
A historically unpopular Labour government, until a few weeks ago led by the joint least popular Prime Minister on record, is presiding over a stagnant economy, a cost-of-living crisis, and the aftermath of its worst council-election performance. Yet the official Conservative Opposition is polling at barely 20% (Politico’s Poll of Polls, August 12, 2026), still below its catastrophic 2024 general election result of 23.6%, and even below the current Labour government itself, benefitting from a modest Burnham Blip.
Traditional political analysis struggles to explain this conundrum. Game Theory, however, offers a credible and strategically useful lens.
But what is game theory ? Defined simply, it is the mathematical and strategic study of decision-making in situations where the outcome for each participant depends not only on their own choices, but also on the choices of others. It analyses how rational (non-emotional) agents, the players, behave when their decisions are interdependent.
It was formally established by John von Neumann and Oskar Morgenstern in their 1944 work Theory of Games and Economic Behavior, and developed further by John Nash, who won the Nobel Prize in Economics in 1994, and whose life is depicted in the film, A Beautiful Mind.
The core idea is that, in ordinary decision-making, we choose the best option given fixed circumstances. In game theory, however, the circumstances themselves change depending on what others might choose. We must, therefore, anticipate the choices of others, while they simultaneously anticipate ours.
The classic question game theory asks is: “What is the best strategy, given that everyone else is also choosing their own best strategy ?”
It has been likened to multiple chess players playing dozens of chess games simultaneously, where the boards are connected so that moves on one board affect all the others; where some games are cooperative rather than competitive; where, on some boards, you cannot see all the pieces; where your opponents can talk to each other and form alliances between games; where your reputation from past games follows you to every new board; and where, occasionally, a sufficiently powerful player can rewrite the rules of one or more games entirely. Got it ? Simples.
Through this lens of strategic interdependence, voter behaviour becomes far more comprehensible, predictable, and, crucially, navigable. This essay applies game theory’s core frameworks to diagnose why voters remain turned off by the Conservatives, and what the Party can do to restructure the strategic environment back in its favour.
The most fundamental obstacle facing the Conservative Party is what game theorists call The Credible Commitment Problem.
In any repeated game, meaning one played multiple times between the same participants, a player’s future promises carry weight only if their past behaviour makes those promises believable. Credibility is not asserted; it must be earned through a track record of solid delivery.
After 14 years of very mixed government, the Conservative Party has established for itself a reputation for Defection, or non-delivery, in its repeated game with the British electorate.
Commitments on economic stability were abandoned catastrophically by Liz Truss in 2022. Promises on immigration were repeatedly made and repeatedly broken. Pledges on public services went unfulfilled. Impropriety manifested itself through Partygate, PPE-gate and Pincher-gate.
In game-theory terms, voters have rationally updated their prior beliefs about Conservative behaviour. Those beliefs now assign a very low probability to any new Conservative commitment being honoured in the way it is conveyed.
The devastating consequence for the Party is that it cannot escape this trap through better promises alone. Voters have been thoroughly conditioned, through repeated experience, rather like Pavlov’s dog, to believe that the payoff benefits of trusting Conservative promises are unfavourable. Until the underlying game structure changes, meaning until there are mechanisms that make future defection, or failure to deliver, genuinely costly for Conservatives, no amount of rhetorical renewal will shift voter beliefs.
Any remediation must therefore be structural, not merely communicative. Conservatives must create credible commitment devices: binding, publicly observable, and independently verified frameworks that structurally constrain future behaviour, similar to the fiscal rules that constrain Labour’s spending.
Published delivery scorecards, independent audits of policy pledges, and transparent accountability mechanisms change the game itself rather than merely reasserting trustworthiness within the existing game. A party that visibly binds itself to its commitments signals something qualitatively and strategically different from one that simply makes promises, even if it makes them slightly more sincerely.
In repeated games, The Folk Theorem establishes that cooperative outcomes can be sustained only when players adopt credible punishment strategies: withdrawing cooperation after a partner defects, and resuming it only after sufficient evidence of changed behaviour has emerged. This is not merely theoretical. It describes precisely how millions of British voters are currently behaving toward the Conservative Party.
In 2019, Johnson won the election with 13.96 million votes, a 43.6% share. Current polling at 20%, using Politico’s Poll of Polls, suggests that just over 6 million voters would support the Party now, a collapse of around 7.5 million.
The 2022 mini-budget was not simply a policy error. It was felt by voters as a further catastrophic defection in a long-running repeated game between the Conservative Party and its core supporters. Homeowners who had trusted the Party with their financial security saw mortgage rates spike almost overnight. The value of the pound collapsed, and pension funds risked severe drawdowns because of the impact on Gilt prices. The pain was direct, personal, and measurable, with money leaving bank accounts and pension values falling in real time.
This problem was compounded by Kahneman’s Loss-Aversion Principle, where losses are felt approximately twice as intensely as equivalent gains. The emotional weight of the 2022 defection was therefore even greater than its objective financial impact.
Voters have responded precisely as Folk Theorem analysis would predict. They have implemented a punishment strategy by withdrawing their cooperation, in this case their votes, and signaling these votes will not be restored until a sufficient punishment phase has elapsed, and genuine Conservative behavioural change has been demonstrated.
The critical game-theory insight is the duration of the punishment phase is determined by the severity of the original defection(s), not by the quality of the offender’s subsequent apologies.
The 2022 mini-budget was a malign defection against the Party’s most loyal supporters. The punishment phase will therefore be long. Attempts to shortcut it through aggressive campaigning or rhetorical renewal will be perceived as further defection: an effort to escape the consequences of past behaviour rather than acknowledge them. Such attempts are likely to extend, rather than shorten, the punishment period.
The strategic implication is clear. Conservatives must accept and endure the punishment phase rather than resist it. There is, however, some hope. Badenoch’s non-involvement in the 2022 Cabinet represents a genuine structural asset. In the language of game theory, she is a new player entering the repeated game, and her personal payoff history is unblemished by the original defection. She should deploy this positioning more aggressively, drawing an explicit and visible line between herself as a new player, and the Party’s historical defectors.
Historical two-party electoral politics approximated to a zero-sum game as one party’s gain was simply the other’s loss, and voters faced a simple binary choice. The strategic logic was straightforward. If you disliked the governing party, you voted for the opposition.
The emergence of Reform, the Liberal Democrats, the Greens and, to a lesser extent, Restore, alongside the nationalist parties as credible alternatives, have fundamentally transformed British politics into a multi-player, non-zero-sum game in which the old binary logic no longer applies.
In a two-player game, a voter dissatisfied with Labour has only one real rational alternative destination, the Conservatives.
But in a four-, five- or six-plus-player game, that same voter can defect to whichever alternative most precisely matches their specific interests, concerns or grievances: Reform for immigration and cultural conservatism; the Liberal Democrats for slightly leftist economic credibility; the Greens for environmental and progressive concerns; the SNP for independence, and so on.
This produces what game theorists call Fragmented Equilibria: stable distributions of voter support across multiple parties that are extremely difficult to reverse because each equilibrium is self-reinforcing. Reform voters feel their identity is validated by Reform; Liberal Democrat voters feel theirs is validated by the Liberal Democrats; and so on.
Breaking these equilibria requires a significant further disruption to the existing payoff structure, not merely a better offer within it.
The multi-player environment also creates a severe Coordination problem for Conservative recovery. Even voters who are sympathetic to returning to the Conservatives face a collective-action problem. Each individual voter’s return is worthwhile only if enough others simultaneously return to make the Conservative vote count.
A voter considering switching from Reform to Conservative in a marginal seat must estimate not only whether the Conservatives are the better party, but whether enough others will make the same switch to produce a useful outcome. This produces a coordination failure: individually rational caution leads collectively to Conservative vote fragmentation that no individual voter wants, but which all individual voters help perpetuate.
Resolving this requires the Conservatives to generate Focal Points, known in game theory as Schelling Points, which coordinate voter expectations around a credible Conservative recovery narrative. Visible momentum, credible polling trajectories, and the deliberate sequencing of coalition-building, for example Reform-leaners first and centrist defectors second, can create the expectation of mass return that makes individual return rational. Tangible examples could include regaining control of Westminster and Wandsworth councils, or winning the constituency of Aberdeen South.
Michael Spence’s Signalling Theory, for which he received the Nobel Prize, addresses how players in games of asymmetric information communicate qualities which cannot be directly observed.
The Conservatives face a classic Signalling challenge. They need to communicate genuine transformation to voters who cannot directly verify whether the Party has really changed, because it is not in power and therefore cannot easily demonstrate that change nationally through government action: only local signalling is possible.
In such games, only Costly signals are seen as credible by players (voters). These are signals whose cost is high enough that only a genuinely transformed party would rationally bear them. Think of Solomon threatening to cut the baby in half to resolve a dispute between two women who each claimed to be the child’s mother. Only the real mother was willing to bear the cost of giving up the child in order to save its life.
To voters, the Conservative Party has so far generated mainly cheap signals: speeches, manifesto pledges and leadership changes. The costs of these are low enough that an unreformed party would produce them just as readily as a reformed one. Voters therefore correctly discount them.
What is required are signals whose cost is genuinely prohibitive for an unreformed party: actions so politically painful that only a party that had genuinely changed its underlying character would rationally take them.
Such signals might include publicly agreeing with Labour on a significant policy question; formally abandoning a previously central position in a way that risks alienating the base; adopting binding independent oversight of party finances and donor relationships; or elevating figures whose backgrounds and perspectives explicitly represent the communities the Party has historically failed to reach.
Each of these actions carries real political cost, which is precisely what makes them credible as signals of genuine change. A party willing to pay these costs is demonstrating something that cannot be faked cheaply.
In previous essays, I have suggested Reform of the House of Lords, which I strongly believe is warranted in any case, but which would also represent the kind of break from the past needed to bring voters back to the Party.
There are, however, tentative signs of hope. Badenoch’s improving personal ratings, from a net score of -32 to +1 over the last 10 months according to More In Common, suggest that her personal signalling is beginning to work. But the ongoing gap between her ratings and the Party’s ratings reflects the reality she has generated far more credible personal signals than the Party has generated institutional ones. Closing this gap requires the Party itself, not merely its leader, to begin producing costly signals, while ensuring that the leader does not lose the signals she has already built.
The Conservative Party’s historic internal-discipline problem is structurally identical to the Prisoner’s Dilemma, arguably the most famous and consequential model in game theory.
Each individual Conservative MP faces a choice between Cooperation, meaning maintaining collective discipline and subordinating personal ambition to the Party’s recovery, or Defection, meaning pursuing their own media profile, factional interest, or individual political positioning.
Counterintuitively perhaps, the payoff structure strongly favours individual defection. An MP who breaks ranks gains personal publicity and signals ideological distinctiveness, while bearing only a fraction of the collective reputational cost their defection imposes on the Party as a whole. One can think of Robert Jenrick’s behaviour before he moved to Reform, and the subsequent improved perception of the Conservatives after he did.
The Nash Equilibrium of this game, meaning the stable outcome when every player pursues their individually rational strategy, is, perversely, persistent indiscipline. The consequence though is that every individual MP’s rational choice produces a collectively catastrophic outcome for the Party, which appears chaotic, ungovernable, and incapable of the discipline that responsible government requires: and this helps explain a great deal of Conservative behaviour between 2010 to 2024.
This is not a failure of individual character per se. It is a failure of game structure. The payoffs are arranged in a way that makes defection individually rational regardless of the collective consequences.
The only solution game theory offers to a Prisoner’s Dilemma is to change the payoff structure, making defection genuinely costly for the individual, not merely for the collective. This requires swift, visible, and consistent sanctions for public dissent, donor controversies, and factional behaviour that undermines the collective recovery project.
When the first high-profile enforcement changes an individual MP’s rational calculation, making cooperation more rewarding than defection, the equilibrium of the game can shift. Until enforcement happens, the Nash Equilibrium of persistent indiscipline will persist, regardless of how many collective calls for unity are issued. Kemi referred to this indirectly as “Spring Cleaning”. Recent examples include Gavin Bardwell having the whip removed from him, or Grant Schapps not being allowed to become a candidate MP.
The strategic competition between the Conservatives and Reform UK for the Right side of British politics maps precisely onto the Hawk-Dove Game, one of game theory’s most powerful models of conflict between aggressive and restrained strategies.
Reform plays the Hawk: maximally aggressive, emotionally uninhibited, confrontational, and unconstrained. It must do so in order to gain initial traction. The challenge for Conservative strategists is whether to play Hawk in response, or to play Dove.
Game theory is unambiguous about the optimal strategy in Hawk-Dove interactions.
If the Conservatives play Hawk against Reform’s Hawk, they risk entering a destructive escalation contest on terrain where Reform has first-mover advantage, greater emotional authenticity, and a more established brand identity. Ironically, both parties suffer, and the primary beneficiary is Labour, which watches the Right destroy itself.
If the Conservatives play Dove, restrained, credible and governance-focused, they may concede the emotional short game, but they retain a chance of winning the strategic long game. The Nash Equilibrium of Hawk-Dove is not mutual aggression. It is one Hawk and one Dove, with the Dove ultimately inheriting the governing position when the Hawk’s aggressive strategy generates unsustainable costs.
Reform’s trust problems over Farage’s £5 million “G(r)ift” are evidence that the Hawk-Dove equilibrium is already emerging organically. Conservatives should not try to accelerate this through aggressive counter-attack, which would merely catalyse the Hawk-Hawk dynamic. Instead, they are better served by consolidating and deepening their Dove positioning, ensuring that when Reform’s Hawk strategy next generates costs, Conservatives are unambiguously the credible, stable, governing alternative. Let the left-leaning parties fight Farage for them.
The broadest structural challenge facing Conservative recovery is a Stag Hunt: a coordination game in which the best collective outcome requires simultaneous mutual commitment from multiple players who cannot directly verify one another’s intentions.
Rebuilding a Conservative majority requires voters, MPs, donors, local associations and media supporters to commit simultaneously to the recovery project. But each player is willing to commit only if they are sufficiently confident that others will also commit. In the absence of that confidence, each defaults to the individually safer option of staying with Reform, abstaining, or maintaining scepticism and inaction.
The Stag Hunt produces two equilibrium outcomes. The first is the high-level equilibrium, where everyone cooperates, the stag is caught, and the Conservative majority is rebuilt. The second is the low-level equilibrium, where insufficient confidence in others’ cooperation leads everyone to defect to safer individual options, and the Party remains trapped at 19% to 20% of the vote share.
The Party is currently locked in the low-level equilibrium. Breaking out requires generating sufficient common knowledge: not merely that each player believes in recovery, but that each player believes every other player believes in it, and also believes that every other player believes that every other player believes in it.
This recursive confidence is the precondition for coordination, and it cannot be generated by individual conviction alone.
The strategic implication is that the Conservative recovery programme must be explicitly designed as a coordination mechanism, creating visible, credible, and mutually reinforcing signals of recovery that make it individually rational for each additional player to join.
A properly resourced target-seats strategy, visible momentum in by-elections, credible polling trajectories, and the public commitment of key figures all serve this coordinating function. Paradoxically, a party that honestly acknowledges a two-election recovery horizon, rather than claiming implausible imminent success, generates more credible common knowledge because its forecast is consistent with observable reality.
Some comfort can also be taken from recent political volatility. Johnson’s 2019 victory was followed by a rapid reversal that produced Starmer’s historic win, which itself has collapsed within two years. We wait to see what a midnight blue T-shirt and eyebrows can deliver.
Conclusion: Restructuring the Game
Game-theory analysis converges on a single, powerful conclusion.
The Conservative Party is not failing simply because its policies are wrong, its leader is inadequate, or its communications are insufficiently polished.
It is failing because the strategic game structure itself is working against it.
Rational punishment strategies are being administered. Coordination failures are trapping potential returners. Cheap signals are being correctly discounted by voters. Internal Prisoner’s Dilemmas are producing collective indiscipline, albeit to a much lesser extent after Jenrick’s exit. And the Hawk-Dove dynamic with Reform, while shifting, has not yet resolved itself in the Party’s favour.
The path back does not lie in winning arguments within the existing game structure. It lies in restructuring the game itself through credible commitment devices that change voter beliefs; costly signals that communicate genuine transformation; coordination mechanisms that solve the collective-action problem; payoff restructuring that resolves the internal Prisoner’s Dilemma; and the patient acceptance of the punishment phase that the Folk Theorem demands.
As Kemi Badenoch said at her 2024 leadership launch: “In engineering, you can’t overcome resistance with rhetoric. You can’t get the machine to work with promises.”
Game theory provides the precise analytical framework that this engineering instinct requires, and its diagnosis is clear. The machine will work only when the structure of the game changes. That change begins not with a speech, but with the binding commitments above.