The Story

Tinker Tailor Soldier Spy is, at its core, a story about institutional failure. The British Secret Intelligence Service, known internally as the Circus, has been penetrated at the highest level by a Soviet mole, and the novel’s central tension is whether the institution is even capable of identifying one of its own.

Control, the aging Chief of the Circus, believes someone on the Fifth Floor, his own senior leadership, is feeding intelligence to Moscow Centre. He’s right. But he can’t prove it, and his actions, informed more from instinct than evidence, set in motion the very catastrophe he is trying to prevent.

The story is set in motion by Operation Testify: a covert mission to extract confirmation of the mole’s identity from a Czech military source. Control sends a trusted officer, Jim Prideaux, to Czechoslovakia to make the contact. The operation is betrayed before it begins, and Prideaux is shot, captured, and broken. The Circus’s Eastern European agent networks, years of patient work bringing in people with families who trusted London with their lives, are systematically rolled up by Soviet counterintelligence. Control is forced out and dies shortly after, while the mole remains in place and the Circus falls under the control of the very faction the mole has been cultivating.

It takes George Smiley, recalled from retirement, to reconstruct the logic of the betrayal after the fact. He pieces together who had access, who had motive, and who had opportunity through months of painstaking investigation that depends entirely on people choosing to trust him over their own institution.

Le Carré described the Circus in rich enough detail to reconstruct a realistic organizational network from his writing, making this novel a good test case. Whether the answer to the novel’s central riddle is visible in the illustrated structure is something only the analysis can unpack. For a first-time reader, deducing who the mole is without following Smiley step by step through his investigation is very difficult, because the novel buries it in character, misdirection, and institutional politics. Tessera Semita’s framework doesn’t read novels. It reads structure, so it isn’t distracted or charmed by complex character behaviors. It just measures position, and position doesn’t lie.

The Question

Would employing Tessera Semita’s proprietary methodology have helped Control find his insider threat?

To find out, we plotted the Circus’s network from the novel as faithfully as possible to a pre-Operation Testify state, using only the kinds of data an organization would actually have on hand before anything went wrong; in our case the book and the network described by le Carré in his novel “Tinker Tailor Soldier Spy”. From the book we extracted data that can be reasonably inferred to exist in the state we are targeting and we assembled it from the descriptions provided. Information that would, in reality, be obtained from personnel logs, internal and external location attendance observations, diplomatic reception attendance lists, professional reporting chains, and social affiliations. All of this would be available to the Circus, and Le Carré describes it in enough detail for us to populate a synthetic network faithful to the pre-Operation Testify state. Where the novel implies roles or functions without naming individuals, we made reasoned additions consistent with the context Le Carré provides. All of this can be observed without the need for an investigation triggering event, all the data could be populated through a precise and deliberate inquiry.

No accusations. No suspicion. No hindsight.

Well, the hindsight can’t be eliminated from the analyst, but it can be controlled for by excluding data that is beyond the scope of our collection effort, a collection effort that is trying to isolate a specific temporal moment: pre-Operation Testify. That effort was taken very seriously because neglecting it would make this demonstration moot due to bias.

The Setting

The British Secret Intelligence Service, the Circus, operates from Cambridge Circus in London. Its senior leadership, the Fifth Floor, manages intelligence collection across Eastern Europe and the Soviet Union through a network of field officers, case officers, analysts, and support staff. The Circus maintains relationships with allied services (CIA, BND), handles sensitive liaisons with Whitehall (the broader British government apparatus), and runs a surveillance section (the Lamplighters) for operational support.

In the period before Operation Testify, the ill-fated Czech operation that would cost Jim Prideaux dearly and Control his career, the organization doesn’t suspect any specific individual of wrongdoing. Control suspects someone, but his suspicion is intuitive, political, and ultimately unactionable because he can’t point to anything definitively.

The structural conditions for insider compromise, however, are observable:

  1. Some personnel hold dual organizational roles that bridge otherwise separated sections, creating access across information domains that no single role requires.

  2. Some personnel maintain social contacts that extend into the diplomatic community, including the possibility of sharing presence with Soviet diplomatic staff.

  3. The combination of broad internal access, external social reach, and proximity to a known hostile intelligence contact in the same individual could create compounding opportunity that routine personnel review might surface but never combine into a definitive operating picture.

The Individual Opportunity Index (IOI) framework asks two questions of this network:

  • Q1: Does the Circus’s organizational structure create measurable opportunity concentrations that, if left unmanaged, could enable an incident?

  • Q2: If the Circus’s organizational structure creates measurable opportunity concentrations, where are they?

1. Baseline Network and Analysis

Structural Validation

Before looking at what the network shows, the first question is whether it behaves like a real organization at all. A network built from a novel could be random, structurally incoherent, or too neatly arranged to be credible. The Conditional Uniform Graph test directly addresses this concern by generating 5,000 random networks conditioned on the same size and density as the Circus network and comparing the observed transitivity against the full distribution.

The result isn’t close. The Circus network’s transitivity falls so far outside the random distribution that the probability of observing it by chance is effectively zero. This network behaves like a real organization and the CUG confirms that.

The Full Network

The network was constructed from le Carré’s text across eight thematic keyword searches of the novel. Every node represents a character or entity mentioned in the book, and every edge, whether professional, social, or co-presence, traces to a described relationship, interaction, or attendance pattern in the source material.

Some nodes carry names from the text. Others carry naming conventions like SOV_EMB_1 or WH_3. Those were populated based on the descriptions of named characters and the narrative’s account of organizational functions, social patterns, and the reasonable presumption that few people exist inside an organization in isolation. An embassy isn’t one person. A surveillance section doesn’t operate with two people.

The only integrity flag in the dataset is Karla, because George Smiley met him personally in Delhi years before the events of the novel and identified him as a senior Soviet intelligence officer. That’s the only piece of prior qualitative knowledge in the system, and everything else is structure. However, having this flag will hardly impact the analysis because this character is never in London, but we did give him a single co-presence at the Soviet Embassy because it can be reasonably presumed that Soviet diplomatic efforts in an adversarial state would have some connection to Soviet intelligence.

Internal Structure

Looking inward at the SIS, this is what the organization looks like from the inside. Every internal location in this network is a collection environment. The canteen, the registry, the fifth floor, the meeting rooms. In the context of an intelligence service, anywhere people gather and information is present is a surface where collection can occur.

External Exposure

Now looking outward. These are the spaces outside internal controls where Circus personnel show up. Diplomatic receptions, clubs, galleries, restaurants. The organization has no authority over who else is in those rooms, and limited visibility into what gets exchanged there. This is where co-presence with external contacts, including adversarial ones, becomes observable.

Risk Assessment

With the internal structure mapped and the external exposure surface visible, the IOI framework scores every node in the network across three dimensions: collection opportunity, information access, and distribution exposure. The result is a risk matrix that plots individual opportunity against the organizational severity of that individual becoming an insider threat (OSI is an index of measures populated by the client’s understanding of their own organizational roles, in this case it’s populated by analyst judgment).

2. What the Structure Shows

Connie Sachs and the Limits of Qualitative Analysis

Connie Sachs is a Circus analyst. She spent years tracking Soviet diplomatic staff and noticed Polyakov’s behavior suggested he was a Soviet intelligence asset. She was right, but when the Circus had him followed, they saw no smoking gun and dismissed Connie’s assessment. When Connie pushed back on that, the Circus pushed her out for it. Her method technically worked, but it had a fatal dependency: someone had to believe her and it had to be verified beyond doubt – a high bar for qualitative analysis that is dependent on degrees of certainty.

This framework doesn’t need Connie Sachs, but her analysis would have a procyclical value add with our own. One key difference between Connie’s analysis and Tessera Semita’s analysis is where it starts. She started from the external threat and worked inward – we go the opposite direction. We don’t need to identify Polyakov as suspicious, because Polyakov is already in the data by virtue of being a Soviet diplomat present in London. His attendance at receptions is intelligence, but for Connie it was a key piece of intelligence and to Tessera Semita it’s a cell value in the bipartite matrix that doesn’t care about his nature/demeanor/prior training; we simply observe his existence and move on. The Circus did too, but without a means to analyze opportunity, Polyakov was just another profile in a stack of files that one analyst, Connie, fixated on while everyone else moved past.

For Tessera Semita, the framework starts from internal structure and asks who has the most opportunity, regardless of whether a known adversary is in the picture. It found Haydon without looking outward at all. While Connie’s analysis would strengthen the urgency of this information, it was not needed to identify the real risk burdening the Circus.

Structural Proximity to a Group of Concern

Through the normal behaviors the book describes, through Haydon’s social life, his attendance at diplomatic receptions, his presence in the circles where Whitehall meets the foreign diplomatic community, there’s a measurable path to Polyakov. Not a direct tie. No phone call, no meeting, no dead drop. A path through shared spaces where co-presence is routine and unremarkable.

The signal isn’t that Haydon is meeting with a Soviet officer. The signal is that Haydon’s normal social pattern creates the shortest structural path from inside the SIS to a known diplomatic presence, and that path runs through spaces where showing up is just part of the job. For whatever reason, he has the shortest distance to an adversary from inside the SIS. The framework doesn’t care about the reason but it does care about the distance.

Information Proximity

Because we’re looking at a limited network, the information proximity visualization will carry some skew. Had we been able to map every person in every space, the picture would shift, but the underlying reality it exposes is still valid. What we’re concerned with is how close to SIS secrets the Soviets are, and how they got that close (something that is incredibly easy to dismiss without a real measurement being taken).

The full view shows that Soviet nodes are closer to internal SIS nodes than some of the Circus’s own people. But those distant internal nodes are foreign station personnel, populated because foreign stations aren’t unoccupied. They’re far away because they’re supposed to be far away.

Remove those foreign station nodes and the picture sharpens. Haydon is closer to the Soviets than would be intuitively clear from knowing he attends diplomatic events. The shortest path is visible in this layout, and it’s shorter than his job title alone would explain.

The Elevated Node’s Neighborhood

We know who we’re looking at. Haydon’s social neighborhood is large, and that isn’t unusual for someone in his position. It includes Soviet nodes, but that isn’t unusual either. He’s an intelligence station head during the Cold War. Contact with the other side’s diplomatic staff is part of the landscape.

Algorithmic Communities

This isn’t showing how the organization classifies its people. It’s showing what the algorithm identifies as Haydon’s communities based on structural ties. The Louvain method groups nodes by who they’re actually connected to, not by where the org chart puts them.

WH_4 appears in the same community as the Soviet nodes. From Haydon’s structural perspective, this Whitehall contact clusters with the diplomatic circuit where Polyakov operates. That’s an interesting finding. This node might know something about Haydon that was never revealed in the book, because Smiley’s method for uncovering the mole after the fact couldn’t have surfaced it. Or it could be an artifact of the synthetic network. In a real engagement, it’d be advised to find out who WH_4 actually is and that might lead to a conversation that could gather more qualitative support for a conclusion.

Cohesive Subgroups

What are Haydon’s tightly knit communities? The maximal cliques, the groups where everyone knows everyone and information flows freely because trust is assumed by proximity.

No Soviets. Not one. Haydon’s most cohesive relationships are entirely internal and entirely high-access – explainable by his role in the organization but also a risk indicator due to the value of his access. A well-run mole doesn’t embed with his handler’s people; he embeds with his own – but this is not an accusation, just an observation.

Every qualitative checkpoint clears him. No direct ties to the adversary. Soviets in his neighborhood but that’s his job. No Soviets in his inner circles. Behaviorally clean. And yet he’s still the highest IOI score in the network, because the framework isn’t measuring intentions or pointing out suspects – it’s measuring opportunity.

Is His IOI Normal for His Position?

It doesn’t matter that his job title muddies the waters qualitatively. The question is whether his IOI is normal for someone in his division. The boxplot shows the distribution for each division, and Haydon sits at the tail end of his division’s range, close enough to look plausible but far enough to suggest he’s the one skewing it. Whether that’s typical for the position or specific to the person occupying it would require iterative testing with variety in who fills the role, something the framework supports but this single snapshot can’t answer.

Articulation Points

Cutpoint analysis identifies nodes whose removal would fragment the network into disconnected components. In a defensive context, these are the personnel whose departure would sever communication channels between sections of the organization. In an adversarial context, the same analysis tells an opposing service exactly where to apply pressure for maximum disruption.

Haydon appears as an articulation point. His removal isolates three nodes: AGENT_1, AGENT_2, and AGENT_3. All three are synthetic nodes from the constructed network, and in a real organization those agents would have other handlers and other reporting chains. This particular finding is a limitation of the synthetic data, not a structural discovery.

But the method also picks up elements of the synthetic network that keep recurring: SOV_EMB_1, SOV_EMB_2, and WH_4, all of whom appeared in Community 7 from the Louvain analysis alongside Polyakov. The same nodes the framework flagged in community detection keep showing up across different methods. That’s either a consistent artifact or a consistent signal. In a real engagement with real people instead of synthetic identifiers, we’d understand which one it is.

Bridge Edges

These are the specific ties whose removal would most disrupt network connectivity. Rather than looking at all bridge edges in the network, this focuses on the edges connected to the highest-scoring IOI node. The result identifies which of Haydon’s specific connections carry the most structural weight and where restricting access would have the most impact.

The Weight of the Structure

Any single finding in this analysis can be explained away. His social proximity to Polyakov? He’s a senior officer who attends diplomatic functions. His information proximity? He runs London Station. His elevated IOI? He holds a dual role. His cohesive subgroups being entirely internal? That’s what a trusted insider’s network looks like. His anomalous IOI relative to his division? Maybe his job is just bigger than his peers’.

Each of these explanations is individually reasonable, and any one of them would satisfy a routine personnel review. But the framework doesn’t evaluate these findings individually. It measures them structurally, and structurally they all converge on the same node. The person with the shortest path to an adversary is also the person with the highest information proximity, the highest IOI, the most anomalous score relative to his peers, and the strongest bridge position in the network. That convergence isn’t something a job title explains. Every angle that could be offered as an alternative explanation has been addressed in this analysis, and none of them change the structural conclusion: Haydon carries the highest concentration of opportunity in the Circus, and it isn’t close.

While this is a far cry from a conviction, it does show how structural opportunity can inform and focus resources for an insider threat program. Precisely why we classify this style of analysis as prevention: we cannot predict intentions, guilt, or behavior by looking at opportunity alone, but in conjunction with current best practices this can focus the attention of a counterintelligence element or insider threat program and increase their efficacy. Furthermore, this analysis doesn’t end with identification; there are recommendations for intervention too that generate measurable and predictable responses from the elevated IOI individual.

Organizational Exposure

Zooming out from the individual to the organization. The SIS was operating at a .886 OEI before Operation Testify, against a theoretical maximum of 2.20. That’s moderate exposure but saying whether or not it is normal would require multiple measurements taken at different points in time (an intensive and lengthy process). But a significant portion of that .886 traces back to a single node. This isn’t an organization with evenly distributed risk. It’s an organization where one structural position is dragging the macro exposure upward, and managing that position would meaningfully change the organizational picture.

3. Percy’s Fool’s Cap: What Happens Without Analysis

We know the outcome of the novel. The network doesn’t. That asymmetry is the point of this exercise, but it also lets us assess something the network alone can’t: the executive errors that compounded the original compromise.

Control suspected a mole on the Fifth Floor, and he was right. But he acted on instinct and a limited analysis, and he zeroed in on Percy Alleline, the man who was loud, political, abrasive, and transparently ambitious. Here’s the uncomfortable truth: Control’s read on Alleline wasn’t unreasonable. It was, in fact, entirely consistent with the behavioral indicators that conventional insider threat programs are designed to flag. Alleline was politically motivated, cultivating external relationships beyond his operational mandate, and openly positioning himself for advancement at the expense of Control’s comfort. A behavioral assessment would have flagged him too, as it should.

The behavioral approach wasn’t wrong about Alleline being worth watching. The problem is that it couldn’t tell Control what it can’t measure: that Alleline didn’t carry the structural opportunity to be the mole he was looking for. Behavioral indicators identify people who look concerning, while structural analysis identifies positions that carry opportunity. Those are different questions, and in this case they pointed at different people. Without the structural dimension, Control’s error, targeting the behaviorally conspicuous officer instead of the structurally dangerous one, is only recognizable as an error in hindsight. With a structural opportunity system, it could be recognized in advance and thus prevented.

What Control did next was, in the framework’s terms, an uninformed Tier 2 intervention: he restructured Alleline’s organizational access without first conducting a structural assessment of the actual threat surface.

The mechanism was bureaucratic. Control ensured that practically the entire Circus would decline to open their books to Alleline. Le Carré describes the result: Alleline was starved of significance. He had the title and the seniority but none of the organizational plumbing, and the network confirms it. Alleline’s IOI reads very low, suppressed because the “Fool’s Cap” worked as a structural constraint; he has Fifth Floor membership and Whitehall contacts, but almost no cross-sectional operational reach.

Control was authorized to do this because he was Chief. Nobody challenged the Fool’s Cap, and nobody needed to approve it, so authorization was never the problem. The problem was that he had the authority to act but not the full analysis to act correctly. Power without insight. That combination is arguably more dangerous than having neither, because it meant he could make consequential structural decisions that reshaped the entire organizational network based on personal judgment alone.

The consequences were catastrophic on three levels:

  1. He constrained the wrong person. The man Control was punishing wasn’t the mole. The actual threat node, Haydon, was left completely unmonitored with full structural access to every section of the Circus.

  2. Shifting Alleline to a loss mindset. By starving Alleline of operational relevance, Control pushed him toward any source of intelligence that could restore his sense of significance. Moscow Centre, through Polyakov and Haydon, handed him exactly that. ‘Witchcraft’, an operation where an “informant” from inside the Soviet camp was providing seemingly high-value intelligence to the SIS – it was a cover for the mole to give high-value intelligence to the Soviets, but I digress; Witchcraft became Alleline’s lifeline. He defended it with the ferocity of a man whose entire career depended on it being real because his career was part of his identity and being put in the “Fool’s Cap” put him in a loss mindset. He was losing the future he thought himself deserving of, his own ego ruined him. Ego is the “E” in MICE, an acronym for remembering key elements for recruiting HUMINT assets. Control was confident that the mole was Alleline and for good reason too, all the behavioral indicators were there, but it was inaccurate and his handling of the situation pushed Alleline towards defection.

  3. He made the signal clearer and missed it anyway. The Fool’s Cap actually made Haydon’s anomalous position more visible by contrast with Alleline. Alleline was constrained and Haydon wasn’t. If Control had run a structural analysis on his own organization, the gap between those two positions could have informed him of his options long before he leapt on the opportunity to get definitive proof of his mole’s identity with Operation Testify.

The framework doesn’t restrict executive authority, and Control still makes the call and still holds the pen. But before he restructures someone’s access, the analysis shows him where the actual structural risk concentrates, which might not be where his instincts are pointing or where a behavioral analysis asserts the biggest risk is.

Same authority, same Chief, but more options, that’s what structural analysis provides.

With Tessera Semita’s framework, Haydon would have been identified rapidly, perhaps alongside Alleline. However, even if Alleline’s behavioral profile warranted attention, the tiered response protocol would never have permitted a Tier 2 structural intervention without first passing through Tier 1 deterrence. Restructuring someone’s organizational access before documenting the structural basis for concern, establishing monitoring criteria, and giving the deterrence effort time to produce measurable behavioral change is sloppy and runs against best practices for an intervention. Control skipped everything. He went from suspicion to structural action in a single step, on the wrong target, with limited analytical basis, and no process to course-correct before the damage compounded.

4. Tier 1: What the Framework Would Have Done

Had Tessera Semita’s structural analysis system been deployed, Haydon’s score would have crossed the monitoring threshold before Operation Testify was approved. Under the tiered response protocol, Tier 1 inquiry would have activated: notification that the organization is aware of an elevated risk concentrated around Haydon would’ve been the method to initiate deterrence. He would’ve become aware of passive monitoring of interactions and assigned self-reporting requirements.

None of his ties would be severed, no access restricted, and no accusations of misconduct. The only change is that he would now know that the organization is watching and more aware than he originally thought.

5. Tier 2: Structural Controls

If deterrence fails, if the behavioral change projected by Tier 1 deterrence projections don’t materialize, the organization escalates to structural action. We identify which specific ties to restructure that would reduce the elevated person’s (Haydon’s in this case) IOI to a tolerable threshold while minimizing disruption to organizational functions.

6. The Answer

Yes.

Had structural analysis been deployed before Testify, the framework would have identified Haydon as the individual whose structural position creates the highest compounding opportunity for compromise. The framework doesn’t require evidence of intent; it only requires social structure information that any competent OSINT collector can obtain.

What would have changed:

Testify proceeds safely. With Haydon’s access structurally constrained under Tier 2, Operation Testify is approved and Jim Prideaux goes to Czechoslovakia. The operation isn’t compromised because Haydon no longer has the structural position to leak it. Prideaux completes his mission and returns with Haydon’s name affirming that Haydon was compromised.

Connie Sachs is validated. Her observations about Polyakov’s behavior patterns are strengthened by the structural data, and the structural analysis benefits from urgency derived from her qualitative analysis.

Control keeps his seat. There’s no catastrophe to investigate. The Witchcraft deception never takes root because the structural opportunity that enabled it was already visible in the data and managed through proportional intervention.

Naturally, wanting a Smiley is fully understandable; he solved the carefully constructed concealment that Haydon, arguably the best spy in fiction, had crafted. However, needing George Smiley in your Insider Threat Department or Counterintelligence element is a systems failure because George Smiley is a fictional character. The approach of the book is still a valid approach, George Smiley asks: “Who is the mole? Who has incentives?”; we approach the problem from another angle, a structural angle, that asks: “Who has the most opportunity?”

Le Carré provides all the insight into what could happen. Tessera Semita tells you how to prevent it.

7. Summary

This demonstration applies Tessera Semita’s structural analysis framework to a fictional organization where the answer is already known. The Circus had a mole, the novel tells us who it was, and the framework identified him from a network diligently extracted from le Carré’s text and populated with synthetic data to bring it closer to what a real organizational network would look like.

What this shows is what other approaches miss. In the novel, behavioral profiling flagged the wrong person. Institutional intuition pointed in the wrong direction. Qualitative analysis identified an external threat, but no one in the book could define the significance of that analysis. Tessera Semita’s structural approach found the right answer to the novel’s riddle because it doesn’t depend on suspicion, intuition, or trust. It measures opportunity, and opportunity is visible in the data whether anyone is looking for it or not.

In fiction, the cost of missing something is a good story, a tragic story but a good one. In practice, the cost of missing something is not romantic. In reality, missing the opportunity indicator can result in some measure of institutional existence being offered up as a sacrifice for someone’s selfish gains; operations compromised, burned networks, destroyed careers, and in the worst case all of these compounded into a forced restructuring. The framework exists because gambling on outcomes shouldn’t depend on whether an organization happens to employ a George Smiley.