The Tenth Man: Israel's Answer to Groupthink, and What Happens When It Is Ignored
Security & Intelligence · Memo · 8 min read
Executive Summary
After the surprise attack of October 1973, Israel did something no other democracy has matched. It built dissent into the structure of its intelligence service: a small unit whose job is to argue that the consensus is wrong, with a direct line to the head of military intelligence. Fifty years later, on 7 October 2023, the warnings existed again. Some came from that unit. They were not acted on. Israel invented the best institutional answer to groupthink, and its own worst failure shows the condition that answer depends on. Dissent only works if someone is obliged to listen to it.
Context
On 6 October 1973, Egypt and Syria attacked Israel on Yom Kippur. Israeli military intelligence, known as Aman, had a great deal of information pointing to war. Its leadership, under Major General Eli Zeira, held to what became known simply as "the concept": a settled belief that Egypt would not launch a war until it had the long-range aircraft and missiles to strike Israel's airfields, and that Syria would not go to war without Egypt. Evidence that did not fit was explained away as exercises or bluff. Even a direct warning from Ashraf Marwan, Mossad's most valuable Egyptian source, on the eve of the attack came too late to change the mobilisation.
The Agranat Commission, set up to investigate the failure, held the intelligence chiefs responsible and Zeira was removed. Among the reforms that followed was a dedicated devil's advocate function inside the Military Intelligence Directorate, known by the Aramaic phrase ipcha mistabra, roughly "the opposite is plausible". The unit is headed by a colonel who reports directly to the head of military intelligence, not to the research department whose conclusions it challenges. Its mandate is to examine unlikely scenarios, attack shared assumptions and write the assessment nobody else wants to write. Its past heads include Yaakov Amidror and Amos Gilad, both of whom rose to the top of Israel's security establishment.
The idea reached a mass audience through the film World War Z, whose "tenth man" rule held that if nine people agree, the tenth must assume they are wrong and investigate. The film's version is fiction. The principle behind it is real, and it is Israeli.
Assessment
Why the design is right. Groupthink is not a failure of intelligence. It is a failure of incentives. Analysts who challenge the consensus risk their careers, and senior officers rarely seek out evidence that their judgement is wrong. Israel's solution was structural. It made dissent someone's job, protected it from the chain of command it criticises, and gave it access to the top.
What happened before 7 October. Israel's intelligence services reportedly obtained Hamas's attack plan, a document of around forty pages that officials called "Jericho Wall", more than a year before the assault. It was assessed as an aspiration rather than a capability. In July 2023 a veteran analyst warned, after observing Hamas training, that the gap had narrowed. In 2024 it emerged that the devil's advocate unit itself had written directly to senior military and political decision-makers on 21 and 26 September 2023, warning of a possible Hamas attack. The Israel Defense Forces' own investigations, presented in early 2025, acknowledged a comprehensive failure.
The lesson is uncomfortable. The mechanism worked. It produced the warning. What failed was the culture around it. A new concept had taken hold, that Hamas was deterred and focused on governing Gaza, and dissent that challenged it was received, filed and outweighed. Fifty years apart, the pattern was the same: the information existed, and a settled belief decided what it meant. Institutionalised dissent with no obligation attached becomes a ritual.
Why this still makes Israel the model. Israel's failure is visible because Israel built a system that recorded the warning. Most organisations would not even know what they had ignored.
The British Parallel
Britain has its own record of settled beliefs. The Chilcot Inquiry found in 2016 that judgements about Iraq's weapons were presented with a certainty that the intelligence did not justify. The Ministry of Defence publishes a red teaming handbook, and some departments use challenge functions. But no British institution gives dissent the protected status and direct access that Israel does, and in political parties, campaigns and boards the idea barely exists. Candidate selection is a good example: once a favourite emerges, almost nobody is paid to argue that the favourite is a risk.
Limits and Criticisms
Critics argue that a permanent devil's advocate becomes predictable and is easily discounted, a contrarian by job title rather than conviction. Others point out that the scale of the 7 October failure, across military intelligence, the Shin Bet and political leadership, shows that no single unit can correct a whole system's assumptions. Both points are fair. Neither is an argument against the principle. They are arguments for attaching consequences to it.
Implications
Every organisation runs on concepts: the candidate is safe, the opponent will not dare, the regulator will not move, the board has seen everything it needs to. The question is not whether those assumptions exist but whether anyone is paid to attack them, and whether leaders must answer when they do.
What Decision-Makers Should Do
Write down your concept. State the assumptions your strategy depends on in plain words, so they can be tested rather than absorbed.
Appoint a tenth man. Give one person or small team the explicit job of arguing the opposite, and protect them from the reporting line they criticise.
Attach an obligation. Require a written response from the decision-maker to every formal dissent. Silence should not be an option.
Rotate the role. Prevent the dissenter becoming a familiar voice that is easily tuned out.
Review after the event. When a dissent proves right, ask why it was outweighed, and change the process.
Forward Look
Israel's post-7 October inquiries will shape how intelligence services everywhere think about warning and decision. Expect more interest in structured challenge across government and the private sector, and more scrutiny of leaders who were warned and chose not to act.
Bottom Line
Israel proved that you can build dissent into an institution. It also proved that dissent alone is not enough. The tenth man has to be heard, and someone has to be accountable for ignoring him.
Comments