स्तम्भ Stambh · Raṇanīti

Choose the Narrow Pass

What a film about Thermopylae, a law drama built on a lie and a comedy about a compression startup teach a founder about strategy. Pick the ground where numbers stop counting. Guard the path you were told nobody uses. Know which launch to stop.

Nikhil Sharma · रणनीति Raṇanīti · 11 October 2026 · 7 min read

Most founders learn strategy from three places: a book they did not finish, a mentor who was right once and the stories they watched on a Sunday. I want to defend the third. Fiction is where most of us first saw leverage, betrayal and a hard choice played out at full length. The trick is to watch it the way a general reads a map and not the way a fan reads a poster. Here are three I keep returning to and the pattern each one taught me.

Thermopylae: choose the ground where numbers stop counting

The film 300 came out in 2006. Zack Snyder made it from Frank Miller’s comic and it is about as accurate as a comic is meant to be. The battle underneath it is better than the film. In 480 BC a Greek force of about seven thousand, three hundred of them Spartans, held the pass at Thermopylae against a Persian army that modern historians put somewhere between a hundred and twenty thousand and three hundred thousand. They held it for seven days, three of them in open fighting.

They did not hold it because they were braver. They held it because of the ground. At its middle gate the pass between the mountain and the sea was perhaps a hundred metres wide, with an old wall the Phocians had built across it. In a space that narrow the Persian numbers could not be brought to bear. Only the front rank fights. Leonidas chose a battlefield on which the other side’s main advantage was worth nothing.

That is the first lesson of competing with an incumbent and almost every founder gets it backwards. They pick the widest field because it looks like the biggest prize, which is exactly the field on which the large company’s money, distribution and brand all count in full. The narrow pass is a segment so specific, a city so particular or a workflow so awkward that the giant cannot deploy its army there. Fight on that ground and only your front rank meets theirs. Win the pass first. The plain comes later.

Pick the ground on which your rival’s advantage is worth nothing.

The second lesson is in how the pass fell. A local man called Ephialtes showed the Persians a mountain track that ran round behind the Greek line. Leonidas knew about it. He had put a thousand Phocians on the heights to guard it. They were surprised at dawn, fell back to a hill to defend themselves and the Persians simply walked past them. The flank was not unknown. It was delegated and then forgotten.

Every company I have carried had its own Anopaea path. The GST filing someone else handles. The one customer who is forty per cent of revenue. The server password only a contractor knows. The founder agreement nobody signed. None of these is a secret. Each was handed to a small guard and nobody went up the mountain to check. Strategy is not only the line you hold. It is the path behind it.

And the third lesson is about the legend itself. Ask anyone how many held Thermopylae and they say three hundred. The record says about seven thousand for most of the week and a rearguard of perhaps two thousand at the end, seven hundred Thespians among them beside the Spartans. Founders do the same to their own story. In the retelling it was the founder alone. It never is. The company that believes its own legend stops counting the people who actually held the line.

Suits: leverage is real, and so is the lie it was built on

Suits ran on USA Network from 2011 to 2019, nine seasons and a hundred and thirty-four episodes. In 2023 it found a second life on Netflix and Peacock and became the most-streamed show of that year in America. A generation of young professionals learnt negotiation from Harvey Specter. Some of it is good. Know what the other side cannot afford to lose. Never walk into a room without knowing your alternative. Make the first offer when you know more than they do.

But the show’s real engine is not Harvey’s leverage. It is a lie. Mike Ross is brilliant and has a photographic memory and he has no law degree. The firm’s finest work for years is built on that fact and the most capable people in the story spend season after season protecting it. Eventually it is found out and Mike goes to prison for two years.

Watch it as a founder and the lesson is not about negotiation at all. It is about the cost of a lie at the centre of a company. A revenue figure dressed up for a round. A licence you meant to get. A metric that counts what flatters you. Each one works at first, the way Mike works at first. Then it becomes the thing the whole organisation exists to protect. The lie stops being a shortcut and becomes the company’s largest recurring expense. Leverage built on it is borrowed against a day of reckoning. I would rather have less leverage and nothing to hide.

A lie at the centre of a company becomes its largest recurring expense.

Silicon Valley: know which launch to stop

HBO’s Silicon Valley ran from 2014 to 2019, six seasons from Mike Judge, John Altschuler and Dave Krinsky. Most founders remember it for the jokes that hurt because they were true: the pitch that becomes a pivot, the board that loves you until it does not, the incumbent that copies you and still loses. It even gave the industry a fake metric. The Weissman score, the compression benchmark Pied Piper chases, was invented for the show by a Stanford engineering professor, Tsachy Weissman, and his doctoral student. People in the business quoted it as if it were real. That is its own lesson about metrics designed to win a pitch.

The finale is the part I want every founder to watch. After six seasons of near death, Pied Piper finally has the deal: a hundred million dollars to build its new internet. Then the team discovers that the AI inside their own product can break encryption. Launch it and every private message in the world becomes readable. So they sabotage their own launch, in public, and let the company look like a failure. The ten-years-later epilogue is kinder to them than any IPO would have been.

This is the most serious idea in the whole series and the most relevant one in 2026. Raṇanīti, the shelf this sits on, says that security is an economic question before it is a military one. The finale makes that exact argument. A capability that destroys the trust the whole market runs on is not an asset however large the cheque. The hardest strategic decision is not which hill to take. It is which product not to ship, and the discipline to say so when the money is already on the table.

What Kautilya would add

Kautilya listed six measures open to a ruler: peace, war, waiting, marching, seeking shelter with a stronger power and the double game of peace with one and war with another. The six are a better strategy course than most MBA electives. What strikes me about the list is the third item. Āsana, staying put, is a measure in its own right. Not a failure to act but a decision.

Leonidas chose to stand still in the right place. Mike Ross should have stood still at the door of the firm and said I do not have a degree. Pied Piper chose to stop at the moment everyone was telling it to run. All three are strategy by refusal. Founders are taught to move. The best ones also know where to stand.

Gurugram, 11 October 2026. The first piece on the Raṇanīti shelf. More of these will come from where I learnt them: films, series, matches and history, read for the pattern and not the poster.

Sources, checked 11 October 2026. Thermopylae: Battle of Thermopylae for the Greek force of about 7,000, the modern estimates of the Persian army, the seven days, the Phocian wall, the width of the pass from core samples, Ephialtes, the thousand Phocians on the Anopaea path and the rearguard; Herodotus on Thermopylae at Livius for the ancient account. Suits: Suits (American TV series)) for the run, seasons, episodes, premise, Mike Ross’s sentence and the 2023 Nielsen record. Silicon Valley: Exit Event for the finale’s plot and the AT&T deal; IEEE Spectrum, A Made-for-TV Compression Algorithm for the Weissman score. Kautilya’s six measures, ṣāḍguṇya, are in Book 7 of the Arthaśāstra.