The Cunning Hare Problem: Why Sima Yi Needed Zhuge Liang Alive

Xi Shan’s The Empty Fort Problem treats the Empty Fort legend as a two-body problem: did Sima Yi correctly read Zhuge Liang’s mind at that gate, or just get lucky about where to stop guessing? Either answer is too easy, and both rest on an assumption neither one defends: that whatever Sima Yi was doing, he was doing it about Zhuge Liang. A reading that has been gaining ground for years, mostly online, mostly among people who would never call themselves Three Kingdoms scholars, drops that assumption entirely.

The theory goes by a name older than the legend itself: 养寇自重, nurturing a threat to inflate your own worth. In this account, Sima Yi doesn’t need to guess whether the fort is really empty, because it doesn’t matter either way. Zhuge Liang alive and dangerous is the reason Wei’s court cannot afford to sideline the one general who can check him. Zhuge Liang dead is a general with no remaining excuse for the army, the authority, and the decade of political patience that follows. Han Xin is said to have put the underlying warning more literally while being arrested by the emperor he’d just finished making, in the Records of the Grand Historian’s biography of the Marquis of Huaiyin: 狡兔死，良狗烹，高鸟尽，良弓藏, the cunning hare dead, the good dog gets cooked; the high-flying birds gone, the good bow gets put away. He was echoing an older warning: nearly three centuries earlier, the strategist Fan Li had used almost the same image to tell a fellow minister of Yue to leave court while he was still useful, because usefulness was the only thing keeping him alive.

Sima Yi’s own career gives the reading teeth, which is more than the Empty Fort story can say for itself. In the years before 249, he perfected the art of seeming harmless, most famously feigning such convincing senility during a visit from a rival’s spy, drooling into his own robe and mishearing every question, that the spy reported back a dying man with nothing left to fear. He was lying. In 249, while his rival Cao Shuang was away with the young emperor on a tomb-sweeping trip, Sima Yi seized the capital in a single afternoon, the Gaoping Tombs incident, and put Wei under his family’s control for good. His grandson Sima Yan forced the last Wei emperor to abdicate in 265, founding the Jin dynasty, and Jin finished in 280 what Sima Yi never lived to see, swallowing the last of the three kingdoms. A 2017 television drama, The Advisers Alliance, gave Wu Xiubo’s Sima Yi two full seasons to build this arc. Its second season, Growling Tiger, Roaring Dragon, is built entirely around his rivalry with Zhuge Liang: not a general who got lucky at a fort gate, but a man patient enough to spend a decade playing dead for a payoff nobody watching him could have safely bet he was actually waiting for.

None of that proves Sima Yi was already running this calculation in 228, at a fort gate he was probably never anywhere near, twenty-one years before Gaoping Tombs and thirty-seven before Jin. If anything it shows the opposite of what it’s usually taken to show. We know how the story ends, and a man who ends a story that well is easy to read a plan into, backward, starting from the ending. Psychologists have a name for the mechanism doing the real work here, and it has nothing to do with Sima Yi’s intelligence. Baruch Fischhoff documented hindsight bias in 1975, the well-replicated finding that people systematically overestimate how predictable an outcome was once they already know it occurred. A grandson founding a dynasty doesn’t make one specific calculation at one specific gate more likely. It makes any story that leads there feel more likely, which is not the same thing, however identical the two feel from inside the retelling.

So is 养寇自重 just hindsight bias with better production values, a coronation retold backward and mistaken for a character read? Not entirely, but answering properly means pulling the theory apart from the moment it usually gets applied to. Nothing supports the claim that Sima Yi, staring at an open gate in 228, was already running a forty-year plan with his grandson’s throne as the payoff, and that specific claim is where hindsight does all the work. But the general incentive underneath the theory doesn’t require Sima Yi to have been a genius or a prophet at that gate. It only requires him to have already known what Cao Cao is recorded as having suspected of him decades earlier. In the Book of Jin’s own phrase, he had a wolf’s glance, the mark of a man too ambitious to stay content as anyone’s permanent subordinate, and Cao Cao had warned his own son to keep Sima Yi away from real power. A man already marked as dangerous has one way to keep a suspicious court from acting on the suspicion: staying visibly, urgently necessary.

It’s worth admitting that this anecdote comes from the same dynastic history Sima Yi’s own descendants later commissioned, which makes it the sort of story a new dynasty’s historians could have shaped after the fact. But it doesn’t need to be true to make the point. It only needs to describe the ordinary pressure a capable subordinate under a suspicious court lives under, whether or not this particular version of it happened to this particular man.

Xi Shan’s essay argues that nobody, including Sima Yi, could ever verify in real time whether a guess about another mind had actually landed or just gotten lucky. The 养寇自重 reading doesn’t solve that problem. It relocates it. Instead of a live guess about what Zhuge Liang was thinking, it becomes a live guess about what kind of man Sima Yi already was, decades before the evidence that would settle the question had even happened. The dog doesn’t get told, while it’s still useful, whether it’s being kept around out of respect or out of appetite. Sima Yi didn’t know either, standing at that gate in the story. He just turned out to be the rare dog who lived long enough, and stayed careful enough, that nobody ever had to find out which one it was.