Are Three-Body Problem and Foundation the same kind of story?
Not quite, but they rhyme. Both ask whether science can predict and steer the fate of an entire civilization. Asimov calls that science psychohistory; Liu Cixin calls it cosmic sociology. Both assume that once a population is large enough, collective behavior becomes statistically predictable, like gas molecules in a chamber. The twist is that they reach nearly opposite conclusions from that shared starting point.
Foundation is the optimistic version. Hari Seldon predicts the Galactic Empire will fall and then does something crucial: he engineers the Seldon Plan to compress thirty thousand years of dark age into a single millennium. Prediction plus intervention. The two axioms of cosmic sociology take the same premise somewhere much colder, deriving the dark forest theory instead of a rescue plan.
Why do two "predict the future" sciences end up so different?
Because the authors disagree about how much a single person can bend the system. Asimov, writing in the optimistic postwar years, believed rational elites could guide history. The Seldon Plan is built on the idea that a small group holding the truth can carry a whole civilization. Even when the Mule appears as an unpredicted mutant variable, the story eventually snaps back to a corrected path.
Liu Cixin writes from a much harsher worldview. In his universe individuals can barely resist structural survival pressure. The Wallfacer Project looks like a mirror of the Seldon Plan, a handful of people executing strategies only they understand, but three of the four Wallfacers fail, and the one who succeeds wins not through salvation but through the threat of mutual destruction.
Which one is closer to how the real universe works?
This is the more interesting question, beyond just optimism versus pessimism. Psychohistory's weak point is that it assumes civilizations coexist for the long term, communicating, trading, and migrating across millions of inhabited worlds. That is exactly the premise the dark forest denies. If the dark forest answer to the Fermi paradox holds, the default state of the cosmos is not a bustling Galactic Empire but a silent field of hidden hunters. On that score, Three-Body confronts the real "why is the sky so quiet" puzzle more directly.
But Three-Body has its own blind spot. The chain of suspicion assumes civilizations can never build trust, which pushes the cost of communication and the odds of goodwill down to zero. In reality, even within human society, that is not always true. Both worldviews are thought experiments that push one real variable to its extreme, not predictions.
If I only read one, which should I start with?
If you want grand scale and dense, escalating set pieces, Three-Body hits harder and faster. Foundation is more of a slow political epic told mostly through conversation. Both are worth your time, but the Three-Body reading order is friendlier: a clean trilogy versus Asimov's sprawling, decades-spanning series that later merges with his Robot novels and tangles the timeline.
Short version: pick Three-Body for a punch-in-the-face cosmic shock, pick Foundation for a slow-simmered tale of rise and fall. They are two answers to the same question, and that question, whether we can predict our own fate, may never have a settled answer.