The moon doesn’t necessarily exist if you don’t look at it. So says quantum mechanics, which states that what exists depends on what you measure. Proving the truth in this way usually involves comparing ambiguous probabilities, but physicists in China have made this point clearer. They conducted a matching game in which two players take advantage of the quantum effects to win each time – something they cannot do if the measurements reveal the truth as it really is.

“As far as I know this is the simplest [scenario] Adan Cabello, a theoretical physicist at the University of Seville who demonstrated the game in 2001. This quantum pseudopathology relies on correlations between particles that are only found in the quantum world, says Ann Broadbent, a quantum information scientist at the University of Ottawa. “We are noticing something that has no classical counterpart.”

A quantum particle can exist in two mutually exclusive states at once. For example, a photon can be polarized so that its electric field wiggles vertically, horizontally, or both directions at the same time—at least until it is measured. Then the two-way state randomly collapses into either vertical or horizontal. Crucially, no matter how the bidirectional state collapses, the observer cannot assume that the measurement reveals only how the photon was actually polarized. Polarization is shown only with measurement.

This last part alarmed Albert Einstein, who believed that something like the polarization of a photon should have a value independent of whether or not it was being measured. He suggested that the particles might carry “hidden variables” that determine how the bidirectional state collapses. However, in 1964, British theorist John Bell found a way to prove experimentally that such hidden variables could not exist by exploiting a phenomenon known as entanglement.

Two photons can be entangled such that each is in an uncertain state in both directions, but their polarizations are interconnected so that if one is horizontal, the other must be vertical and vice versa. Investigating entanglement is difficult. To do this, both Alice and Bob must have a measuring device. These devices can be oriented independently, so that Alice can test whether her photon is horizontally or vertically polarized, while Bob can’t detect his angled detector. The relative orientation of the detectors affects how closely their measurements are related.

Bell imagines that Alice and Bob randomly guide their detector through several measurements and then compare the results. If the hidden variables determine the polarization of the photon, then the correlations between the Alice and Bob measurements can only be very strong. But he argued that quantum theory allows them to be stronger. Several trials saw those strongest associations and excluded hidden variables, albeit only statistically across many trials.

Now, Xi-Lin Wang and Hui-Tian Wang, physicists at Nanjing University, and colleagues have made this point more explicitly with the Mermin-Peres game. In each round of the game, Alice and Bob share not one, but a pair of entangled photons to make whatever measurements they want. Each player also has a 3D grid and fills each square in it with 1 or 1 – depending on the result of those measurements. In each round, the judge randomly chooses one of Alice’s rows and one of Bob’s columns, which overlap in one square. If Alice and Bob have the same number in that square, they win the round.

It looks easy: Alice and Bob put 1 in each square to secure a win. Not so fast. Additional “equivalence” rules require that all entries across Alice’s row must multiply by 1 and those below Bob’s column must multiply by -1.

If the hidden variables predetermine the results of the measurements, then Alice and Bob cannot win every round. Each possible combination of values for the hidden variables effectively defines a network that is already filled with -1 and 1. The results of the actual measurements only tell Alice which one to choose. The same goes for Bob. But, as is easily demonstrated in pencil and paper, no single network can satisfy Alice and Bob’s equivalence rules. Therefore, their grids must differ by at least one square and, on average, they can win eight out of nine rounds at most.

Quantum mechanics allows them to win every time. To do this, they must use a set of measurements devised in 1990 by David Mermin, a theorist at Cornell University, and Asher Peres, a theorist who worked at the Israel Institute of Technology. Alice makes the measurements related to the squares in the row selected by the referee, and Bob’s measures those of the squares in the selected column. Synapsis ensures agreement on the number in the main square and that their measurements also conform to the valence rules. The entire chart works because the values are only shown while the measurements are being taken. The rest of the grid is irrelevant, as there are no values for the measurements that Alice and Bob never made.

Xi-Lin Wang says that simultaneously generating two pairs of entangled photons is impractical. So instead, the experimenters used a single pair of photons to entangle in two ways — through polarization and what’s called orbital angular momentum, which determines whether a photon is in a left- or right-hand spiral waveform. The experiment isn’t perfect, but Alice and Bob won 93.84% of 1,075,930 runs, overtaking the 88.89% cap with hidden variables, the team reported in a press study at physical review messages.

Others have demonstrated the same physics, says Capello, but that Xi-Lin Wang and colleagues “use exactly the language of the game, which is cool.” He says the show could have practical applications.

Broadbent takes into account real-world use: checking the workings of a quantum computer. This task is necessary but difficult because a quantum computer is supposed to do things that a normal computer cannot. However, Broadbent says, if the game is integrated into software, monitoring it can confirm that the quantum computer is manipulating the entangled states as it should.

Xi-Lin Wang says the experiment was primarily aimed at demonstrating the capabilities of the team’s favorite technology — photons are entangled in both polarization and angular momentum. “We want to improve the quality of these highly entangled photons.”