Two barometers hang on the wall. One reads clear, the other says partly cloudy. Which weather do you trust? That is roughly the question facing a group of climate scientists when two independent methods predicted different futures for the Pacific Ocean in 2026.

The first method builds a climate network: 14 points in the central and eastern equatorial Pacific connect to 193 surrounding points if their temperatures fluctuate in sync. When the network becomes too "chummy," it signals that El Niño, a warm current that reshuffles weather from Peru to Indonesia, may arrive the following year. According to the network data, the threshold was not crossed in 2025, so the probability of El Niño in 2026 is only 8.6%.

The second method looks not at connections but at the "complexity" of temperature at a single ocean point. If the temperature series becomes more unpredictable, it resembles a tremor before an event. By this method, complexity has already exceeded the threshold, and El Niño 2026 is expected to be weak, with a probability of 71.4%.

The two barometers disagree. The authors did not choose who was right. They recalled Laplace's rule, an 18th-century formula devised to estimate the probability of sunrise. Over the past 40 years, there were six cases where the entropy method shouted "El Niño!" while the network stayed silent. In two of those, the event did occur. Applying Laplace's rule to this track record yields a 37.5% probability of El Niño in 2026.

If El Niño does come, it will most likely be of the eastern Pacific type, with warm water concentrated off South American shores. If it does not, there is a 70.9% chance of a neutral state, while La Niña (the cold phase) remains at 18.2%. The barometers did not agree, but now we have a probability map instead of a dispute.