Since the first exoplanets were discovered in the 1990s we have found more than 3,500 worlds beyond our sun. Roughly a third of these are less than twice the size of Earth. It is no surprise we are beginning to wonder if some these worlds could be not just Earth-size (more or less), but also Earth-like. Unfortunately, the data we currently have cannot tell us.

Planet detection techniques typically provide only one of two pieces of information: a planet’s size, either in terms of its radius or its minimum mass (or sometimes both); and the amount of radiation that reaches the planet from its star.

Knowing the size of a planet is only broadly informative. If it has a mass or radius similar to Jupiter, then we can reasonably assume it is a gas giant with a deep atmosphere and no habitable surface. If the planet is approximately Earth-size, then it is more likely to have a rocky surface with a thin atmosphere. Ironically, most of the exoplanets that have currently been found sit in between these two regimes. These so-called “Super Earths” have no analogue in our solar system, and may either be giant rocky planets or mini gas giants.

The small number of cases where we have been able to measure both radius and mass, and can thus calculate an average density, have only added to the confusion. Roughly speaking, a planet 40 percent larger than Earth is likely to be gaseous, but there are many exceptions in both directions.

Interpreting the radiation received from the star is equally vexing. This energy can be converted into an equilibrium temperature equivalent to the surface temperature on a planet with no atmosphere. For Earth, our equilibrium temperature is 0 degrees Fahrenheit (around –18 degrees Celsius), well below the freezing point of water. Our atmosphere traps heat to lift this to a global average of 59 degrees F (15 degrees C)—a value conducive to liquid water and life. The heat-trapping properties of an atmosphere are highly variable, however, and depend on its thickness and gas composition. Our atmosphere raises Earth’s temperature to from below freezing to a life-friendly level. On Venus, by contrast, where the equilibrium temperature would be a comfortable 80 degrees F (about 27 degrees C), its thick atmosphere, made mostly of the greenhouse gas carbon dioxide, boosts the actual temperature to a hellish 860 degrees F (460 °degrees C), turning this Earth-size planet into an inferno hot enough to melt lead.

Clearly, a planet’s equilibrium—or atmosphereless—temperature says nothing about its habitability. Our planet’s life-supporting conditions stem from a long list of intertwined properties that include a magnetic field that protects against solar flares, retention of water (which is essential for life as we know it, and which, in vapor form, is a greenhouse gas) and the composition of its rocks, which along with plate tectonics create a carbon–silicate cycle.

Moreover, with only one example of a planet that can support life, we do not know how variations in any of these properties affect habitability. Even if we could measure them, we have no way to know how quickly we change from Earth-like conditions as we deviate from the Earth value.

In short, a quantitive measure of habitability is impossible.

If this sounds depressing—take heart! The next generation of telescopes will be able to study the absorption of starlight as it passes through an exoplanet’s atmosphere. Known as spectroscopy, this can reveal what gases are in its air, which is linked to what is happening on its elusive surface. The right mix of gases may indicate rock type and geology and perhaps even, the whiff of life.

The challenge is that time on these new instruments will be in very short supply. So how do we select the best candidates for observation from the list of 3,500 possible targets? This is why astronomers and planetary scientists have developed metrics to rank planets for sample selection.

These metrics are familiar to those browsing the news. The “habitable zone” is the region around a star that receives a similar amount of radiation as Earth. If an Earth-clone were to orbit in this area, it could maintain liquid water on the surface. A second metric is the Habitability Index for Transiting Exoplanets, or HITE. This uses a slightly modified habitable zone and favors planets with smaller radii and more circular orbits. A third is the ESI, or Earth Similarity Index, which compares a planet’s properties to Earth’s value.

These metrics zoom in on the subset of planets most similar to Earth in size and radiation levels. What they do not do is measure the probability of a planet being able to support life. Without a way to measure surface conditions, it is completely impossible to say that a planet with a high metric value is more likely to be habitable than the planet below it in ranking.

This is easy to demonstrate. In the sun’s habitable zone lies Earth, but also its barren moon. Similarly, if we were to observe Venus as an exoplanet and measure size and equilibrium temperature, we would calculate an ESI of 0.9, where 0.8 is declared threshold for “Earth-like.” Exoplanet Kepler 442 b scores higher on the HITE metric than Earth; but all we really know about it is its radius and radiation levels.

Over-interpreting these selection tools to claim they measure habitability is a dangerous game. It is absurd to suggest we can assess something as complex as life-supporting conditions based on just two properties, neither of which directly probe the relevant environment. It is equivalent to judging someone’s personality based on height and the distance between the eyes. The proliferation of such statements both in the popular media and even scientific literature risks planetary scientists being taken less seriously.

If the public loses faith in the field, it will become difficult to secure funding for future planet-finding missions. Why would money be allocated to identify habitable planets if scientists have been claiming they have already been discovered?

In a comment to Nature Astronomy we therefore suggest a change in language is needed so sample-selection metrics no longer include “habitability” and related terms in their name. The habitable zone, for example, could be renamed “the temperate zone” to place the focus on the amount of radiation a planet receives, rather than on its ability to support life. Such metrics should also go to zero if detectability is not possible, such as when a planet is too distant for current or planned instruments.

In the last 20 years we have learned planets are common and form in the strangest of circumstances. We’ve found worlds orbiting binary stars like Star Wars’s Tatooine, circling the corpses of dead stars, and others with no star at all. We are now on the verge of sending missions that will probe what these alien worlds are really like. It is a plan that will help answer where we come from and whether we are alone. Let’s not screw this up for a quick headline.