In the days leading up to April 6, 2009, citizens of L’Aquila, a small town in the Abruzzo region of Italy, began feeling tremors in the earth. However, local seismologists determined the shocks were not strong enough to warrant an evacuation. This was an unfortunate underestimation—disaster struck when an earthquake registering 6.2 on the Richter scale hit the town, killing more than 300 people.

The incorrect analysis would come back to haunt the group of six researchers and one government official as they found themselves defending their reasoning before a jury more than two years later. On Oct. 22, 2012, the seven were convicted of involuntary manslaughter, handed six-year prison sentences and mandated by the court to pay nearly $10 million to the families of earthquake victims. Their appeal began Oct. 10, 2014.

The scientists were not on trial for failing to predict an earthquake. They were accused of providing citizens with “inaccurate, incomplete and contradictory communications” and of failing to discuss preempting risks and evaluate whether the activity leading up to the earthquake could have led to the disaster. This statement implies that the scientific consensus about earthquake predictability is wrong.

Earthquake prediction is not a firm science. The Italian government expected a wholly accurate forecast without accepting the margin of error that is built into probability.

“Mother nature is an inherently chaotic system,” said Jennifer Strauss, external relations officer for the University of California, Berkeley Seismological Laboratory.

According to Strauss, certain earthquakes have repeatable characteristics that allow for some estimation of probability, but factors of time—how long earthquake activity has been measured in a specific region—limit the ability to forecast.

“We [can’t] say on this given day, at this time, at this location, there’s a certain probability of an earthquake,” she said. “If you want scientists to be making statements about what hazard probabilities are, you also need to have some mechanism on the other side to either say ‘We’re going to accept these probabilities or we’re not.’”

No reputable scientist would claim to be able to say with conviction when and where a quake is going to happen and how strong it would be.

Earth scientists can identify particular areas at risk of earthquakes depending on the amount of information they have, according to the Geological Society of London. Forecasts about the likelihood of a major earthquake in a certain area during a given period of time are currently much more reliable in the long-term than predictions that deal with days or months. The U.S. Geological Survey states that its approach has been to “focus on providing long-range forecasts of the likelihood, locations and impacts of damaging earthquakes.” Although Italy has a long history of earthquake activity, based on these assessments, the seismologists at L’Aquila would have had better luck trying to predict an earthquake 10 years down the road rather than the one that ended up being only days away.

If the Italian legal system deems their witchhunt necessary, it would be more sensible to blame the officials who, based on the analysis, made faulty claims about the likelihood of the earthquake. Bernardo De Bernardinis, a Civil Protection Agency official and one of the seven convicted men, said that the foreshocks were an indication of a continuous discharge of energy and not a sign that an earthquake was imminent. After the earthquake occurred, these claims came under fire as misleading at best and completely inaccurate at worst.

Denying the scientists’ appeal would place the blame on the science, rather than the policy. The court must provide evidence of the defendants’ intentions to misinform officials and downplay the risk of a seismic event when they had no motivation to do so.

If the Italian government does not to rule in favor of the researchers, it could really drive its point home by also indicting the city’s civil engineers for not anticipating how L’Aquila’s antiquated infrastructure would fare in a natural disaster. The Italian government could also blame the Department of Civil Protection itself, which was cited in an October 2010 article in the Journal of Natural Resources Policy Research for being “undeniably poorly developed and under-resourced. In short, it was unable to cope in any way with a major disaster.”

The seven men in question were convicted based on an emotional argument that overrode objective analysis of the initial message they intended to deliver. The burden on government should be to prove that the convicted men intentionally deceived their fellow citizens in hopes that an unpredictable earthquake might actually occur. If they can’t prove that, the charges have no merit. That makes more sense than locking them up for not forecasting the impossible to begin with.