Is the world overreacting to the act of terrorism at the Boston Marathon? Judging from the available research, what looks like a knee-jerk response might actually be quite sensible.

U.S. authorities are not alone in stepping up security after the Boston attack, which left three people dead and more than 170 injured. British officials called for a review of the security measures planned for the London Marathon, the next major international event in the sport. In France, security patrols in public places have been reinforced.

It may seem ridiculous for governments to rush in police reinforcements and make people take off their shoes in airports after a successful terrorist attack. Their actions will not bring back the victims, and succumbing to paranoia merely plays into terrorists' hands. The increased security always slackens after a while, so why bother?

That logic, however, would hold only if terrorist attacks were random. If one is merely tossing a coin, there is no point buying insurance against the outcome of any particular toss. Yet the attacks follow a different pattern than coin tosses. And there is a statistical logic to boosting security after a highly publicized attack and slackening it when it is clear that the most recent burst is over and the "quiet" period has begun.

The consensus among academics involved in "physics of terrorism" research is that attacks are governed by a power law, with a small percentage of major incidents accounting for an outsized share of deaths. The distribution in terms of death toll and frequency is similar in just about every country that has an active rebel underground, and pretty much the same globally. This statistical evidence suggests there is some kind of underlying social mechanism to all modern terrorist warfare that makes irrelevant specific considerations like the perpetrators' cause or the city they choose to victimize.

A seminal 2009 article in Nature magazine described yet another statistical pattern common to terrorism: its "burstiness." Prolonged quiet periods are followed by frantic and deadly activity. The researchers suggested that this may be the result of competition for media attention. Terrorism, after all, is so called because its goal is to instill terror rather than kill the maximum number of people or blow up the most buildings. One group, or even a lone fanatic, gets terrorism back into the media's sights, and other groups or individuals jump on the bandwagon to boost the effect.

As far as the physics of terrorism is concerned, Boston was not a severe event. In a 2012 article, Aaron Clauset and Ryan Woodard set the threshold at 10 casualties - a category that included about 8.3 percent of all terrorist attacks from 1998 to 2007. After correcting for Iraq and Afghanistan, which have caused a dramatic increase in the number of deadly events in recent years, the share of severe events remains fairly constant - another sign that global terrorism obeys consistent statistical laws.

If attacks come in bursts, it is conceivable that insurgents could follow with larger acts, seeking to take advantage of people's frayed nerves after the Boston event. According to Clauset and Woodard, there is a 19 percent to 46 percent chance of a "catastrophic" event on the Sept. 11 scale occurring in 2012-2021. The forecast is as imprecise as any estimate based on historical data. The point is that the probability is too high to ignore.

It may be that the modern world's universal connectedness is part of the mechanism that assures the non-randomness of terror. If the "burstiness" of attacks is a direct result of media attention, then people who value their freedom of information will have to put up with the inconveniences of tightened security, particularly in the wake of any well- publicized attack. The alternative would be to suppress information, and that is definitely the greater evil.

Leonid Bershidsky is an editor and novelist based in Moscow. She wrote this for the Bloomberg News.