Murphy’s Law

In the late 1940s, at Muroc Army Air Field in California's Mojave Desert, soon to be Edwards Air Force Base, a small team was slamming rocket sleds to a violent stop. The project was MX-981. The point was not a land-speed record for its own sake. Pilots were dying in crashes, and no one had solid numbers for how much sudden deceleration a person could survive if the straps and the seat did their jobs. The physician in charge was John Paul Stapp. He often rode the sled himself rather than ask someone else to take the risk.

The early Gee Whiz runs at Muroc, beginning in 1947, produced serious injuries and valuable data; Stapp endured very high g-forces on those runs. His most famous ride came later, on 10 December 1954, at Holloman Air Force Base in New Mexico, aboard Sonic Wind I: 632 miles per hour in five seconds, then a stop in 1.4 seconds at 46.2 g. His eyes filled with blood from burst capillaries. He used the public attention that followed to push for better crash protection, including automobile seat belts.

Murphy's Law enters this history in 1949, during the Edwards and Muroc work. Captain Edward A. Murphy Jr., an engineer from Wright Field in Dayton, brought strain-gauge sensors to measure the violent stops more accurately. On one run the instruments read zero. The sensors could be wired two ways, and they had been wired the wrong way.

Stapp later told people that the team stayed careful by taking Murphy's Law seriously. The earliest published link of Stapp to that name is Lloyd Mallan's 1955 book Men, Rockets and Space Rats, which quotes Everything that can possibly go wrong will go wrong as a saying Stapp used. The gloomy idea itself is older than any of them. In 1877 and 1878 the engineer Alfred Holt wrote that anything that can go wrong at sea generally does go wrong sooner or later, and still earlier writers had said much the same.

The useful part of the law is the part Stapp's work actually lived by. Expect the failure path. Design so the wrong path is hard to take.

There are at least two different laws hiding under the same name, and they lead to different habits.

The popular version is fatalistic: the universe is out to get you, toast falls butter-side down, and planning is almost a dare. That version is catchy, and it is not very helpful. It also is not a law of physics, even though people sometimes joke that it is a cousin of entropy.

The engineering version is practical. If a connector, a switch, a form, or a procedure can be completed in more than one way, and only one of those ways is safe, then over enough repetitions someone will take the unsafe way. They will be tired, hurried, new to the job, or simply human. The failure is then not a mystery and not a curse. It is a path the design left open. The reply is to close that path: make the part fit only the right way, add a check that must be passed, or make the wrong state obvious before it can do harm. Aviation mechanics in the 1950s already phrased it this way: if an aircraft part can be installed incorrectly, someone will install it that way.

That is why our cartoon turns the dead sensors in Murphy's hand into a design lesson. A sensor that fit in both ways is an invitation. Build every part so it cannot be put in wrong is the productive half of the story. The grumpy half, anything that can go wrong will, is what most people remember. It is also the half that quietly excuses a sloppy design, because it blames luck instead of a two-way fitting.

Look for anything that can be assembled, switched, or confirmed in more than one way, with only one way being safe. If you can plug it in backwards, wire it in reverse, sign the wrong line, or skip a check and still make the setup look complete, assume that path will be taken eventually. Then change the design: key the part, add a second check, or make the wrong state impossible to miss. Use the law as a planning tool before a test, a launch, or a busy day, not as a shrug after something breaks. And keep the tone right. The point is not that the universe is against you. The point is that a two-way fitting is a decision, and you can usually make a better one.