Love and loss in the time of colliders
Alas, even the most promising data can let us down, writes Katie Mack.
Isaac Asimov once said, “The most exciting phrase to hear in science, the one that heralds new discoveries, is not ‘eureka’ but ‘that’s funny’”.
While the popular representation of scientists is that they are focused, single-minded workers jumping from discovery to discovery, the reality is a lot more complicated. Finding something strange in the data, rather than having a “eureka” moment, is usually what leads us to a new or more complete theory.
That’s why physicists around the world got so excited when, in late 2015, they detected a small but tantalising bump in a plot produced by the Large Hadron Collider (LHC).
In particle physics, confirming a theory is an important (and very satisfying) step, but anomalies are the path to real progress. The Standard Model of Particle Physics – the fantastically successful framework for all particle physics seen so far – has passed every experimental test we’ve thrown at it.
But it has some major theoretical problems. Theories such as supersymmetry and string theory are designed to solve those problems, but to know if we’re on the right track, we need to see something in the lab that doesn’t fit with what the Standard Model predicts.
The plot that got everyone excited last year looked like it might show just such a disagreement.
The LHC works by colliding protons together and collecting information about what comes out. Sometimes, collisions produce, among other debris, two photons (particles of light).
In general, if you chart the number of collisions producing two photons, you’ll see that low-energy photons are produced much more often than high-energy ones, and the transition between low and high is a nice smooth curve. That’s what the Standard Model predicts.
But what physicists saw in the plot was not a smooth curve, but one with a bump – a high-energy “excess” registering 750 gigaelectronvolts (about 750 times the energy of a single proton at rest).
IN PARTICLE PHYSICS, CONFIRMING A THEORY IS AN IMPORTANT STEP, BUT ANOMALIES ARE THE PATH TO REAL PROGRESS.
When the results were announced, the particle physics community erupted in speculation. A bump like that could mean the LHC had produced a new particle, one the Standard Model didn’t include – a sure sign of new laws of physics at work.
The “diphoton excess” became the talk of the town. New theories were created, papers were published, and old theories were tweaked to “predict” the excess after the fact.
But there was a catch. Particle colliders count the number of events (in this case, photon pair productions) above a background level of random production of particles from other kinds of events.
That background is always present, and we know what it is on average, but sometimes there’s more and sometimes less. A bump could be just random chance.
The diphoton excess certainly looked significant: early estimates said there was only a small chance of it being a fluke. Still, in a long-running experiment, low likelihood events do occur once in a while.
Scientists needed more data to know for sure whether the blip was, in fact, a new particle or just an unfortunate coincidence in the background events.
In early August, LHC scientists had their answer, and the news was not good. The diphoton bump had vanished back into the noise. It was a random fluctuation after all. That meant more than 500 papers now contained detailed dissections of a signal that did not exist.
On the Netflix series Stranger Things, a science teacher quips, “Science is neat, but I’m afraid it’s not very forgiving”.
Even more apropos is a quote from American writer H. L. Mencken (slightly reworded): “For every complex problem, there is an answer that is clear, simple and wrong.”
As scientists, we always have to accept the possibility that a new and exciting development will be mercilessly killed by the next batch of data, and when that happens, we have to adapt.
No matter how elegant our theory, or how well we think it solves some long-standing problem, if the data don’t agree with it, we have to let it go and move on. At the moment, that means setting aside our diphoton theories and going back to the data to see what else we might find to challenge the Standard Model.
The next time something goes bump in the night, we’ll still be here, and we’ll be ready.