Yesterday both Higgs boson-hunting experiments at the Large Hadron Collider separately reported signs of the Higgs boson with 5 sigma of statistical significance. That means there's only a 1 in 2 million chance that the result is due to background processes - and not a Higgs boson - enough to declare a discovery. But what if you combined the two results? Independent physicist Philip Gibbs, not a member of the teams responsible for detectors at the LHC, has done his best to do just that. The calculation gave him a staggering new confidence level 7 sigma - though it should be treated with caution as the calculation isn't official.
Both the ATLAS and CMS experiments separately combined data collected in 2011 and 2012 to produce the 5 sigma results. Most of the evidence came from two channels: a Higgs that decays into two photons, and a Higgs that decays into two Z bosons that then decay further into four leptons.
These are two of the easiest ways to look for the Higgs at the LHC because they can be measured precisely, unlike some other decay products like neutrinos that escape the detector without being seen. But the Z boson decay happens only once for every 12,000 Higgs particles, and the photon signal can be drowned out by other processes that also produce photons. Combining both channels helps overcome each one's weaknesses and gives more confidence in the result.
The same could be said for combining results from the two experiments - each has its own set of possible experimental errors, which the other could cancel out. But that option "is a controversial one, because once we combine results to get the final observation we can no longer use each experiment as a crosscheck for the other", cautioned Aidan Randle-Conde of ATLAS at the Quantum Diaries blog on Monday, two nights before the result was announced.
The fact that each experiment cleared five sigma on its own makes that less of a concern, so Gibbs, who blogs at viXra, did a rough combination on his own.
Combining both channels in both experiments over the past year and a half yielded a striking confidence level of 7.4 sigma. That means that the chance that both results were produced by background processes in the detector are less than 2 in 10 billion. Three sigma is considered "observation" and five sigma "discovery", so perhaps 7 sigma could be considered "certainty"?
Still, as the Associated Press reported on Monday, CERN spokesman James Gillies said that he would be "very cautious" about unofficial combinations of ATLAS and CMS data:
"Combining the data from two experiments is a complex task, which is why it takes time, and why no combination will be presented on Wednesday."
So the final count is still to come. Stay tuned.