To Higgs or not to Higgs

Fabiola Gianotti, the ATLAS spokesperson, Guido Tonelli, the CMS spokesperson and Rolf Heuer, the CERN director at the end of the seminar on the Higgs boson Tuesday this week. The picture is from Jiri Kvita:

I’ve been a Higgs skeptic for years, but I must admit my skepticism feels less justified after Tuesday’s seminar. Don’t get me wrong, I am not convinced one way or another by what was shown this week by ATLAS and CMS, but it leaves a strong impression. I must remind myself that what has been seen so far by both LHC experiments can still vanish in the next year. It’s a good thing that we developed all these fancy statistical analysis tools, because our intuition would have us believe that we found it.

Let me remind you of the status of the Higgs search as of last summer, which I discussed here. At the Lepton-Photon conference at the end of August, the Higgs could still potentially have a mass between 114.4 GeV and 141 GeV, or above 476 GeV. These constraints had been set by a combination of ATLAS and CMS results. Basically, if the Higgs boson had a mass outside of these ranges, some excess in the data would have prevented the constraints to be extended that far.

One thing that makes the search for the Higgs boson fairly tricky is that it won’t look the same in our detector depending on its mass. It remains however that for a given mass, we know exactly what it should look like (or else, it can’t be called the Higgs boson). The Standard Model of particle physics allows us to predict the most likely mass of the Higgs boson but this estimate is limited by the precision of other experimental measurements like the mass of the W boson and the mass of the top quark. Nevertheless, we know that the Higgs is far more likely to be found between 114.4 GeV and 141 GeV than it is to be found above 476 GeV.

So what does the Higgs boson look like in our detectors in this low-mass range? Well the first thing to know is that the Higgs boson is unstable, so it will quickly decay to known particles. These known particles are what we will observe in the detector, but they will show up in special configurations. But even these special configurations can be faked by more ordinary events in the LHC collisions, so picking up a special arrangement of known particles among a flurry of other known particles is quite the challenge.

The trick to discover something new is to do a counting experiment. We know what the not-so-special events that would fake our Higgs signature looks like, and we know how often they occur. We can estimate how much of these we should expect. Then, we look at the data we collect at the LHC and we count how many events we observe, and we compare to how many events we should expect if the Higgs was not there. A strong excess over our expectation would indicate that the Higgs is there. To locate the Higgs on its possible mass spectrum, you do this kind of counting experiment in a number of finite mass ranges (from 115 to 120, from 120 to 125 and so on).

So you now know the basics to understand the new results. ATLAS and CMS didn’t combine their results yet, so I will have to go through what each experiment saw.


ATLAS managed to close down the gap quite a bit. The allowed range at low-mass has been narrowed to 115.5 GeV to 131 GeV. There isn’t much space left for the Higgs to hide. But that isn’t what is causing all the excitement.

The ATLAS result is a combination of three different kinds of searches for the low-mass region. There is the diphoton channel, where the Higgs simply decays to a pair of photons. There is also the so-called golden channel (because it is very obvious), where you end up with 2 pairs of electrons or muons, which is very unique and easy to detect at ATLAS. To complete this little family, we have the WW channel where the Higgs first decays to a pair of W bosons, each of which decay into a neutrino, and an electron or a muon.

There is a small excess seen in the diphoton channel around 126 GeV. This excess is reinforced by three potential signal events from the golden channel. Combining the three channels together, the excess is still there, and even a little bit stronger.


CMS has very similar results, although they have been able to take the exclusion range a bit further down by including more channels. The window where the Higgs could be according to CMS is between 114.4 GeV and 127 GeV. They also see excesses in the same channels, but what they see is weaker than what ATLAS sees.

Remember when I talked about how many events you would expect if the Higgs is not there? CMS has the sensitivity to exclude all the way down to 117 GeV. They would expect  to take the exclusion that far down if the Higgs wasn’t there. However, they couldn’t because they see too many events below 127 GeV. So what does this all mean?

It means that…

…we still don’t know. The excesses both experiments see are still too small to categorize as signal. Everybody knows from tossing coins that what you see never really match the probabilities you would assign to heads or tails, but you get closer and closer to these probabilities if you just keep on tossing coins. If you throw 10 coins, you won’t be too upset if you get 6 heads and 4 tails. It would be just a fluke. However, it would be a lot more surprising to throw 1000 coins and obtain 600 heads and 400 tails.

Given how few potential signal events we have observed, it is just too soon to rule out the possibility that these excesses are flukes. There is also something else that needs to be taken into account. A positive fluke could appear anywhere on the mass range we are scanning for the Higgs boson. The chances of this happening is indeed higher than the chances of a fluke appearing for one particular mass that we picked. You may have heard of this as the look-elsewhere effect.

Even if CMS and ATLAS see excesses at about the same place in different channels, you still need to take the look-elsewhere effect into account. In the end, neither CMS or ATLAS can claim to have their hands on a signal strong enough to justify claiming to see the Higgs.

We will know soon enough. As people in both collaborations keep working on refining their analyses, they can identify more signal through the same 2011 data. Also, other search channels (other ways for the Higgs to decay) will be added to the combination. If the Higgs really is where it looks like it could be, we could have a claim as early as next January. However, if it really is a statistical fluctuation, the 2011 data won’t be enough to tell for sure. But the 2012 data should be enough for that.

So there it is. Notice that I haven’t been my hard-nose, skeptical and toning-down self this time. That’s because I feel that the excitement is justified. Whether the Higgs boson exists or not, we are rapidly closing in on a definitive statement. I also think that CERN public relations handled this one very well. I am proud of my collaboration (which is ATLAS), and I am also proud that we have CMS as fierce competitors.

Note: If you think that the fact that Comic Sans MS was used on the ATLAS slides at the seminar is something worth having long discussions on the Internet, you are completely missing the point. Get over it, it’s a freaking font. Fabiola have more respect than you will ever have. Some of us are trying to not be irrelevant in this world, leave us alone.

This entry was posted in Science and tagged , , , , . Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s