The analogy was conceived by Eilam Gross in discussions with his colleagues. Eilam is a competitor from the ATLAS experiment (I work in CMS); but Eilam and I are also co-workers, in a way, since as selected members of the respective Statistics Committees of the two experiments, we participate in highly interesting and educative joint sessions, where we try to converge to a common statistics language and practice at CERN.

Get aboard the Higgs Car:

So, the idea is the following: imagine a rocket or even a car as an analogy to the SM. The car is moving very fast and smooth. We (experimental physicists) wonder how it runs. We come up with one major prediction: it must have a first class engine. That is our Higgs Boson.I think this is a simple and nice explanation of what we are currently doing at the LHC, and why the 95% upper limits on the Higgs boson cross section are to us just as interesting as anybody else's fancy 2-sigmaish bump in a mass distribution.

So we chase the car with our test vehicle (accelerator) in order to examine it. However the car moves very fast, so we need to increase our test vehicle speed (energy). As we get closer, the Higgs car is trying to increase its speed, but at one point we finally catch up. And then we are sure when examining it while running that it has an engine.

So we either find the engine or alas.... No fancy engine! No nothing: the SM car is running full speed and there is no engine inside. At least not one that we are familiar with. Isn't that even more interesting then finding an engine, as fancy as it is ? The wonder of all wonders: a car is moving smoothly but there is no engine inside. So how does it move?

Thanks Eilam for the contribution to this blog... And if you fancy it, be sure to accept an invitation to a guest post on a topic of your choice here!

Tommaso,

A quick statistics question on the usual way of presenting the Higgs search

If you look at enough possible Higgs mass points on a single graph wouldn't you expect to get "2 sigma signals" by pure chance? Or are the data points stringly correlated ?

I guess i don't understand what "2 sigma" means when the search is over many possible mass points,

Thanks

Paul