Earlier this year the CDF collaboration published a measurement of the mass of the W boson which caused something of a stir. I even made it the topic of my slot at the rescheduled “Nine lessons for Curious People” in April. The interest arose principally because the measurement did not agree with the Standard Model of particle physics. It still doesn’t.

The Standard Model is very good as far as it goes, but it leaves important questions unanswered, so we are always on the lookout for a bigger, better theory with even more explanatory power. A very important method of “being on the look out” is to make precise measurements of things which the Standard Model predicts, and see if the predictions are right.

If they aren’t, that should be a big clue to the sought-after better theory.

This is a bit of a different approach from the “make up new particles then go look for them” methodology we are sometimes accused of, but it does sometimes motivate the introduction of new, hypothetical particles when a discrepancy like this shows up. Theorists try to come up with extensions of the Standard Model which would accommodate the discrepancy (in this case a slightly-too-massive W boson) and if possible include more explanatory power at the same time. These extensions do sometimes imply new particles.

So this is where a preprint some colleagues and I put on the arXiv today comes in, which is why I’m writing this post now.

There is a class of models that can explain the W mass discrepancy called “Type II seesaw models”. The preprint describes them in detail, but essentially these models introduce more Higgs-like particles, which can give quantum corrections to the W mass, making it agree with CDF.

An attractive feature of these models is the fact that they connect to neutrino physics; they explain why neutrinos have such very, very small masses. In fact, they were invented to deal with neutrino physics, so the possibility that they might also line up the W mass is a bonus, and potentially very interesting.

To cut to the result:

• Julian, one of the authors on our new preprint, determined in a previous paper the parameters of these models that would explain the W mass discrepancy. These parameters implied that my experiment (ATLAS) and our colleagues across the other side of the LHC (CMS) might be able to see the new particles being produced in our collisions, and that the searches for new particles that had been done so far did not rule this out.
• We simulated the production of these particles and found that if they were being produced, there were measurements that ATLAS has already made that should have shown them up. These included the four-lepton measurement I wrote about here, as well as two other measurements involving the production of leptons (see the preprint for references). These configurations of leptons can be produced by Standard Model processes, but can also be produced by the new particles in the see-saw model.
• The measurements agree well with the Standard Model; no sign of the new particles. So, unfortunately, we’d better come up with a different solution.

So it’s a null result, but one that saves a lot of nugatory work searching for these new particles when the answer was already in our data.

I find this way of working satisfying.

The amount of effort that goes into making the measurements is enormous – not just the building and running of the experiments, but also the data analysis, making sure your measurement is as accurate and as unbiased by theory assumptions as possible. The reward is that the measurement, once made, is very widely applicable, as we see here. The ATLAS measurements pre-date the CDF measurement and Julian’s first paper, but they have informed the discussion very quickly, and relatively easily. (This is also thanks to the excellent software tools that exist to simulate the consequences of new theories.)

This is the way I hope we do physics over the next decade or so at the LHC. We should focus less on searching for new particles, and more on measuring what actually happens – then compare it to the Standard Model and whatever new theories become interesting.

Null results are important, but their value depends on how much credibility you give the theory you were testing. If we had ruled out the Higgs boson, for example, that would have been as big a deal scientifically as finding it – arguably even bigger, since it would have ruled out the Standard Model. But designing a dedicated search and ruling out every new model our theory colleagues cook up (or, optimistically, every model but one) is not a good use of resources, in my opinion. Some theories may be better motivated than others, but in the end it is quite a subjective judgement call. More importantly, it is beholden on us as experimentalists to characterise nature in the new regimes of physics to which the LHC gives us access, and to quantify how well the Standard Model is doing at describing them.

If we do this right, then one bonus is papers like ours today; a relatively rapid recycling of a lot of work to get new information on an intriguing puzzle.