3 ms·

> The hard part is tweaking the properties of the family of Higgs Bosons so they predict the results of the experiments where the 125GeV and 95GeV are seen, but

> The hard part is tweaking the properties of the family of Higgs Bosons so they predict the results of the experiments where the 125GeV and 95GeV are seen, but they don't cause weird results in all the other thousands of experiments where the Higgs Bosons have a negligible effect. These don’t sound like “risky predictions” at all! We already know everything that the model should output? Popper is spinning in his grave. And a real theoretical physicist would have been able to tell you “95GeV” before seeing it.

gus_massa 5y ago
I agree that "predictions" is not the right word. It's easy to explain the problem looking at the Figure 7 of the preprint. The fist graph shows how many times an experiment produced a pair of photons. The x axis is the energy of the photons, and the y axis is the number of events (I'm not sure about the scale here). The dashed line is the theoretical result without the second boson IIUC. The green zone is the range of a difference of ±σ. The yellow zone is the range of a difference of ±2σ. The continuous line is the experimental data. As expected, the experimental data is close to the theoretical result, but it is not exactly equal. It moves most of the time in the green zone of ±σ. Sometimes it moves in the yellow zone of green zone is the range of a difference of ±2σ. The interesting part is that in some parts it goes outside the yellow some. For those energies the theoretical model is too different of the experimental data. Back to your objection, the idea in the paper is to make a model with a few parameters, and then optimize the parameters until the curve of the new model fits the experimental data as much as possible. It's not enough to say 95GeV, the model has more parameters and also they must verify that the new model modifies the range of energies where the difference is big, but does not modify too much the other ranges.
- progrus 5y ago
  Well also, in my partly-serious physics trolling I left out that maybe the model could then predict a value for the next in the Higgs family (if such a thing exists), which according to my limited understanding doesn’t seem impossible.

gus_massa 5y ago

I agree that "predictions" is not the right word. It's easy to explain the problem looking at the Figure 7 of the preprint. The fist graph shows how many times an experiment produced a pair of photons. The x axis is the energy of the photons, and the y axis is the number of events (I'm not sure about the scale here). The dashed line is the theoretical result without the second boson IIUC. The green zone is the range of a difference of ±σ. The yellow zone is the range of a difference of ±2σ. The continuous line is the experimental data. As expected, the experimental data is close to the theoretical result, but it is not exactly equal. It moves most of the time in the green zone of ±σ. Sometimes it moves in the yellow zone of green zone is the range of a difference of ±2σ. The interesting part is that in some parts it goes outside the yellow some. For those energies the theoretical model is too different of the experimental data. Back to your objection, the idea in the paper is to make a model with a few parameters, and then optimize the parameters until the curve of the new model fits the experimental data as much as possible. It's not enough to say 95GeV, the model has more parameters and also they must verify that the new model modifies the range of energies where the difference is big, but does not modify too much the other ranges.

progrus 5y ago
Well also, in my partly-serious physics trolling I left out that maybe the model could then predict a value for the next in the Higgs family (if such a thing exists), which according to my limited understanding doesn’t seem impossible.