@Manishearth oh, I just saw that

@annav to continue our discussion from physics.stackexchange.com/a/13172/124, there's something I'm still not understanding. Earlier you were saying that the numerator and denominator of the "expected" ratio would be the same with large statistics, but now you're saying the numerator would be less than the denominator where there is no Higgs boson?

Perhaps I need to see an example of how these things are calculated... e.g. suppose you have 5/fb of data and you calculate the cross section of $pp\to\gamma\gamma +X$ in a particular energy bin to be 50 fb using the Higgsless standard model. Does that mean $\sigma_{SM}$ is 50 fb?

And for that energy bin, how is the "expected" value determined? You said it's a ratio of two MC results... the denominator is a simulation with large statistics, so shouldn't it produce a value of 50 fb? And the numerator would be based on 5/fb, so shouldn't it produce a value taken from a Poisson distribution with mean 10?

@DavidZaslavsky I didn't expect a delete, just a closevote :/ Alright, though--I can see why you may want it deleted ;)

@Manishearth It's kind of a network-wide thing, jokes shouldn't displace legitimate questions

@DavidZaslavsky Aah, just checked meta: meta.stackoverflow.com/questions/127909/…

*shrugs*

Good morning. Lets take the curve "expected" . The denominator can either be a large statistics monte carlo run or a direct calculation if possible. The numerator will have the poisson statistics of few event for the same calculation with the monte carlo. The x axis is the mass of the Higgs assumed in the modeling. For each bin a different Higgs mass is expected. The numerator not only has low statistics but is also an addition of events from many channels.

If you look at the crossections of higgs producing channels you will see curves going up and down according to the opening and increase/decrease of that particular channel crossections. The addition of these crossections together with the small datata statistics will create the wavy nature of expected, which diminishes as the real data statistics increases. Because of construction when the real data reaches high statistics, the numerator that is simulating the real data statistics

will be the same as the denominator and the expected curve will go to 1.

Now the "observed" is what nature is serving us, and if a higgs exists it will have one specific mass, so the ratio with the calculated crossection withe the varying Higgs mass, if the observed data had very large statistics, would be at 1. The rest of the x axis exploration of higgs mass would be below 1 because it would be missing the crossection contribution of the higgs at that mass. At that point the brazil plots are useless.

One would plot the Higgs mass at the x axis and it would show as a Gaussian or breit-wigner and the theoretical expectation would be overlayed on the plot. These brazil plots are just so as to extract any possible information from the data by excluding regions where the Higgs might be and highlighting regions where it is. Theoreticians get excited by this but the real

meat will be in the higgs mass plots for individual channels. IMO this is almost sleight of hand to increase statistics and expectations.

that should read "the higgs might not be"

That was for @DavidZaslavsky I have never made a brazil plot in my life. It is what I have gathered from the talks and plots.

Maybe I should clarify the :"The rest of the x axis exploration of higgs mass would be below 1 because it would be missing the crossection contribution of the higgs at that mass." The denominator in the "observed" is theory calculated with the specific higgs mass at that x. The real world data will have only one mass for the Higgs, the rest of the x axis will be a miscalculation with extra crossection for a fictitious Higgs at that x/mass.

as far as the denominator is concerned.

@annav Have you looked at this yet: meta.physics.stackexchange.com/questions/1159/… ??

@Manishearth if you put links to questions or answers on a line of their own in chat, you get a clever embedding thing. Like this:

@Energy I know, never thought of using it.. Thanks anyway :)

Works with wikipedia links and iirc pictures as wel

Good day

Is it possible for a black hole's absolute horizon (event horizon) to expand faster than the speed of light in a frame outside the hole? I don't see any causality issues with this--but I'm not sure...

Well, if the theory depends on gravitons transferring changes in gravitation then they cannot move faster than the speed of light, no?

@annav I guess not--but I was thinking of classical general relativity-- in this case the event horizon CAN be used to send a signal faster than light--but receiving the signal will amount to being swallowed by the hole and we're OK with causality going boom inside a singularity

As far as I can tell, the rate of increase of size of event horizon depends upon the rate of gain of energy. If we beam some photons with high energy, there's nothing stopping the hole from expanding faster than light

So, it doesnt violate causality, and looks possible (within the framework of GR). But I may be making a mistake in my logic