Race For the "God Particle" Heats Up 397
SpuriousLogic writes "CERN is losing ground rapidly in the race to discover the elusive Higgs boson, its American rival claims. Fermilab say the odds of their Tevatron accelerator finding it first are now 50-50 at worst, and up to 96% at best. CERN's Lyn Evans admitted the accident which will halt the $7B Large Hadron Collider until September may cost them one of the biggest prizes in physics."
Re:...its a 50-50 chance (Score:4, Informative)
Wow thats a worse investment than that stimulus package
You mean it in jest, but the "stimulus package" (aka handout for the rich) is going to provide more ammunition for the robber barons to shoot at us with, whereas these colliders are going to lead to developments in science whether they find this particle or not.
Re:race? (Score:1, Informative)
Why is there a race?
The Tevatron is nearing the end of its working life. Once the LHC is working, it will be hard to justify continuing to fund it. However, at the minute, it's the only game in town, and might just scoop the LHC on finding the Higgs.
they would say that, wouldn't they (Score:4, Informative)
Re:race? (Score:3, Informative)
Now boil your brain on the fact that the very same thing exists in medical research, and feel the creeping horror at what that implies.
Re:How do you give odds for that? (Score:2, Informative)
Re:How do you give odds for that? (Score:5, Informative)
To discover the Higgs, we must show that given a theory without the Higgs, our data would only occur 1 in 2 million times we did an experiment like this, (5 sigma significance, standard for particle discovery) and of course the the difference in the data is consistent with a Higgs.
To exclude the Higgs in a certain mass range, we must show the opposite: if there were a Higgs, our data would only occur some very small percentage of the time (I can't remember the exact significance, but it is less stringent than discovery, again standard).
LEP already excluded masses below 114 GeV/c^2, and the Tevatron has excluded a small mass range around 160 or 170 GeV/c^2.
However, all that said, I disagree with the apparently official Fermilab line (50/50). We have a small chance of excluding all the available mass ranges, but the amount of data needed to go from excluding it if its not there to discovering it if it is there is huge. We would need several times as much data as we will have unless we keep running for quite a bit longer. Maybe we can get a chunk of the gov't stimulus package?
Without any data to base your odds on, you're just making some shit up. Not only is their level of precision low, but there is zero confidence.
Quite the contrary, sir, and I do somewhat resent remarks like these, although I understand they were made in haste in your frenzy to get first post. We have a tremendous amount of data, and we have theories that describe exactly what we're looking for. It's almost just a statistical game now. Our level of precision is in fact quite high (although not as high as is achievable at a lepton collider), and as I said above, we have excluded some potential Higgs masses to a high level of confidence.
You keep using that word.I do not think it means.. (Score:3, Informative)
...what you think it means.
Falsifiability: Which you don't have for mathematical axioms either.
http://en.wikipedia.org/wiki/Falsifiability [wikipedia.org]
Re:How do you give odds for that? (Score:2, Informative)
Probably pretty good. After all, God doesn't exist, but millions of people convince themselves that they are finding Him all the time.
There. Fixed that for you.
Re:How do you give odds for that? (Score:3, Informative)
They aren't searching. They are performing experiments for which current theory predicts certain results.
Re:How do you give odds for that? (Score:5, Informative)
Here is a nice graphic of what you described about the exclusions for a light-mass Higgs:
http://www.symmetrymagazine.org/breaking/wp-content/uploads/2009/02/higgsexclusionplotfy08.jpg [symmetrymagazine.org]
Where the percentages come from (Score:3, Informative)
I'm not a physicist. But I know that the Tevatron, since it is lower-energy than the LHC, relies on aggregating the data from many collisions to produce a data set in which to look for proof of the Higgs. In the article they said that they already have 8 collision events which seem to provide good hints that the Higgs does exist. Presumably they will need many more good hints and/or a lucky collision that produces direct proof.
There is not unlimited time--the actual quote was a 50% to 96% chance of finding the Higgs FIRST (before the LHC). It seems like this could be calculated by estimating (based on performance so far) how many tries it takes to produce an adequate data set and how long each try takes. Divide into the amount of time left until the LHC begins operating at full capacity and searching for the Higgs (The article says about two years).
Of course if the Higgs does not exist then neither cyclotron will find it and all bets are off.
Re:How do you give odds for that? (Score:3, Informative)
Nope. Here's how it works. Other observations [wikipedia.org] show that the Higgs has to have a mass between 170 and 285 GeV/c2, with 95% confidence. Assuming a given Higgs mass, Fermilab can do Monte Carlo simulations [wikipedia.org] of the results of their experiments, and they can determine a probability that the signal will show up in their data, and show up with a certain level of confidence. For instance, the article says, "And the chances are even higher - 96% - if its mass is around 170GeV (giga-electron volts)." What this means is that they set some level of statistical significance that would make them confident enough to publish a paper claiming to have found a Higgs. They expect to have some noisy spectrum with a fairly crappy-looking peak sticking up out of it, and they're willing to publish if the statistical significance of that peak passes some predetermined statistical test. Okay, so they run the Monte Carlo simulation 10,000 times, putting in a Higgs mass of 170 GeV. Out of those 10,000 simulations, 9,600 of them produce a simulated peak, at the right energy, that passes their criteria for statistical significance.
It's also possible that the Higgs doesn't exist [wikipedia.org]. There are models that are consistent with that, and also consistent with the experimental data. However, there are fairly model-independent reasons to believe that something new must happen in this energy regime, and it's likely that any experiment that can probe that energy range will detect the whatever-it-is.
What is not possible is that the correct description of physics consists of the standard model minus the Higgs. Such a theory doesn't have the necessary self-consistency and consistency with well-established experimental results.
Re:How do you give odds for that? (Score:3, Informative)
"I can divide by zero"
No you can't, whilst extrapolation may seem to imply the answer is infinity, division by zero is completely illogical thus has no place in mathmatics.
Re:How do you give odds for that? (Score:5, Informative)
To put it another way, reminding people of the way they were taught to divide in primary school, dividing by anything is splitting it up into that many groups. Exactly how can you divide something into zero groups? The answer isn't infinite, because that would imply creating more stuff to put in those groups. If you divide by zero, whatever it is your dividing has nowhere to go.