Understanding Finetuning: What Does it Mean & How is it Decided?

[ChrisVer]

What does it mean to say that you need a quantity to be tuned? I mean why would you ask for it? How could someone decide on it?

 

[ZapperZ]

What does it mean to say that you need a quantity to be tuned? I mean why would you ask for it? How could someone decide on it?

You really should include the context that this is used. Otherwise, this becomes utterly vague!

Zz.

 

[ChrisVer]

I'm sorry...I thought by getting a general idea of what tuning is might give me an answer to my original question...
Well the context I'm talking about is the K^{0}-\bar{K}^{0} mixing in SQCD... and more specifically the last equation in this attachment... Although I had another topic with that, I guess this topic is somewhat different... (see attachment for the next).
You have a susy contribution of the mixing diagram, which has two parameters I guess which you can tune to make this contribution small...
One is m_{d} the universal squark mass... and the other is Δm_{i} their difference... If I want the contribution to be small then I can either send very up high the universal mass of the squarks, or make very small their differences... (I don't think I could play with the U matrices of flavor-mixing appearing from going to flavor eigenstates).
Is there any physical reason I can choose one out of these two?
One possibility is that I couldn't send to infinity the mass of squarks, otherwise the Higgs radiative corrections would "go crazy"... On the other hand I don't see why one could choose the mass differences to be very small... In quarks that's not the case, since their masses range from ~5MeV up to roughly 175GeV...

https://www.physicsforums.com/attachment.php?attachmentid=70229&d=1401647024

 

[Miralansa]

I report you what I understand of fine-tuning.

My english is not much good, so forgive me.

When you build a new theory, you have a lot of free parameters, in Standard Model they are for example the masses of quarks, Cabibbo's angle, ecc
You have to set values for parameters that can reproduce experimental results. If you have many degrees of freedom you can have the same output, with a large space of parameters ( for two you can visualize the X-Y plane ).
When you have a lot of experiments against you theory, you have a small space of parameters, fine-tuning, which means to give values to parameters, you haven't freedom for them, it's called fine because if the parameters are a little different, your theory isn't working anymore.

The situation in susy is this, because the Higgs is 125 Gev, you need to give at the s-top a mass of ~10 Tev, but you hope, for LHC's sake, that it's ~4-5 Tev. You also need to put the other s-quark >> 10 Gev . You say that you can't find a justification to put masses very close, it's true, but it's what fine-tuning means, you have a little space of unnatural parameters. Unfortunately the Higgs's mass says that to us, if it was less, you would say that Susy was right. If it was more you would say that SUsy was wrong, in the middle you can say that with fine-tuning the theory is right, but it's the reasone because Susy is losing a lot of supporters right now.

I hope that this will help you, send me a message if you want to talk about this.

Edit: My mistake I wrote Gev but thought Tev

 

[p-brane]

the Higgs's mass says that to us, if it was less, you would say that Susy was right. If it was more you would say that SUsy was wrong, in the middle you can say that with fine-tuning the theory is right, but it's the reasone because Susy is losing a lot of supporters right now.

There's not a single bit of experimental evidence that supersymmetry is wrong and an enormous amount of theoretical evidence that it's right. All experiment indicates right now is that the lightest supersymmetric particle is too heavy to for the LHC to produce. Get your facts straight.

 

[Miralansa]

I do an error in my previous post, now you can understand that I say the same thing, it's unlikely that s-quarks will be produced to LHC.

I don't understand what do you mean with "theoretical evidence" .

The experiments say that standard model works fine, like for B_s\rightarrow\mu^{+}+\mu^{-} , so you can always say "s-quarks are to higher energies" , yes but it isn't an experimental evidence.

I suggest to read this http://arxiv.org/pdf/1211.0004v1.pdf

From your nick I understand what you study, I'm sorry to say something that you don't want to hear. But the way, I'm working on Susy right now, I hope that it will be right.

Please don't assault me, I'm tired to argue always on the same things.

Best regards.

 

[ChrisVer]

Be careful p-brane... When we are talking for SuSy and phenomenology, we are dealing with models (in this case MSSM)... Models can be ruled out by experiment... For example the Higgs mass at the moment is at extreme point of MSSM-SM predictions... if it was a bit higher, then MSSM would be ruled out because it cannot give the correct Higgs mass from its vev+ radiative corrections.
You can of course play with the model, but it turns out that you should have LSP at the TeV scale, otherwise then MSSM will be ruled out (otherwise SuSy won't be able to describe the gauge hierarchy problem that is one theoretical reason it exists).

For example I found 3 ways to deal with that problem...
1. Is to send the universal mass very large... at least for the first and second generation, since their yukawa couplings are small, and thus you don't get hierarchy problems... In that case all FCNCs loops are suppressed...
Also the prevention of large loop corrections of U(1)_{Y} D-terms to the Higgs field then require the \sum_{i} Y_{i} m_{squark,i}^{2} < O(1TeV^{2}) - I don't really get this problem.
Another problem is when you implement such spectrum at high energies. In that case you have two loop contributions to the Renormalization Group equations due to SU(3)_{c} interactions which tend to drive the stop quarks masses to negative values (thus charge+color symmetry breaking- isn't that weird? SU(3) breaking itself? also isn't it good for the matter-antimatter asymmetry?)
But this model also allows contraints on flavor conserving CP-violating amplitudes (eg from those yet unobserved magnetic dipole moments of neutron and electron) although soft susy terms contain violating CP phases of the order of unity.

2. Alignment between quarks and squarks mass matrix so that they both get diagonal in the same basis.

3. make the masses almost universal. In that case you can still have high amplitudes for FCNCs which are though suppressed by a super-GIM mechanism.

 

[ChrisVer]

MSSM on the other hand could indeed lose supporters, but that's not the case for SUSY in general (as a mathematical object), since it's indeed needed for the string theory to "hold" (but who can ensure me that it "holds"? he shouldn't tell me- he should tell the world and get the nobel prize)...
For example you need local susy in order to get the correct vacuum energy after you break susy... at least to me it seems a better choice to have gravitational mediation rather than choosing to charge your hidden sector with some symmetries (which would need the existence of extra particles which are yet unobserved). Also this kind of thing could allow the gravitino to be the LSP...
gravitino sounds fine, since it comes as the spartner of graviton - and the local susy mathematics/algebra gives you Gravity itself...
(that's getting way out of the topic, I'm not to support SuSy here but try to understand how this FCNCs are suppressed and what the consequences are for each choice)

 

[dauto]

The bottom line is that MSSM is fine tuned (Good I have always been fairly superstitious that the MSSM is oversimplified. It's good to see it go). SUSY on the other hand is alive and well. In fact, the models I worked on in my Phd based on SO(10) GUT kept producing Higgs masses in the range of 125 to 130 GeVs. My adviser would say "These masses are coming up higher than usual but I thinks that's OK".

 

[mfb]

There's not a single bit of experimental evidence that supersymmetry is wrong

Which possible experimental result would you regard as such evidence?
How could experiments do better than excluding parts of the parameter space?

All experiment indicates right now is that the lightest supersymmetric particle is too heavy to for the LHC to produce.

if it exists
This is also true for all other proposed extensions of the SM (with an alternative "the couplings are too weak" instead of the mass for some models). The experiments don't suggest "there is a particle too heavy to see" - they just say "where we can look*, there is no particle".

*this is not limited to the LHC energy due to contributions to other processes, like B_s -> µµ, for example.

 

[Bill_K]

this is not limited to the LHC energy due to contributions to other processes, like B_s -> µµ, for example.

For the benefit of ignorant persons like me, here's an explanation of what B_s -> µµ might have to do with supersymmetry.

 

[ChrisVer]

could b & s energetically couple to top and Ws? I mean this process should be suppressed by a factor \frac{1}{M_{W}^{2} m_{t}}

 

[dauto]

I said in post #9 that I'm superstitious. I meant to say I'm suspicious. (Darn auto-correct...)

 

[mfb]

For the benefit of ignorant persons like me, here's an explanation of what B_s -> µµ might have to do with supersymmetry.

Just as an addition, as the article is from 2012: Both LHCb and CMS found the decay in the meantime, with a significance of roughly 4 sigma each. And directly at the SM value.

 

[kurros]

There's not a single bit of experimental evidence that supersymmetry is wrong and an enormous amount of theoretical evidence that it's right.

There is experimental evidence that it is wrong if you believe that the hierarchy problem is in fact a problem.

My favourite analogy is this: consider a table, with 100 cups sitting upturned on it. I tell you that I have hidden a coin beneath one of the cups. You then look under 50 of the cups; the 50 cups you thought it most likely that I hid the coin under, for whatever subjective reason. By this stage, you probably start to get suspicious that I lied to you and there is no coin. Certainly by the time you looked under 99 of the cups and still did not find the coin, you would think this.

The SUSY model space is vast, sure. But solving the hierarchy problem gave us good a-priori reasons to think certain of those models more likely to be describing what Nature does than others. The LHC has indeed now looked under many of those cups we (or at least many people) thought most probably contained the coin, so we are quite justified to increase our subjective belief that maybe SUSY doesn't exist after all.

It is certainly not definitive, and the conclusions vary with your thoughts about the SUSY model space, but it is perfectly rational, and certainly evidence. It is a subjectivist argument, sure, but you would be somewhat mad not to accept the subjectivist argument in the coin case, and I argue that the SUSY case is not fundamentally different, just more complicated. The subjective chance that SUSY will be found one day is certainly not zero, for me at least, but it is quite a bit lower than it once was.

 

[ChrisVer]

For once again- it's better to avoid arguments like "SUSY is wrong"... SUSY is not wrong because it's a mathematical object, and it has the freedom to exist anywhere between TeV to the Planck's scale... Of course, in the last case it cannot solve the gauge hierarchy problem... but that's OK... the hierarchy problem might be the strongest theoretical trump card of SUSY, but there are still more indications of its existence (like the DM particle content or string theories).
For example :
http://www.physics.ntua.gr/corfu2013/Talks/g_ross1@physics_ox_ac_uk_01.pdf
(C)GNMSSM gets a better value (check pg76)...

Also page 9 gives some results about the searches in ATLAS... where you can see there are energy scales which we haven't been able to "see".