Was Halton Arp hard done by? Need some clarification. some pictures

[Chalnoth]

Wow, when you have to recur to such ofensive expressions as asinine and "you've lost your mind" I might be on the right track.

Or you could just learn a little bit.

In cosmology, the primary effect of dark energy is on the rate of expansion. So what we're fitting for is a function H(z). To date, we have a tremendous number of individual observations, from supernovae to galaxy distributions to the CMB, which all provide separate, independent constraints on this function. It is a fundamental impossibility for the addition of a single parameter to fit any potential H(z). It cannot be done.

Anyway I was developing what Twofish-quant posted:"If it turns out that we've messed up something, then LCDM is out the window. Before 1998, CDM was the model. People added L when CDM wouldn't work. I wouldn't be terribly surprised if we start talking about xLCDM."
is his characteization also asinine and is he out of his mind or is it just me?

No, his statement was perfectly reasonable. Yours was asinine, because it claimed something completely and utterly false, while at the same time being insulting.

 

[TrickyDicky]

I'll say it again, you tell me where the insulting part is: when you have a model that can be made to accommodate anything just by adding one more letter to its acronym (this was referring to this words by twofish-quant:"I wouldn't be terribly surprised if we start talking about xLCDM") their proponents might just be either fooling themselves or pulling your leg.

 

[turbo]

I'm pretty sure that his idea that these objects were ejected from galaxy cores was pretty nonsensical from the start, because not only was there no model whatsoever for what they actually were within Arp's idea, but they also had spectra of the quasars at that time. With no known physical method for producing such large redshifts except by either large recession velocities or gravitational redshift, Arp's idea that these redshifts were "intrinsic" was nutty from the start.

Well, put yourself in their shoes. (Arp, Margaret and Geoffrey Burbidge, et al). Yes, quasars have some pretty impressive redshifts. Logically, though, their redshifts can't possibly originate from the peculiar motions of quasars, since that would place the Earth in a very special place in the universe, with each and every quasar receding from us. It was not such a leap to consider that quasars may have some property(ies) such that their redshifts are intrinsic. It's easy to call such an idea "nutty" in retrospect, but look back a few decades, and try to imagine a better reason for the redshifts.

 

[Chalnoth]

I'll say it again, you tell me where the insulting part is: when you have a model that can be made to accommodate anything just by adding one more letter to its acronym

This is the insulting part, because it just isn't the case. I'll say it again: many other proposed models of the observed acceleration have failed. LCDM has survived because it fits the data better.

 

[TrickyDicky]

This is the insulting part, because it just isn't the case. I'll say it again: many other proposed models of the observed acceleration have failed. LCDM has survived because it fits the data better.

And that sentence in no way contradicts what you claim here, other models just don't have the plasticity or malleability to adapt to observations like supernovae Ia light curves. I don't know how that can be insulting, unless it is insulting only when I say it, I'm precisely highlighting that property of the concordance model, its ability to survive any observation.

 

[Chalnoth]

And that sentence in no way contradicts what you claim here, other models just don't have the plasticity or malleability to adapt to observations like supernovae Ia light curves.

The amount of placticity/malleability is set by the number of parameters. Adding one single additional parameter to the CDM model is a simple as you can possibly get and still explain the observations. And what's more, the particular parameter that is "added" was in the theory all along, it was just assumed to be zero because simple arguments showed it had to be very small. Nobody, however, has ever found a convincing way to set it to zero, so in a sense, it was never added in the first place. Instead it had been taken out previously for no good reason.

I don't know how that can be insulting, unless it is insulting only when I say it, I'm precisely highlighting that property of the concordance model, its ability to survive any observation.

Except it isn't able to survive any observation. If the CMB wasn't there, or if the cosmological parameters derived from CMB observations didn't match the cosmological parameters from baryon acoustic oscillations or supernova, then the theory would have basically been falsified. It is very very easy to come up with potential observations that wouldn't fit. But this just isn't the case.

 

[twofish-quant]

And this only goes to show that when you have a model that can be made to accommodate anything just by adding one more letter to its acronym, their proponents might just be either fooling themselves or pulling your leg.

You can't accommodate everything. If the primordial helium abundance goes below 20%, we have serious problems. There are a lot of things that we *could* have observed that would have tossed out the LCDM, but we didn't observe them.

The thing about LCDM is that it fits the vast amount of data that we have, so we are pretty sure that whatever the right answer is, it has to be close enough to LCDM to explain what LCDM explains. Just like it may be that GR and Newtonian physics may be wrong, but we know that at some levels it's close to what is going on.

Maybe it's not GR that is wrong but the specific solutions with problems at large scales.

GR gives specific predictions for what happens at large scales. If the actual universe deviates from those predictions then GR is wrong. There is an industry that in f(R) models in which you invent a theory of gravity that looks like GR at small scales, but is different at large scales. The reason you need things to agree with GR at small scales is that we have a lot of small scale experiments that puts limits on how far you can move from GR.

Are you serious? Atoms are currently easy to observe. Can't you make the distinction between indirect and direct observation?

And black holes are not that much more difficult to observe than atoms. Also, I'm not sure what the difference is between direct and indirect observation or how we see black holes less indirectly than atoms.

Show me a picture of a BH, not that something that could also reasonably be a different thing.

I can show you a picture of a black hole. If you insist that it's not really a black hole that's something different, but you can also play that game with atoms. Show me a picture of a atom that can't possibly be something other than an atom.

By definition, you can't directly observe a black hole.

Yes you can for a reasonable definition of "directly observe." You can come up with a silly definition of "directly observe" but it's really hard to come up with a definition by which black holes cannot be directly observed and the cup in front of me can be. Black holes don't directly emit light, but neither does the cup in front of me. I can see the cup in front of me because the cup reflects and distorts light from other sources. Same for black holes.

Black holes are black. So is my winter overcoat, and it's easy to spot something that is black if it happens to be in the middle of a white room.

Yes you can play lots of philosophical word games, but what I'm saying is that black holes are easier to observe than exo-planets or ocean currents and cosmology is no more speculative than botany, chemistry, or oceanography. It's actually a lot *less* speculative than economics or sociology.

 

[twofish-quant]

I'm pretty sure that his idea that these objects were ejected from galaxy cores was pretty nonsensical from the start, because not only was there no model whatsoever for what they actually were within Arp's idea, but they also had spectra of the quasars at that time.

I'd argue that in 1965, it was not as nutty as assuming that quasars were at cosmological distances.

With no known physical method for producing such large redshifts except by either large recession velocities or gravitational redshift, Arp's idea that these redshifts were "intrinsic" was nutty from the start.

However it was known from the start that if quasars were at cosmological redshifts, then it required an energy source that was vastly more efficient than nuclear fusion. Nuclear fusion only convert a few percent of mass energy into energy, whereas quasars require 50% of mc^2 to get converted into energy.

One thing that is cool is that you read the textbooks from 1965, and you have supporters of distant quasars coming up with very weird and unconvincing models of how quasars are so energetic. The two ideas that people were using was that somehow there was antimatter in quasars or chain reactions of supernova. Both these ideas have serious problems.

It wasn't until the 1970's that people figured out that you could get the energy from dropping matter into a black hole, and that nicely explained why there were no nearly quasars. Black hole eats up all the gas. No gas. No quasar.

It's fun to read old textbooks, and watch people try to figure out a mystery.

 

[twofish-quant]

Wow, when you have to recur to such ofensive expressions as asinine and "you've lost your mind" I might be on the right track.

That's a heuristic, but often not a good one.

Anyway I was developing what Twofish-quant posted:"If it turns out that we've messed up something, then LCDM is out the window. Before 1998, CDM was the model. People added L when CDM wouldn't work. I wouldn't be terribly surprised if we start talking about xLCDM."
is his characteization also asinine and is he out of his mind or is it just me?

The reason I wouldn't be surprised is that while we do have very good data for the rate of universe expansion at recent eras, we are still in the process of getting that information for early eras. That information is rapidly coming in from WMAP. Once we have expansion rates for the early universe, we can see if it fits into a cosmological constant scenario. It wouldn't surprise me if we found that it doesn't and we have to add another parameter to characterize the expansion. On the other hand, it wouldn't surprise me if we didn't.

However that's a very different thing from saying that anything fits.

 

[Chalnoth]

I'd argue that in 1965, it was not as nutty as assuming that quasars were at cosmological distances.



However it was known from the start that if quasars were at cosmological redshifts, then it required an energy source that was vastly more efficient than nuclear fusion. Nuclear fusion only convert a few percent of mass energy into energy, whereas quasars require 50% of mc^2 to get converted into energy.

One thing that is cool is that you read the textbooks from 1965, and you have supporters of distant quasars coming up with very weird and unconvincing models of how quasars are so energetic. The two ideas that people were using was that somehow there was antimatter in quasars or chain reactions of supernova. Both these ideas have serious problems.

It wasn't until the 1970's that people figured out that you could get the energy from dropping matter into a black hole, and that nicely explained why there were no nearly quasars. Black hole eats up all the gas. No gas. No quasar.

It's fun to read old textbooks, and watch people try to figure out a mystery.

Right, but my point was that the spectra of quasars that they had starting in 1962 were extremely strong evidence that they were at cosmological redshifts. These other ideas were pretty crazy, but they definitely aren't as obviously wrong as Arp's idea.

 

[twofish-quant]

And that sentence in no way contradicts what you claim here, other models just don't have the plasticity or malleability to adapt to observations like supernovae Ia light curves.

Which is why they are considered to be wrong.

Where Chalnoth and I very strongly disagree with you is the idea that LCDM can fit *anything*. It can't. If the supernova Ia light curves were something very different from what we observed, then you couldn't tweak LCDM to fit it. The expansion rates, deuterium abundances, helium abundances, CMB spectrum, galaxy correlation functions, yadda, yadda, yadda are such that we can tweak LCDM to match the data. If we observed something *very* different, then LCDM wouldn't work.

There are things that LCDM doesn't quite fit, but right now those things are not big enough to suggest that we got something basically wrong.

I don't know how that can be insulting, unless it is insulting only when I say it, I'm precisely highlighting that property of the concordance model, its ability to survive any observation.

Except it doesn't have that property. There are thousands of things that we *could* have observed that would have killed the LCDM. If you find a star with less than 20% helium, then BBN will not work. If the galactic correlation spectrum is something different that what we see, then LCDM won't work. I remember when the first COBE results came out and showed that the CMB was this perfect isotropic blackbody. I remember it because we were talking about the results, and someone pointed out that if we wait another three months and those results *still* show a perfect isotropic blackbody then it means that we got something very seriously wrong, and then a week later COBE released something showing variations in the CMB.

We *could* have observed a thousand things that could have killed LCDM. It just so happens that we didn't.

Two points...

1) One point is that cosmology is data driven. It's not any different from botany or oceanography. Observing the early universe and trying to make sense of the data isn't that different from observing ocean currents or orchids in the Amazon rain forest (and we have better maps of the early universe than we do some parts of the Amazon rain forest).

2) The other point is that theorists really get disappointed when everything matches. It's no fun when there are no mysteries. I remember when a high energy physics theorist said that he was jealous of the cosmologist because there were so many interesting problems in cosmology when in HEP everything matches the standard model. Also one reason I got into supernova theory, is that no one knew exactly how core collapse supernova work, and we still really don't.

 

[twofish-quant]

Right, but my point was that the spectra of quasars that they had starting in 1962 were extremely strong evidence that they were at cosmological redshifts.

At which point you had a major problem with energy generation. Now if you assume that quasars are nearby then you also have problems, but as of 1965, it wasn't clear what was nuttier. By 1970, people had observed things that changed the debate, namely CMB and pulsars, and by 1975, people figured out quasar evolution.

You can imagine an alternative universe when by 1970 we'd discovered an "megastar" which is something that produces high enough gravitational fields that causes major gravitational redshift when generating hydrogen and it turns out if you look at the CMB closely it consisted of point sources from ultra z=3000 galaxies that through GR effects caused ultra-redshifting which prevents Olber's paradox from happening.

You can imagine that, but that's not what we ended up seeing.

Also I saw this sort of something similar first hand in the early-1990's when you had the same debate over gamma ray bursters. If you assume that gamma ray bursters are at cosmological distances, then you have problems with energy generation, but it turns out that you can come up with plausible energy generation mechanism and beaming, and I knew people that were involved in that effort.

Now again you could imagine this alternative universe when people tried for a few years to come up with GRB energy generation methods and then found out that nothing worked, and then you take more observations of GRB and find that their position is coorelated with either the ecliptic or the galactic plane.

Again, you can imagine this, but that's not what we ended up seeing.

These other ideas were pretty crazy, but they definitely aren't as obviously wrong as Arp's idea.

Obviously wrong in 1975. Not in 1965. All of the ideas for energy generation in 1965 turned out to be wrong.

One other thing is that in 1965, the big bang still hadn't won (people hadn't seen CMB). One argument against cosmological quasars that would have worked in 1965 is that if you accept cosmological quasars then you have to accept the idea that the universe as a whole was evolving, which gets you problems like "what happens before the BB" that you don't have to answer if you assume the universe was in steady-state.

It's a perfectly good philosophical argument that would likely win if cosmology were about philosophy rather than observation.

But it isn't.

One reason I like science is that you *can't* figure things out from pure thought. At some point you just have to look into the telescope and how you believe how the universe works is based on what you see,

 

[twofish-quant]

Also in addition to red shifts there is a lot of other pretty direct evidence now that quasars are at cosmological distances:

1) you see gravitational lensing of quasars,

2) you see things like the Lyman-alpha forest or the Gunn-Peterson trough. What happens is that if you see a quasar at redshift of say z=1, you'll see a whole bunch of absorption lines from hydrogen clouds at redshifts between z=0 and 1. If the quasar is sufficiently high redshift, then you see a general depression from the un-ionized early universe.

These are thing that you see everyday.

 

[TrickyDicky]

However that's a very different thing from saying that anything fits.

Which is why they are considered to be wrong.

Where Chalnoth and I very strongly disagree with you is the idea that LCDM can fit *anything*. It can't. If the supernova Ia light curves were something very different from what we observed, then you couldn't tweak LCDM to fit it. The expansion rates, deuterium abundances, helium abundances, CMB spectrum, galaxy correlation functions, yadda, yadda, yadda are such that we can tweak LCDM to match the data. If we observed something *very* different, then LCDM wouldn't work.

There are things that LCDM doesn't quite fit, but right now those things are not big enough to suggest that we got something basically wrong.
Except it doesn't have that property. There are thousands of things that we *could* have observed that would have killed the LCDM. If you find a star with less than 20% helium, then BBN will not work. If the galactic correlation spectrum is something different that what we see, then LCDM won't work. I remember when the first COBE results came out and showed that the CMB was this perfect isotropic blackbody. I remember it because we were talking about the results, and someone pointed out that if we wait another three months and those results *still* show a perfect isotropic blackbody then it means that we got something very seriously wrong, and then a week later COBE released something showing variations in the CMB.

We *could* have observed a thousand things that could have killed LCDM. It just so happens that we didn't.

Ok, I see where you and Chalnoth are confused about what I wrote:
I was not claiming LCDM can fit anything, since I was not referring to LCDM specifically but to the way of constructing models by adding parameters. When you said that if we find a new observation that the LCDM couldn't fit you wouldn't be terribly surprised if we started talking about xLCDM, that is defining a mechanic, a model producing process, so that when we'd find another observation that would make xLCDM not work we could add a new parameter so that we'd start talking about $xLCDM.
Can't you see the pattern?, it's this pattern that has the plasticity, but you should be able this is the pattern of the ptolemaic epicycles. Nothing wrong with this pattern, in a way it works, it adapts to empirical observations, but eventually it comes to a point where the number of fudge factors is absurdly high, then you have two options: either you come up with a different model that explains observations without the fudge factors,and that probably make you reconsider something you thought it was obvious, or if you don't find the alternative you keep the model with the alphabet soup but with a suspicious rat smell you can't get rid of.

 

[twofish-quant]

The other thing about universe expansion is that if you start with the premise that

1) there are no preferred directions or location in the universe
2) GR is more or less right

There aren't that many different ways that the universe can expand.

 

[TrickyDicky]

The other thing about universe expansion is that if you start with the premise that

1) there are no preferred directions or location in the universe
2) GR is more or less right

There aren't that many different ways that the universe can expand.

Does this belong to this thread?

 

[twofish-quant]

When you said that if we find a new observation that the LCDM couldn't fit you wouldn't be terribly surprised if we started talking about xLCDM, that is defining a mechanic, a model producing process, so that when we'd find another observation that would make xLCDM not work we could add a new parameter so that we'd start talking about $xLCDM.

That's not what I'm saying. Again it's all about observation, and you have to go into the details of the observations.

Right now we have very good expansion curve observations for relatively late periods, but we don't have that sort of information about the early universe. It so happens given our knowledge about the early universe, we *could* parameterize everything with a single number for lambda and still hit observational constraints. It may turn out that we can't once the observations are better, at which point we have to put in a function for lambda.

Can't you see the pattern?, it's this pattern that has the plasticity, but you should be able this is the pattern of the ptolemaic epicycles. Nothing wrong with this pattern, in a way it works, it adapts to empirical observations, but eventually it comes to a point where the number of fudge factors is absurdly high, then you have two options: either you come up with a different model that explains observations without the fudge factors,and that probably make you reconsider something you thought it was obvious, or if you don't find the alternative you keep the model with the alphabet soup but with a suspicious rat smell you can't get rid of.

Sure but reality is messy. However...

1) If you can figure out something with fewer fudge factors that explains the observations, then people will go gaga over this. The problem is that people can't. It's not that people aren't looking or questioning models. The problem is that if you reduce the number of parameters, you just don't fit the data.

Also once you have something that works with a number of fudge factors, you don't give up. At that point you look at each of the numbers that you put in and ask why that number is that number.

2) The number of parameters in cosmological models really isn't that large. Twelve or so numbers and you've explained the universe. Think of the number of parameters you need to model the aerodynamics of a car or ocean currents. You typical solar model has dozens of parameters.

But these parameters aren't just random numbers. They *mean* something. If you have to set lambda to something to make everything work, then you scratch your head and think about what lambda is. And you can't set those numbers to anything. If you have to set the baryon mass of the universe to -1 to get it to work, then you have a problem.

3) Talking about Ptolemy is interesting because if you look at any model of the solar system, what you end up it quite more complex than anything Ptolemy every came up with, and it turns out that any model of the solar system has a lot of parameters that you have to set to make everything work (and curiously a solar system model has more parameters than a cosmology model).

It turns out that those aren't *free* parameters, but they correspond to physically relevant qualities (namely the mass of all of the planets). Same with LCDM. People aren't putting random numbers into the models. Each of the numbers *means* something, and then you can figure out the implications of the parameters being what they are.

It's not a matter of not looking. If you can get rid of a parameter, great! It's a matter of getting things to work. It's not that people aren't very actively looking for alternatives.

Also more data makes the problem worse. Our solar system models are a lot more complicated than anything that Ptolemy came up but we can predict the location of the planets to within 1 meter and within fractions of seconds, and that's nothing that he was able to do.

Ultimately you have to deal with the data. If you can come up with a simple model to deal with the data, then GREAT! But if you have to come up with something really messy and complex to deal with the data because you can't come up with anything better, than that's the just the universe works. And if the universe turns out to be more complicated than you like, that's something you have to take up with God since I can't help you with that.

 

[Chalnoth]

Can't you see the pattern?, it's this pattern that has the plasticity, but you should be able this is the pattern of the ptolemaic epicycles.

This is just wrong, wrong, wrong.

For quite a while now, basically every year we have doubled the amount of astronomical/cosmological data we have obtained. This means that every year we obtain as much new information as in the rest of human history combined. And despite this, the LCDM cosmology remains the most accurate fit to the data.

Since the cosmological constant was suspected by theorists in the early 80's, we have layered on a wide swath of new sorts of observations that all support a cosmological constant, from supernovae to CMB to baryon acoustic oscillations to cluster counts to weak lensing. We have gone, in the last 15 years, from not being able to detect the accelerated expansion at all to nailing down the time variation of dark energy to within a few percent.

One way to see this is to take a simple extension of LCDM, where we allow the pressure of the dark energy to be a constant times its energy density. For a pure cosmological constant, this constant w = -1. The latest combination of WMAP, supernovae, and baryon acoustic oscillations constrains w = -0.999 +/- 0.056 (http://lambda.gsfc.nasa.gov/product/map/current/params/owcdm_sz_lens_wmap7_bao_snconst.cfm ), so it's a cosmological constant to within 5%, and future experiments will nail that down even further.

 

[TrickyDicky]

This is just wrong, wrong, wrong.

For quite a while now, basically every year we have doubled the amount of astronomical/cosmological data we have obtained. This means that every year we obtain as much new information as in the rest of human history combined. And despite this, the LCDM cosmology remains the most accurate fit to the data.

Since the cosmological constant was suspected by theorists in the early 80's, we have layered on a wide swath of new sorts of observations that all support a cosmological constant, from supernovae to CMB to baryon acoustic oscillations to cluster counts to weak lensing. We have gone, in the last 15 years, from not being able to detect the accelerated expansion at all to nailing down the time variation of dark energy to within a few percent.

One way to see this is to take a simple extension of LCDM, where we allow the pressure of the dark energy to be a constant times its energy density. For a pure cosmological constant, this constant w = -1. The latest combination of WMAP, supernovae, and baryon acoustic oscillations constrains w = -0.999 +/- 0.056 (http://lambda.gsfc.nasa.gov/product/map/current/params/owcdm_sz_lens_wmap7_bao_snconst.cfm ), so it's a cosmological constant to within 5%, and future experiments will nail that down even further.

You are wrong, wrong ,wrong. I'm talking about the pattern, not the specific observations, do you know what a pattern is? It does not depend on the specifics.
You keep answering you prejudiced notions of what you think I'm saying, would you try for a change and respond to what I actually say?

 

[twofish-quant]

For quite a while now, basically every year we have doubled the amount of astronomical/cosmological data we have obtained. This means that every year we obtain as much new information as in the rest of human history combined. And despite this, the LCDM cosmology remains the most accurate fit to the data.

And you can do it with 12 or so parameters. That's pretty damn good as scientific models go.

Standard model of particle physics has about 20. Solar system models have about 20. Solar models and supernova have about 30 parameters. If you put in climatic models and ocean current models, you end up with tons of parameters.

 

[TrickyDicky]

And you can do it with 12 or so parameters. That's pretty damn good as scientific models go.

Standard model of particle physics has about 20. Solar system models have about 20. Solar models and supernova have about 30 parameters. If you put in climatic models and ocean current models, you end up with tons of parameters.

Once again I'm talking about the pattern of ever adding new parameters, not about the specific number of parameters a specific model uses.
Do you agree that in general a good property for a model is that it explains observations with the minimum number of arbitrary parameters?

 

[twofish-quant]

You are wrong, wrong ,wrong. I'm talking about the pattern, not the specific observations, do you know what a pattern is? It does not depend on the specifics.

I don't see how we can have this conversation without talking about specific observations. Cosmology and science is all about specific observations. If the observations were different, then things would be different. If we could explain the universe with simple models, then we use simple models. If you can't use simple models to explain the universe then we use complex models. If it turns out that the universe is not modelable at all, well then life stinks.

But all this depends on what people actually observe and how people respond to what people actually see, and what I'm telling you is that you *can't* have this conversation without reference to specific observations.

You keep answering you prejudiced notions of what you think I'm saying, would you try for a change and respond to what I actually say?

The problem is that I think you are arguing that astrophysicists have philosophical beliefs that they don't actually have, which causes problems because you are arguing with astrophysicists that don't believe what you think we believe.

You have said that cosmology is speculative, and I'm trying to tell you that it's as data-driven as oceanography. What cosmologists do in modelling the universe is no different than what oceanographers do in modelling ocean currents or petroleum geologists do in modelling oil wells.

I think you are saying that I think X when in fact I don't think X.

 

[TrickyDicky]

Cosmology and science is all about specific observations. If the observations were different, then things would be different.

Sure, but not only about observations, you need models to deal with the observations, and that is all about patterns, ever heard about how important patterns are in physics and math?

 

[Chalnoth]

Once again I'm talking about the pattern of ever adding new parameters, not about the specific number of parameters a specific model uses.

Then you're just talking about your own personal feelings that have no relevance whatsoever to reality. I show how specific observations could have required extra parameters (specifically, a parameter allowing dark energy to vary), and despite massive improvements in experimental accuracy, from not being able to detect the acceleration at all to being able to nail down its variation to within 5%, there fails to be any evidence that such extra parameters are needed.

This is pretty much as strong a proof as you can get that your gut feeling of a pattern has no relationship whatsoever to reality.

Do you agree that in general a good property for a model is that it explains observations with the minimum number of arbitrary parameters?

Which is precisely why the cosmological constant is the most likely explanation for the accelerated expansion. It has the minimum number of parameters required to explain the observations: one.

 

[twofish-quant]

Do you agree that in general a good property for a model is that it explains observations with the minimum number of arbitrary parameters?

Sure, but I don't see what this has to do with LCDM, since none of the parameters in the model are arbitrary.

If you say that you need X amount of dark energy to make the model work, then you have to go out and start looking for that dark energy, and that X amount of dark energy is going to affect a thousand other things. If you find that a reason that you can't have more than 0.5 X amount of dark energy in the universe, then you have a puzzle.

This actually makes it different from the standard model of particle physics where you do have parameters that are arbitrary. It's also different from models of the sun and stellar evolution where you have parameters like convection and mass loss that aren't constrained by fundamental physics. Anytime you have convection and/or magnetic fields, then you end up having to but in messy fudge factors, but the processes that force you to put in arbitrary fudge factors don't seem to be present at cosmological scales.

As scientific models go, LCDM is a pretty "clean" model.

Again we can go into deep philosophical discussions about the nature of scientific modelling, but I don't see why you are picking on LCDM, when models of say hurricane behavior have a lot more arbitrary parameters than LCDM has.

 

[twofish-quant]

Sure, but not only about observations, you need models to deal with the observations, and that is all about patterns, ever heard about how important patterns are in physics and math?

Yes, from popular science books that get it wrong...

Part of the reason that I get into these conversations is that there are a lot of popular misconceptions about how astrophysicists actually think.

 

[TrickyDicky]

Then you're just talking about your own personal feelings that have no relevance whatsoever to reality.

This is pretty much as strong a proof as you can get that your gut feeling of a pattern has no relationship whatsoever to reality.

So far, you're the only person who's gotten emotional in this thread, maybe it's your "gut feeling" that needs revision.

Yes, from popular science books that get it wrong...

I see, so it is wrong, I think you are the only one that holds that . Curious.

Again we can go into deep philosophical discussions about the nature of scientific modelling, but I don't see why you are picking on LCDM, when models of say hurricane behavior have a lot more arbitrary parameters than LCDM has.

I'm not picking on LCDM, I was drawing the inevitable pattern from something you said.
If you disagree now with what you said is your problem.

 

[Chalnoth]

So far, you're the only person who's gotten emotional in this thread, maybe it's your "gut feeling" that needs revision.

You are the one that brushed away hard data based upon your own gut feeling.

 

[TrickyDicky]

You are the one that brushed away hard data based upon your own gut feeling.

This is not about hard data and i can assure you I don't have any feeling about this, I derived a pattern from a proposition put forth by twofish-quant. It was a logic exercise, nothing to do with data. Maybe eventually you'll get it

 

[Chalnoth]

It was a logic exercise, nothing to do with data.

You can pretend all you want, but ultimately this has everything to do with the data. Heck, even your original (asinine) claim was very much data-based.

You see, one of the important things to learn about patterns is that most of the time, they are spurious. This becomes blindingly obvious when you start to investigate any science in a noticeable amount of detail. We humans are really, really good pattern-finding machines. Too good, in fact, very often seeing patterns where none exist.

And if you want to know whether something which you think is a pattern really is, you have to delve into it in detail. We've been showing that when you do this, there simply isn't anything at all to this "pattern" you claim to see.