Arp objects, QSOs, Statistics

[DeiRenDopa]

(part omitted)

not once, in BAC's posts before this (and, I think, to date), has "the mainstream's theories" been stated or referenced

ABSOLUTELY FALSE AGAIN. I've stated (and referenced) all the portions of the mainstream theory needed to do my calculations. Do you really want to deny that mainstream theorists claim quasars are continuously distributed (as opposed to quantized) with regards to distance and therefore redshift? Do you really want to deny that the mainstream says quasars are randomly located in the sky with respect to our viewing location?

(rest omitted)

.
Hmm, ok ...

Would you please remind me, by giving the post numbers, where you mentioned the following? I think you'd agree that they are all essential parts of the theory, or theories, you claim your hypothesis sets out to test.

* gravitational lensing, both weak and strong

* clustering; more generally, large-scale structure (groups, clusters, filaments, super-clusters, ...)

* definition(s) of 'quasar'.

In each case, would you please indicate the extent to which you have incorporated each of these - quantitatively - into the statement of the hypothesis you claim to be testing, and what the null hypothesis is?

 

[DeiRenDopa]

(part omitted)

As I noted earlier, this becomes quite interesting, if not necessary important, when considering:

* how to reconcile the L-C&G paper (no statistically significant minor axis anisotropies in the magnitude range spanned by Chu et al., no statistically significant minor axis anisotropies within 1o) with the quasar data in the Chu et al. paper

What is interesting is how you continue to misrepresent what the L-C&G paper concluded. Here's more of what the paper stated: "In the previous subsection, we showed that, given the actual distribution of SDSS QSOs, a random distribution of position angles in the RC3 galaxies should not show the level of anisotropy found; that is, the anisotropy is not due to the peculiarities of the QSO distribution itself. ... snip ... This gives 135 galaxies (instead of 71 galaxies with total coverage) ... snip ... whose significance according to Monte Carlo simulations is 4.4-sigma (a probability of 1 in 90,000). For z > 0.5 the significance according to Monte Carlo simulations is 4.8-sigma (a probability of 1 in 600,000)."

And the only place in the paper where there is a discussion involving an angular distance (I'll substitute THETA for the Greek character in the quote below) of 1 degree or less is here:

"Other values of THETAmax (see Fig. 3) show the existence of anisotropy but the best significance is for THETAmax ?3 degrees. For lower angles, there seems to be a lower value of ALPHA (BAC - anisotrophy), although for THETAmax < 1 degree the number of QSOs is too low to draw any conclusions. "

Did you miss that qualification regarding the data and conclusions? It states very clearly that there are too few data points in that range to justify any conclusions ... in other words, YOUR conclusions. Is this your normal scientific approach, DRD ... to ignore the warnings of the authors of papers?

(rest omitted)

.
We have already discussed this, much earlier in this thread.

For example, in post#97 (some formatting lost, and the links):

M. López-Corredoira, C. M. Gutiérrez, 2006 ("L-C&G") and Arp, H. C., 1999b, A&A 341, L5 ("A QSO 2.4 arcsec from a dwarf galaxy - the rest of the story", Arp (1999b)):

Arp (1999b) finds that the probability of having six out of six QSOs aligned within ±15o of the minor axis of NGC 5985 to be only 10−8 to 10−9.

.
Arp's six "QSOs" range in (V) magnitude from 16.4 to 19.0, and are at distances from 12' to 90', with all but one within 1o.

L-C&G find that no minor axis anisotropy for SDSS quasars with (g) mag <19.2 (I added some bolding):

Examining Fig. 6, one should also realize that there are two structures in the counts: two overlapping peaks, one with a maximum at mg ≈ 19.2 and another at mg ≈ 20.2. Apparently, it is the second group of QSOs that is responsible for the anisotropy, and this is shown over mg > 19.4 because this is the range where the number of QSOs in the second group is relatively significant.

.
Nor is there any significant anisotropy at angular separations <~1o.

Of course, V and g magnitude scales are not the same, but quasars/QSOs with these redshifts don't have extreme (V-g) colours, so there's no wiggle room there.

How to intrepret "the probability of having six out of six QSOs aligned within ±15o of the minor axis [... is] only 10−8 to 10−9" in light of the much larger study by L-C&G?

Care to comment, BeAChooser?

.

Re 'within 1^o': if Arp finds an anisotropy of extraordinary statistical significance with just one galaxy and just six quasars (or two and eleven, or three and twelve, or ...), how come L-C&G can say "the number of QSOs is too low to draw any conclusions", even though they examined ~70 galaxies and thousands (8,698, according to L-C&G) of quasars?

How can a very small sample produce such a strong signal yet a much larger one, designed explicitly to test the same idea, produce a statistically non-significant result?

How can 5/11/12 quasars be more than enough to draw strong conclusions, yet thousands of quasars be too few to draw any conclusions?

By the way, NGC 5985 has an estimated distance (~35 Mpc) that is nicely in the middle part of range spanned by the 71 galaxies L-C&G chose (~10 to ~140 Mpc).

 

[DeiRenDopa]

(part omitted)

As I noted earlier, this becomes quite interesting, if not necessary important, when considering:

* how to reconcile the L-C&G paper (no statistically significant minor axis anisotropies in the magnitude range spanned by Chu et al., no statistically significant minor axis anisotropies within 1o) with the quasar data in the Chu et al. paper

What is interesting is how you continue to misrepresent what the L-C&G paper concluded. Here's more of what the paper stated: "In the previous subsection, we showed that, given the actual distribution of SDSS QSOs, a random distribution of position angles in the RC3 galaxies should not show the level of anisotropy found; that is, the anisotropy is not due to the peculiarities of the QSO distribution itself. ... snip ... This gives 135 galaxies (instead of 71 galaxies with total coverage) ... snip ... whose significance according to Monte Carlo simulations is 4.4-sigma (a probability of 1 in 90,000). For z > 0.5 the significance according to Monte Carlo simulations is 4.8-sigma (a probability of 1 in 600,000)."

And the only place in the paper where there is a discussion involving an angular distance (I'll substitute THETA for the Greek character in the quote below) of 1 degree or less is here:

"Other values of THETAmax (see Fig. 3) show the existence of anisotropy but the best significance is for THETAmax ?3 degrees. For lower angles, there seems to be a lower value of ALPHA (BAC - anisotrophy), although for THETAmax < 1 degree the number of QSOs is too low to draw any conclusions. "

Did you miss that qualification regarding the data and conclusions? It states very clearly that there are too few data points in that range to justify any conclusions ... in other words, YOUR conclusions. Is this your normal scientific approach, DRD ... to ignore the warnings of the authors of papers?

(rest omitted)

.
We have already discussed this, much earlier in this thread.

For example, in post#97 (some formatting lost, and the links):

M. López-Corredoira, C. M. Gutiérrez, 2006 ("L-C&G") and Arp, H. C., 1999b, A&A 341, L5 ("A QSO 2.4 arcsec from a dwarf galaxy - the rest of the story", Arp (1999b)):

Arp (1999b) finds that the probability of having six out of six QSOs aligned within ±15o of the minor axis of NGC 5985 to be only 10−8 to 10−9.

.
Arp's six "QSOs" range in (V) magnitude from 16.4 to 19.0, and are at distances from 12' to 90', with all but one within 1o.

L-C&G find that no minor axis anisotropy for SDSS quasars with (g) mag <19.2 (I added some bolding):

Examining Fig. 6, one should also realize that there are two structures in the counts: two overlapping peaks, one with a maximum at mg ≈ 19.2 and another at mg ≈ 20.2. Apparently, it is the second group of QSOs that is responsible for the anisotropy, and this is shown over mg > 19.4 because this is the range where the number of QSOs in the second group is relatively significant.

.
Nor is there any significant anisotropy at angular separations <~1o.

Of course, V and g magnitude scales are not the same, but quasars/QSOs with these redshifts don't have extreme (V-g) colours, so there's no wiggle room there.

How to intrepret "the probability of having six out of six QSOs aligned within ±15o of the minor axis [... is] only 10−8 to 10−9" in light of the much larger study by L-C&G?

Care to comment, BeAChooser?

.

Re 'within 1^o': if Arp finds an anisotropy of extraordinary statistical significance with just one galaxy and just six quasars (or two and eleven, or three and twelve, or ...), how come L-C&G can say "the number of QSOs is too low to draw any conclusions", even though they examined ~70 galaxies and thousands (8,698, according to L-C&G) of quasars?

How can a very small sample produce such a strong signal yet a much larger one, designed explicitly to test the same idea, produce a statistically non-significant result?

How can 5/11/12 quasars be more than enough to draw strong conclusions, yet thousands of quasars be too few to draw any conclusions?

By the way, NGC 5985 has an estimated distance (~35 Mpc) that is nicely in the middle part of range spanned by the 71 galaxies L-C&G chose (~10 to ~140 Mpc).

 

[DeiRenDopa]

(part omitted)

* how to treat quasars in the environs of NGC 3516 which are/will be discovered after publication of the Chu et al. paper.

Again, this is a red herring. You haven't proven that any quasars have been discovered in the environs of NGC 3516 since the Chu paper. And given that Chu et al were specifically looking for quasars in that environ, it's rather doubtful that many, if any, have been found unless they just missed the obvious. And I doubt they did. Furthermore, you presume any new quasars would not also be aligned with respect to significant galaxy features or have redshifts close to Karlsson values. Check back in when you can actually provide new data on quasars near NGC 3516 that contradict the conclusion of the study. The ball is in your court.

(rest omitted)

.
It's only a red herring if, in your view of how the sciences of astronomy, astrophysics, and cosmology should be done models need not be predictive, and hypotheses cannot be generalised or extended to potentially address new data (and so on).

I somewhat doubt that that's what you intend, in terms of doing science, though it might be what you intend with regard to the specific set of calculations you posted, for the 'NGC 3516 Chu et al. quasars'; would you clarify please?

You may find it worthwhile to re-read the Chu et al. paper; they explicitly state that they set out to take spectra of only the five BSOs Arp identified.

I do not presume anything, re any quasars that may have been found in the environs of NGC 3516 since publication of Chu et al., or which may be found in future.

But the bottom line is, I think, that it is a hypothesis (or two) of yours that we are discussing, so the ball remains firmly in your court .... unless, perhaps, you are hinting at yet another difference between how you think astronomy, as a science, should be done and how it is actually done?

 

[DeiRenDopa]

(part omitted)

Now if he had used this approach to test some hypothesis about Karlsson peaks, or minor axis anisotropies, the approach might have some value ...

Tell us, DRD ... since clearly the alignment in NGC 3516 was discovered AFTER Karlsson identified his specific quantization values, wouldn't you admit that my calculation tests the hypothesis that Karlsson was right and not the mainstream which claims redshifts are not quantized? Or is the logic of that beyond you?

.
Hey BAC, it's your hypothesis ... if you want to re-state it in terms of a test of Karlsson peaks (or whatever), feel free to do so.

If/when you do, then we can discuss what conclusions you draw from it.

But don't forget that for everyone reading this thread, this is completely new territory: I (and I trust other readers) acknowledge that you are presenting only your own opinion (so we cannot take anything else as substantive), and I think we are all agreed that the method you have used, in the calculations you presented, is not legitimate within astronomy (etc) as a science as it done by professionals (so you are presenting 'BAC's scientific methods for use in astronomy').
.

If the hypothesis being tested is solely concerned with 'mainstream theories', then Karlsson peaks are irrelevant ... any 3/5/10 numbers in [0,3] will do.

ABSOLUTELY FALSE AGAIN. You have quite a string of those going, DRD. You will not get the same probabilities if you use any 3/5/10 numbers. Let me demonstrate for NGC 3516.

Instead of the Karlsson values of z = 0.3, 0.6, 0.96, 1.41, 1.96, 2.64, let's use z = 0.10, 0.48, 0.75, 1.20, 2.25, 2.80 as the values (call them DRD values). Note that I even tried to keep them around the actual data points (helping you) rather than picking *any* five.

The quasars have z = 0.33, 0.69, 0.93, 1.40, 2.10. Therefore the spacings to each of the nearest DRD values are -0.15, -0.16, +0.18, +0.20, -0.15. Now double those to find the increment to use in the calculation. Let's just assume they are all 0.15 (helping you again, by the way) so the increment will be 0.30. Between 0 and 3.0 there are 10 zones of that increment. So the probability of those 5 quasars showing up close to the DRD values is 1/(10*9*8*7*6)/(5*4*3*2*1) = 0.004. That compares to a probability of 0.0000003 in my calculation using Karlsson's values. That's not even REMOTELY close. So you clearly don't know what you are talking about when it comes to discussing probability and statistics. You're just blowing smoke and have been since the beginning of this thread.

(rest omitted)

.
Of course you don't get the same 'probabilities' if you use different inputs!

My point was, and still is, that if the hypothesis that you claim to be testing, with the calculations you present, is intended to be one of 'mainstream theories', then there's nothing special about Karlsson peaks (you can use any set of five numbers to do the test).

So, in a sense you are quite correct, I do not understand this method you are using.

Perhaps an example might help.

In the 30' field around NGC 4030, there are 23 'QSOs' listed in NED; here are their redshifts, in order of increasing z:
0.132, 0.538, 0.735, 1.005, 1.011, 1.151, 1.211, 1.234, 1.257, 1.275, 1.326, 1.499, 1.567, 1.675, 1.812, 1.85, 1.892, 1.945, 2.063, 2.093, 2.2, 2.568, 3.592. (note that all but two have SDSS designations; presumably only two were known before the field was observed by SDSS).

Now NGC 4030 is not a Seyfert (as far as I know), so doing a 'BAC analysis' may not be relevant (to your hypothesis); however, could you please walk us through how you would determine 'probability' for these quasars around this galaxy, as a test of your hypothesis?

And how relevant would it be to repeat this, using quasars within 30' of a randomly chosen Seyfert (excluding the ones we are already discussing)?

 

[DeiRenDopa]

(part omitted)

The Chu et al. paper carefully explains how they chose which objects to observe (in order to measure redshifts); the "average density" of quasars you need to use in this part of your calculation is that which would be obtained if the search method used in Chu et al. were to be used over the whole sky.

Curious. I STILL don't see you actually providing a number.

.
Um, er, ... BeAChooser, it's your hypothesis we are discussing (not mine).

I am merely pointing out that, in order to be correct, the numbers you use in your calculation need to be consistent with how the data were obtained.

Or are you saying that this kind of consistency is not important, in terms of how you think astronomy (etc) should be done?
.

I don't think you can because you not only don't understand the calculations I did, you don't even understand Chu's paper. We already know you didn't actually read the Chu paper because you claimed it found a result over a domain that the paper itself stated there was too little data to draw any conclusions.

.
Huh?!?

I think you may have meant to write "M. Lopez-Corredoira and C. M. Gutierrez"; if so, then please refer to an earlier post of mine today.
.

You know what you are, DRD? A TROLL ... whose only purpose is to keep this thread going I think. You've demonstrated three (or is it four times) that you didn't actually read the paper you were claiming to know about and falsely claimed supported your view of things. You best study yourself in that troll study you *claim* to be doing, because I'm tired of your dishonesty and probably won't continue "feeding the troll" for much longer.

.
Well, you are, of course, entitled to your own opinion ... and readers of this thread can make up their own minds too.

But if you'd please clarify - by "the paper" ('You've demonstrated three (or is it four times) that you didn't actually read the paper you were claiming to know about and falsely claimed supported your view of things'), are you referring to the L-C&G one ("First tentative detection of anisotropy in the QSO distribution around nearby edge-on spiral galaxies"), or the Chu et al. one ("Quasars around the Seyfert Galaxy NGC 3516")?
.

But I will thank you for motivating me to flesh out certain details of the calculations and lower the probabilities even further than initially.

(rest ignored)

.
I'm glad that you found the discussion fruitful.

It has been quite an eye-opener for me too.

For starters, I hadn't appreciated just how poorly the importance of formulating quantitative, testable hypotheses (including an explicit statement of the null hypothesis) was understood (and still is, apparently).

Second, I hadn't realised how difficult it would be to demonstrate the flaws of the a posterori approach you have used. If you had continued to be involved in discussion, perhaps I could have demonstrated this to your satisfaction, by using random examples from NED, perhaps.

Third, and best of all from my point of view, I think I can see what the fundamental difference is: you seem to have a view of how astronomy (etc) should be done that is both poorly formulated and clearly at odds with the way the professionals go about actually doing astronomy. In this respect, I'm a bit disappointed you're not going to continue discussion; for one thing, I'm quite curious to know how you would go about actually testing any of the ideas, models, or theories you seem to be enamoured with. For example, suppose you were offered lots of time on any one of the leading astronomy facilities - the VLA, the HST, Spitzer, XMM-Newton, INTEGRAL, the Geminis - using any combination of instruments, what would you observe? and how would you expect your observations to advance the science of astronomy (astrophysics, cosmology)?

 

[DeiRenDopa]

Hey BAC, you seem to have missed post #294, where I looked at 'the NGC 5985 case' for the first time!

I know that you have asked, several times, for comments on your calculations, well, in that post I point out that not one, not two, but three (yes, three!! ) of your inputs seem wrong:

* the 0.35 object is a Seyfert, not a quasar
* the '0.59' quasar in fact has a redshift of 0.69
* the 1.90 quasar is missing.

I also pointed out that there seem to be at least five other quasars within 30' of NGC 5985 (in addition to the two mentioned in the Arp paper).

Two of the questions I asked, in that post, are directly relevant to your calculations:

> I wonder how many [of the five 'new' quasars] lie near the minor axis of NGC 5985?

> I wonder what sort of results one would get by applying a BAC-type analysis to these 104 'near NGC 5985' objects (or even just the seven quasars)?

If you're still with us, would you mind having a go at addressing the apparent errors in the inputs, and also replying to at least these two questions?

Thanks in advance.

 

[BeAChooser]

I think the paper in which 'the NGC 5985 case' is covered is A QSO 2.4 arcsec from a dwarf galaxy - the rest of the story, by H. Arp (the link is to the arXiv preprint abstract).
... snip ...

This doesn't align very well with what went into BAC's calculations (quoted here):
* the 0.35 object is a Seyfert, not a quasar
* the '0.59' quasar in fact has a redshift of 0.69
* the 1.90 quasar is missing.

I used was http://articles.adsabs.harvard.edu//full/1999A&A...341L...5A/L000006.000.html , the published version (Astronomy and Astrophysics, v.341, p.L5-L8, 1999) of "A QSO 2.4 arcsec from a dwarf galaxy - the rest of the story". But you are correct that in using values of z = 0.35, 0.59, 0.81, 1.97, 2.13 for the quasars I made several mistakes. First, the 0.59 value was a typo on my part that got carried into the calculation. The correct value is 0.69. Second, I identified the object at z=0.35 as a quasar because the abstract of the paper states "it turns out that a total of five quasars, plus the dwarf galaxy, are accurately aligned along the minor axis of this Seyfert with the quasars in descending order of redshift, i.e., z = 2.13, 1.97, 0.81, 0.69, 0.35." But the body of the paper does indeed say the quasars are at z = 0.69, 0.81, 1.90 (which I did not use in my calculation because it's not aligned with the minor axis), 1.97, 2.13 and that the z = 0.35 case is a Seyfert instead of a quasar.

So how does all this alter my calculation results for this case? Let's see.

I'll now ignore the z= 0.35 Seyfert. That leave just 4 quasars aligned with the minor axis at z = 0.69, 0.81, 1.97 and 2.13. The spacing to the nearest Karlson values are +0.09, -0.15, +0.01 and +0.17. Since there's so much difference between them, I'll just estimate individual probabilties and combine them.

The increment for each is 0.18, 0.30, 0.02, 0.34. The number of increments for each is then 3.0/0.18 , 3.0/0.30, 3.0/0.02 and 3.0/0.34 which equal 16, 10, 150 and 8. The combined probability is then 1/16 * 1/10 * 1/150 * 1/8 = 5 x 10^-6 compared to my previously calculated value of 3 x 10^-8 . That is a factor of 174 higher. Furthermore, now there are only 4 quasars aligned with the minor axis. The probability of that happening is 12 times higher than getting 5 quasars aligned within a 15 degree zone so overall, the new probability for NGC 5985 must be 12 * 174 = 2088 times larger than I calculated earlier. Thus the corrected probability for NGC 5985 is 1.1 x 10^-5 * 2088 = 0.023.

Admittedly that's a lot higher than before, but it's still pretty small considering that it assumes one looks at the entire population of quasars and galaxies ... when in fact Arp et al found this example after looking at a small fraction of them. In the previous estimate I also conservatively ignored the quasar at z = 1.90. But I don't really need to do that and including it would lower the probability by a factor of about 25. But in any case, to compensate for that now higher (but still small) probability I could always also just add another case into the mix.

How about we add NGC 1068?

http://www.journals.uchicago.edu/doi/abs/10.1086/311832 "A Group of Quasi-stellar Objects Closely Associated with NGC 1068, E.*M.*Burbidge, 1999, ... snip ... It is shown that three of the compact X-ray sources detected by ROSAT close to the nearby classical Seyfert galaxy NGC 1068 are quasi-stellar objects (QSOs) with redshifts , 0.385, and 0.655. Also, using previous optical studies, we show that a total of 11 QSOs brighter than mag and with redshifts that range from 0.261 to 2.108 lie within a radius of 50' of NGC 1068. The distribution and very high surface density of these QSOs strongly suggest that they are physically associated with NGC 1068 and were ejected from it."

Every one of the quasar redshifts identified above is close to a Karlsson value. But what about the others that are mentioned but not listed in that source?

http://arxiv.org/abs/astro-ph/0111123 "Further Evidence for Large Intrinsic Redshifts, M. B. Bell, 2001" lists the following as z for the 11 quasars:0.261, 0.468, 0.726, 0.649, 1.054, 1.552, 1.112, 0.385, 2.018, 0.684, 0.655 .

And then I found this new additional quasar we can add:

http://www.sciencedirect.com/scienc...serid=10&md5=596c8badf26d1a60f6786ae0bfcae1d6 "A new quasar identified in the vicinity of NGC1068, Zhu Xing-fen, Zhang Hao-tong and Chu Yao-quan, 2002 ... snip ... An X-ray source close to the classical Seyfert galaxy NGC 1068 is identified as a quasar with a redshift of 0.63. The very high surface density of quasars around NGC 1068 suggests that the quasars may be physically associated with this active galaxy."

So let's do the calculation for this case assuming the same Karlsson peaks as before (z = 0.06, 0.3, 0.6, 0.96, 1.41, 1.96, 2.64) with 12 data points having z = 0.261, 0.385, 0.468, 0.63, 0.649, 0.655, 0.684, 0.726, 1.074, 1.112, 1.552 and 2.018.

The distance to the nearest Karlsson value for each quasar is: -0.039, +0.085, +0.132, +0.03, +0.049, +0.055, +0.084, +0.126, +0.104, +0.152, +0.142, +0.058 .

These distance are the following percentage of the distance to the next nearest Karlsson value: 16%, 28%, 44%, 8%, 14%, 15%, 23%, 35%, 23%, 34%, 26%, 9% . So most of them do seem to be with 30% of a Karlsson value and there's no specific trend that would make assuming an even distribution of quasars between z = 0.0 and 3.0 unconservative.

Double the distances to find suitable calculation increments: 0.078, 0.170, 0.264, 0.06, 0.098, 0.11, 0.168, 0.252, 0.208, 0.304, 0.284, 0.116.

Let's arrange them in order from lowest to highest: 0.06, 0.078, 0.098, 0.11, 0.116, 0.168, 0.170, 0.208, 0.252, 0.264, 0.284, 0.304 .

Let's treat the first 5 as if they have an increment of 0.11 (i.e., 3.0 / 0.11 = 27), the next 3 as if they have an increment of 0.21 (i.e., 14), the next 4 as if they have an increment of 0.30 (i.e, 10). The probability of encountering all these observations near any given galaxy is therefore (5*4*3*2)/(27*26*25*24*23) * (3*2)/(14*13*12) * (4*3*2)/(10*9*8*7) = 1.6 x 10^-10.

And how many galaxies in the entire sky are likely to have 12 quasar within 50' of them? Or even one degree? Remember that the SDSS authors (a supposedly relatively complete survey) estimate there are about 410,000 observable quasars in the entire sky. Let's double that number just in case they missed any to 820,000 quasars. But again, let's halve the number for those that aren't anywhere near a low redshift galaxy (a conservative view and note this galaxy has a very low redshift of z = 0.0038). So we're back to 410,000. Then let's again assume (VERY conservatively) that only half of the remaining are distributed in groups of 12 or less. That leaves 205,000. Dividing that by 12 give us the maximum possible number of cases out there with 12 nearby quasars ... 205,000 / 12 = 17083.

Thus, the probability of finding those specific quasars around NGC 1068 if we were to have looked at every single possible quasar / galaxy association in the sky is 17083 * 1.6 x 10^-10 = 3 x 10^-6 = .000003, which is another VERY low likelihood. I hate to tell you, DRD, but it's becoming more and more obvious that the mainstream theory is missing something vital ... that the mainstream theory about quasars is deeply flawed.

And by the way, in the calculation above, I ignored the alignment of the objects with respect to the galaxy (i.e., minor axis, etc.). But Bell finds in his various papers that the quasars are indeed aligned in a curious and improbable fashion.

http://www.journals.uchicago.edu/cgi-bin/resolve?ApJ54273PDF "On Quasar Distances and Lifetimes in a Local Model, M. B. Bell, 2001 ... snip ... It was shown previously from the redshifts and positions of the compact, high-redshift objects near the Seyfert galaxy NGC 1068 that they appear to have been ejected from the center of the galaxy in four similarly structured triplets."

And not only that, there may be evidence that Karlsson missed a few values in his series which would lower the probability even further:

http://arxiv.org/pdf/astro-ph/0208320 "Evidence that an Intrinsic Redshift Component that is a Harmonic of z=0.062 May be Present in Every Quasar Redshift, M.B. Bell, 2002, After estimating and removing all ejection-related Doppler components from the redshifts of the QSOs near NGC 1068, the remaining redshift is assumed to be intrinsic. This well-studied case is the first example in which it has finally been possible to separate the intrinsic redshift componen tfrom the cosmological and other Doppler components. It is shown that this leads to intrinsic redshift components that occur at exact multiples of z=0.062".

So my conclusions don't really have to change ... do they, DRD. No, perhaps it's time to admit your conclusions need changing.

A quick check using NED turns up 104 (extra-galactic) objects within 30' of NGC 5985; in addition to the two in the Arp paper, there are:
* 41 radio sources (without listed redshifts)
* 30 galaxies (types unspecified, no listed redshifts)
* 4 IR sources (without listed redshifts)
* 2 x-ray sources (without listed redshifts)
* 1 emission line source (without listed redshift)
* 1 galaxy triple (without listed redshift)
* 1 group of galaxies (without listed redshift)
* 10 galaxies with redshifts (types unspecified; z = 0.006, 0.01, 0.01, 0.01, 0.07, 0.09, 0.11, 0.19, 0.19, and 0.25)
* 2 groups of galaxies (z = 0.009, 0.01)
* 1 cluster of galaxies (z = 0.14, determined photometrically)
* 4 UV excess sources (photometric redshifts of 0.98, 1.28, 1.28, and 1.78)
... and 5 QSOs (redshifts of 1.54, 1.78, 1.97, 2.13, and 3.88).

Perhaps there is some unrecognised duplication, a radio source may be the same as a galaxy, for example?

I wonder how many of the objects without redshifts are, in fact, either quasars or S1 or dSp galaxies?

Introducing another red herring? You are perfectly free to find out and tell us the location of each. And then prove that in any way they invalidate my or Arp et. al's conclusions. In fact, given what a nuisance Arp et. al. have been to your community, I'm a little surprised that no one like you has even tried. Could it be that they did look and just couldn't make the case?

I wonder what would happen if we looked right out to 91' from NGC 5985?

Just curious. What's 91' out from NGC 5985? Now don't be coy.

But, above all, I wonder what sort of results one would get by applying a BAC-type analysis to these 104 'near NGC 5985' objects?

But why would one want to apply it to all 104 objects given that only a few of those various type objects are hypothesized as being ejected from that galaxy or any galaxy for that matter? Perhaps you STILL don't understand the nature of the calculations and method?

 

[BeAChooser]

Using your statistics BAC, how can you tell a random placement from a causal one?

David, the fact that can ask this question with a straight face after all that's been posted in the way of explanation on this thread proves you actually know and understand very little about probability and statistics ... that you even lack common sense and the ability to read. And that's why I chose not to engage you further in debate. It's a complete waste of time.

 

[BeAChooser]

BAC, the hypothesis you claim you are testing, per many many posts in this thread, is one concerning "mainstream theories".

Likewise, I grow weary of going around and around in circles with you, DeiRenDopa. More and more, you seem to be trying to emulate David. You don't listen, you keep misrepresenting what I and various sources have said, you don't answer questions asked of you, and you clearly know a lot less about probability and statistics than you claim.

And rather than continue with it in this thread, I'll start a new thread, in the next day or so.

So you run. Well if you do, I will take the opportunity to summarize the results of the calculations on this thread and post them there with a link back to this one so that everyone can see the context of their derivation and your inability to logically and factually defend your "mainstream" theory against those data, calculations and rather straight-forward logic.

 

[BeAChooser]

I think you'd agree that they are all essential parts of the theory, or theories, you claim your hypothesis sets out to test.

* gravitational lensing, both weak and strong

WRONG AGAIN. That's becoming a habit with you, DRD. Gravitational lensing has no effect on the calculations I made. It doesn't affect the value of z of each quasar (i.e., it's apparent quantization) nor would it cause the alignment of the quasar "images" to be along the minor axis of a galaxy (at least you haven't provided any mechanism to explain how it could).

* clustering; more generally, large-scale structure (groups, clusters, filaments, super-clusters, ...)

WRONG. Again, that has nothing to do with any input or assumption in my calculations.

* definition(s) of 'quasar'.

Over and over in my calculations of probability, I've given your side the benefit of the doubt as to the number of quasars (whatever they are) and their distribution around galaxies. I've even asked you to tell us the number I should use. But you refuse, because you sense where that will lead.

 

[BeAChooser]

Originally Posted by BeAChooser
"Other values of THETAmax (see Fig. 3) show the existence of anisotropy but the best significance is for THETAmax ?3 degrees. For lower angles, there seems to be a lower value of ALPHA (BAC - anisotrophy), although for THETAmax < 1 degree the number of QSOs is too low to draw any conclusions. " Did you miss that qualification regarding the data and conclusions? It states very clearly that there are too few data points in that range to justify any conclusions ... in other words, YOUR conclusions.

We have already discussed this, much earlier in this thread.

No, we did not discuss this issue. You never once mentioned nor did I note that the source you were using to make your claim specifically states there is not enough data to draw any conclusions in the range where THETA is less than 1 degree ... where you want to make a conclusion. And beyond THETA equal to 1 degree, that source agrees that there IS an unexplained and significant anisotropy. Your dishonesty in this is apparent.

Re 'within 1o': if Arp finds an anisotropy of extraordinary statistical significance with just one galaxy and just six quasars (or two and eleven, or three and twelve, or ...), how come L-C&G can say "the number of QSOs is too low to draw any conclusions", even though they examined ~70 galaxies and thousands (8,698, according to L-C&G) of quasars?

One more example of you demonstrating quite clearly that you do not understand the nature of these calculations and probability in general, and are dishonest in your arguments. Here's why ...

How can 5/11/12 quasars be more than enough to draw strong conclusions, yet thousands of quasars be too few to draw any conclusions?

One, Arp did not look at just one galaxy in his studies.

Two, L-C&G only looked at alignment to the minor axis. They did not look at redshift quantization.

Three, the calculation of probability is for the entire estimated population of quasars and thus conclusions can be drawn.

Four, you must have misinterpreted L-C&G's paper AGAIN. It's easy enough to check. Suppose each of the galaxies was more than several degrees away from each other so that those 1 degree zones did not overlap. Out to a distance of 1 degree, each galaxy would occupy about 3 square degrees. Since there are 41,250 square degrees in the entire sky, each galaxy would occupy 0.0000727 of the entire sky. So 70 galaxies would occupy 0.0051 of the entire sky. So if they had claimed that there were 8698 quasars within 1 degree of 70 galaxies, then they would have been implying a total number of observable quasars in the sky of 8698/0.0051 = 1,705,000 ... more than 4 times the number that the SDSS study estimates (and it's the best survey yet with a claim that estimate accounts for more than 90% estimate of all observable quasars).

Five, you stated that:

L-C&G find that no minor axis anisotropy for SDSS quasars with (g) mag <19.2 (I added some bolding):

L-C&G's paper:
Examining Fig. 6, one should also realize that there are two structures in the counts: two overlapping peaks, one with a maximum at mg ? 19.2 and another at mg ? 20.2. Apparently, it is the second group of QSOs that is responsible for the anisotropy, and this is shown over mg > 19.4 because this is the range where the number of QSOs in the second group is relatively significant.

Once again we find you making too broad a conclusion. You must have missed the caption to Figure 6 ... which states the plot is for THETA = 3 degree data. Were you trying to apply that to the THETA < 1 degree data? It would appear so. Once again you drew an erroneous conclusion because you misinterpreted a paper and ignored an explicit statement that there's too little data in the THETA < 1 regime to draw any conclusion. Once again we find you throwing out smoke to hide the flaws of your logic and position. Well I certainly hope you don't apply such *careful* attention to detail and logic in your *scientific* professional life, DRD. Or at least I'm thankful you are NOT building or designing anything on which people's lives will actually depend. Astronomers have a rather *ivory tower* job in that respect. Thank God ... or we'd all be depending on gnomes for our lives.

 

[BeAChooser]

It's only a red herring if, in your view of how the sciences of astronomy, astrophysics, and cosmology should be done models need not be predictive, and hypotheses cannot be generalised or extended to potentially address new data (and so on).

Wow! Talk about blowing smoke. The ball is in your court, DRD. Provide ANY justification for claiming that more quasars have been discovered around NGC 3516 and that any which have will affect the conclusions of the probability calculation I made.

The models by Karlsson, Narlikar, Arp (etc.) are indeed predictive since long before NGC 3516 was looked at, they were predicting that cases like it would exist and be relatively common (and I've demonstrated they are common in this thread) ... cases that rather straightforward and simple probability calculations prove would be exceedingly unlikely under mainstream thinking regarding quasars ... even if one looked at every single observable quasar/galaxy association in the sky.

But the bottom line is, I think, that it is a hypothesis (or two) of yours that we are discussing, so the ball remains firmly in your court .... unless, perhaps, you are hinting at yet another difference between how you think astronomy, as a science, should be done and how it is actually done?

Another nice red herring ... or is it a strawman? But then I think we are learning that's all we can expect from you now since you obviously don't/can't/won't understand the probability calculations and what they mean.

 

[Dancing David]

David, the fact that can ask this question with a straight face after all that's been posted in the way of explanation on this thread proves you actually know and understand very little about probability and statistics ... that you even lack common sense and the ability to read. And that's why I chose not to engage you further in debate. It's a complete waste of time.

Funny that BAC, the mothodology you use can't address the question. Period.

So I wll type it in large fon because you are saying you will ignore it:

How can your methdology show a difference between a random placement and a causal one?

The truth is BAC, you haven't addressed this question at all in this thread. So pretend all you want, you haven't and the reason you don't answer is simple. You know you are wrong.

I challenge you:

Show me where in this thread that you have answered this specific question. Otherwise I will go to the Community Forum and call you a liar in a thread deidicated to that purpose.

I say that you haven't explained how you can use your methodlogy to tell a random placement from a causal one.
Which means you are a liar. You have until next Monday. Of I will post a thread called "BAC and Falsehoods" in the Community forum.

So you can pretend that you are some sort of mental wizard BAC, you would get a 'D' in high schools rhetoric. You might get an 'A' on Elegant Manipulation and Rhetorical Evasion.

If you show to me your explanation on the is thread I will post a "BAC was RIGHT" thread here and on the Community forum.

You can't and you won't, why? because

Your methodology can not distinguish a causal event from a random one.

 

[BeAChooser]

Hey BAC, it's your hypothesis ... if you want to re-state it in terms of a test of Karlsson peaks (or whatever), feel free to do so.

ROTFLOL! I hope everyone can see how DRD is now ducking and weaving.

But don't forget that for everyone reading this thread, this is completely new territory

Perhaps, but it's also rather straightforward math and common sense which I believe most people will understand. That's why I did the calculations step by step, like you insisted I should.

I (and I trust other readers) acknowledge that you are presenting only your own opinion

WRONG. I'm not presenting opinion. This is straightforward math and rather simple logic. You see an event (actually, several) take place that seem to have a certain pattern to them. Mainstream astronomers have told us certain other observations regarding quasars and galaxies, and how they should be distributed according to their theories. We just apply proven math (not opinion) to see if observing those events and patterns is to be expected. And it's clearly not. Not even if we had looked at every single quasar/galaxy association ... we we haven't. The probability of our seeing those specific observations given what the mainstream claims is the number and distribution of quasars is so unlikely that ... well, let me put it in terms we'll all understand ... Arp et. al. should go out and buy lottery tickets if they are really that lucky.

I think we are all agreed that the method you have used, in the calculations you presented, is not legitimate within astronomy

ROTFLOL! I don't believe "we are all agreed" to that. And certainly a number of very renown astronomers don't agree. On the other hand, I think there might be a growing consensus that you don't know what you are talking about when it comes to probability and statistics.

BeAChooser wrote: So the probability of those 5 quasars showing up close to the DRD values is 1/(10*9*8*7*6)/(5*4*3*2*1) = 0.004. That compares to a probability of 0.0000003 in my calculation using Karlsson's values. That's not even REMOTELY close. ... snip ...

Of course you don't get the same 'probabilities' if you use different inputs!

You stated that "as every 'particular configuration' is unique (or nearly so), its probability will be exceedingly small, as estimated by the BAC method, no matter what 'theories' are used!" Using a single example that I pulled out of my head without any difficulty, I proved you were wrong. And you haven't even got the honesty to admit that. Instead you spin and blow more smoke into this discussion. Every "particular configuration" does not give a probability that is exceeding small. The difference between those two probabilities is that in one case we expect to see cases like that if we look at every possible quasar/galaxy sample, while in the other case it will still be a very rare and unlikely event. And we certainly don't expect to see 4 or 5 (or more) such cases (like we have) given the limited amount of detailed observations that have actually been made. Fundamentally, DRD, you do not understand the nature of the probability calculations being made here. And I wonder if that isn't the problem with mainstream astronomers in general. They really don't understand probability ... just as they don't understand electromagnetism and it's affect on plasmas. So they avoid responding to peer reviewed papers on both subjects.

My point was, and still is, that if the hypothesis that you claim to be testing, with the calculations you present, is intended to be one of 'mainstream theories', then there's nothing special about Karlsson peaks (you can use any set of five numbers to do the test).

No, your point was to make our readers erroneously believe that I'd have gotten small probabilities out of my calculation no matter what peaks I put into the calculation. But you were wrong ... again.

In the 30' field around NGC 4030, there are 23 'QSOs' listed in NED; here are their redshifts, in order of increasing z: 0.132, 0.538, 0.735, 1.005, 1.011, 1.151, 1.211, 1.234, 1.257, 1.275, 1.326, 1.499, 1.567, 1.675, 1.812, 1.85, 1.892, 1.945, 2.063, 2.093, 2.2, 2.568, 3.592. (note that all but two have SDSS designations; presumably only two were known before the field was observed by SDSS).

Have a photo of the galaxy and where the quasars are located? And shall I use the latest (Bell's) list of quantized values (it's a more refined one then Karlsson's early list)?

 

[BeAChooser]

Um, er, ... BeAChooser, it's your hypothesis we are discussing (not mine).

But it's the mainstream hypothesis that you are defending ... rather unsuccessfully so far. Or is your reticence about providing an estimate for the number of observable quasars in the sky just *coyness* on your part.

Or are you saying that this kind of consistency is not important, in terms of how you think astronomy (etc) should be done?

No, I'm saying that I don't think you really know what you are talking about, don't know what most of the papers you've cited actually say, and are behaving pretty much like a troll who just wants to argue and make the thread as long as possible rather than actually reach a conclusion based on facts, logic, etc.

 

[Dancing David]

Perhaps, but it's also rather straightforward math and common sense which I believe most people will understand. That's why I did the calculations step by step, like you insisted I should.

And it can not tell a random placement from a causal one.

WRONG. I'm not presenting opinion. This is straightforward math and rather simple logic. You see an event (actually, several) take place that seem to have a certain pattern to them.

And you can't tell a random arrangement from a causal one with your method. They will give the same numbers every time.

Mainstream astronomers have told us certain other observations regarding quasars and galaxies, and how they should be distributed according to their theories. We just apply proven math (not opinion) to see if observing those events and patterns is to be expected.

Math that can not tell a random placement from a causal one.

And it's clearly not. Not even if we had looked at every single quasar/galaxy association ... we we haven't. The probability of our seeing those specific observations given what the mainstream claims is the number and distribution of quasars is so unlikely

Except you haven't addressed the probability of distribution, your numbers would be the same for a detrmined causal relationship and a random one.

that ... well, let me put it in terms we'll all understand ... Arp et. al. should go out and buy lottery tickets if they are really that lucky.

You still don't have a clue. How can you tell a random distribution of winning tickets? especially with your method.


Your method could not tell a random placement from a determined weighted placement. the numbers would be the same.

ROTFLOL! I don't believe "we are all agreed" to that. And certainly a number of very renown astronomers don't agree.

All 20 of them?

I hope they don't write actuarial tables.

On the other hand, I think there might be a growing consensus that you don't know what you are talking about when it comes to probability and statistics.

the irony, it burns.

You stated that "as every 'particular configuration' is unique (or nearly so), its probability will be exceedingly small, as estimated by the BAC method, no matter what 'theories' are used!"

your method can't tell random placement from a causal weighted one.

Using a single example that I pulled out of my head without any difficulty, I proved you were wrong.

No you have proved you overestimate your statitical skills. Insurance does not use your methods. Nor do bookies.

And you haven't even got the honesty to admit that. Instead you spin and blow more smoke into this discussion. Every "particular configuration" does not give a probability that is exceeding small.

Your method can't tell a random from a non random distribution. The numbers would be exactly the same the way you do it.

The difference between those two probabilities is that in one case we expect to see cases like that if we look at every possible quasar/galaxy sample, while in the other case it will still be a very rare and unlikely event. And we certainly don't expect to see 4 or 5 (or more) such cases (like we have) given the limited amount of detailed observations that have actually been made. Fundamentally, DRD, you do not understand the nature of the probability calculations being made here. And I wonder if that isn't the problem with mainstream astronomers in general. They really don't understand probability ... just as they don't understand electromagnetism and it's affect on plasmas. So they avoid responding to peer reviewed papers on both subjects.



No, your point was to make our readers erroneously believe that I'd have gotten small probabilities out of my calculation no matter what peaks I put into the calculation. But you were wrong ... again.

You sure are full of yourself. You don't know the difference between a causal relationship and a random one.

Have a photo of the galaxy and where the quasars are located? And shall I use the latest (Bell's) list of quantized values (it's a more refined one then Karlsson's early list)?

You might as well just say "goddidit","thebiblesaysso" and "Isayiamright".

You can't prove your method could tell a random distribution from a causal one.

BAC is a writer of falshoods.

 

[BeAChooser]

Hey BAC, you seem to have missed post #294, where I looked at 'the NGC 5985 case' for the first time!

See post #308.

I know that you have asked, several times, for comments on your calculations, well, in that post I point out that not one, not two, but three (yes, three!! ) of your inputs seem wrong

Actually, only two were wrong (and one of those was in error because the abstract of the published paper contained an error). The reason I didn't use the z=1.90 datum in the earlier calculations was clearly explained. You apparently missed that.

I also pointed out that there seem to be at least five other quasars within 30' of NGC 5985 (in addition to the two mentioned in the Arp paper).

Actually, the Arp paper mentioned 5 quasars. And the 5 you mentioned "in addition to the two in the Arp paper", two of those appear to have the same redshift as those in Arp's paper. As to the other 3 quasars, one is outside the 0-3 range and the other two we know nothing else about (like where they are located).

I wonder how many [of the five 'new' quasars] lie near the minor axis of NGC 5985?

There aren't 5, only 3, and one of those has a z well outside the 0-3 range of my calculation. As to the location of the other two, feel free to find out for us, DRD. I'll be happy to include them in the calculation although the probability will only end up smaller.

 

[DeiRenDopa]

I think you'd agree that they are all essential parts of the theory, or theories, you claim your hypothesis sets out to test.

* gravitational lensing, both weak and strong

WRONG AGAIN. That's becoming a habit with you, DRD. Gravitational lensing has no effect on the calculations I made. It doesn't affect the value of z of each quasar (i.e., it's apparent quantization) nor would it cause the alignment of the quasar "images" to be along the minor axis of a galaxy (at least you haven't provided any mechanism to explain how it could).

* clustering; more generally, large-scale structure (groups, clusters, filaments, super-clusters, ...)

WRONG. Again, that has nothing to do with any input or assumption in my calculations.

* definition(s) of 'quasar'.

Over and over in my calculations of probability, I've given your side the benefit of the doubt as to the number of quasars (whatever they are) and their distribution around galaxies. I've even asked you to tell us the number I should use. But you refuse, because you sense where that will lead.

.
Well, please do accept my apologies for so seriously misunderstanding the hypothesis that you say you have clearly, quantitatively, and unambiguously stated.

But let's review the history of what you have posted, re that hypothesis, shall we?

I'll pick it up from post #283, because the trail before that is easy to follow:

(part omitted)

In this case, the thing missing from BAC's posts on this topic (so far) is a clear statement, preferably quantitative, of just what the hypothesis he is testing is, and what the null hypothesis is.

I'm beginning to think you actually do have a reading problem. I've stated my hypothesis very clearly ... that the calculations I made (including the Bayesian portion) strongly suggest that the mainstream's explanation for quasars is WRONG.

(rest omitted)

.
Leaving aside the fact that "the calculations I made (including the Bayesian portion) strongly suggest that the mainstream's explanation for quasars is WRONG" is not a quantitative hypothesis, and leaving aside the fact that there is no null hypothesis, there is the question of what "the mainstream's explanation for quasars" actually is.

BAC, would you please state what this explanation is?

Then, from it, derive a clear, unambiguous, quantitative hypothesis that you seek to test?

At the same time, state, clearly and unambiguously, what the null hypothesis is?

.
Next, your reply in post #288:

Leaving aside the fact that "the calculations I made (including the Bayesian portion) strongly suggest that the mainstream's explanation for quasars is WRONG" is not a quantitative hypothesis

It most certainly is ... and the numerical results surely suggest the mainstream theory is faulty in some important manner.

there is the question of what "the mainstream's explanation for quasars" actually is. BAC, would you please state what this explanation is?

Why the one that leads to the mainstream's claim that all redshifts equate to distance, are randomly distributed across the sky, and are distributed in distance in the non-quantized manner I linked to earlier. Since that's all I use from the mainstream's explanation in my calculations, that's all I need to state. My calculation simply tests whether that explanation, whatever it is, is consistent with now three observations. And it doesn't appear to be.

Then, from it, derive a clear, unambiguous, quantitative hypothesis that you seek to test?

I guess after all my effort you still don't understand what those probabilities I calculated mean nor why I made that Bayes' Theorem calculation.

At the same time, state, clearly and unambiguously, what the null hypothesis is?

I did. You just didn't recognize it ... for the umpteenth time.

.
On to #291 (extracts only):

BAC: The calculation simply determines the probability of that particular configuration being found amongst the entire population of quasar/galaxies that we could observe assuming the mainstream's theories.

DRD: not once, in BAC's posts before this (and, I think, to date), has "the mainstream's theories" been stated or referenced ... how then can they have been used in any calculations?

BAC (bold added): there is nothing in the theory that requires quasars be at all the Karlsson values around any given galaxy at any one time.

DRD: If the hypothesis being tested is solely concerned with 'mainstream theories', then Karlsson peaks are irrelevant [...] If Karlsson peaks are important, then the hypothesis being tested cannot concern 'mainstream theories' alone.

Note that, here, 'the theory' refers to something other than 'mainstream theory' ...

#292 (extracts only):

BAC (bold added): Now let's add this latest observation into the Bayes' Theorem look at the likelihood the mainstream's theory is correct.

DRD: So, back to the 'hypothesis' question: what is the hypothesis that you claim this method (and data) tests? As I mentioned in my last post, it cannot be one that is, or involves, solely 'mainstream theories', if only because 'Karlsson peaks' are not part of any mainstream theory (that I know of).

#293(extracts only):

DRD: And as I have just pointed out, if this is so, then the hypothesis you claim to be testing cannot give any special status to 'Karlsson peaks', because they are not part of any 'mainstream theory'.

DRD: there is the question of what "the mainstream's explanation for quasars" actually is. BAC, would you please state what this explanation is?

BAC: Why the one that leads to the mainstream's claim that all redshifts equate to distance, are randomly distributed across the sky, and are distributed in distance in the non-quantized manner I linked to earlier. Since that's all I use from the mainstream's explanation in my calculations, that's all I need to state. My calculation simply tests whether that explanation, whatever it is, is consistent with now three observations. And it doesn't appear to be.

DRD: I don't know what you used as a source, that lead you to conclude that this is an accurate summary!

Maybe it's worthwhile to go through this, in some detail, if only so you can understand just how inaccurate it is?

.
And then there's more, going over much the same ground, again and again.

Two things seem to me to be clear:
* nowhere does BAC actually state the hypothesis he claims tests, in a quantitative fashion, "mainstream theories"
* the calculations BAC presents contain false premises, and cover at least three different things ('quantized redshifts' or Karlsson peaks, quasars on minor axes of a small number of galaxies, and the number of quasars around those galaxies).

Quantized redshifts and Karlsson peaks are easy to address: simply look at the distribution of quasar redshifts from SDSS for example (making very sure to understand what 'quasar' means, in this survey); they do no show any quantization, much less have redshifts at only the Karlsson peaks.

If BAC has a hypothesis, re Karlsson peaks, that he would like to present, then let's see it!

Minor axis anisotropies are also easy to address: those found by Arp (et al.) around NGC 5895 and NGC 3516 seem to be mere chance, and L-C&G's paper is clearly inconsistent, quantitatively, with the stated conclusions in the Arp et al papers. BAC has yet to present an account reconciling the relevant data in L-C&G with his calculations.

The numbers of quasars 'near' a galaxy is, perhaps, the one part of BAC's calculations that might be relevant to a test of 'mainstream theories' ... but if so then we should surely be discussing the 2005 Scranton et al. paper, shouldn't we? and asking how BAC's calculations are consistent with that.

But wait! We have already had that discussion!!

BAC, I know this is getting you very frustrated, but the only two people who are bothering to respond have, very clearly, not understood your hypothesis.

Would you please re-state it?

 

[DeiRenDopa]

re #308:

I wonder what would happen if we looked right out to 91' from NGC 5985?

Just curious. What's 91' out from NGC 5985?

.
According to Arp, RXJ 15509+5856 (actually only 90.4').

NED gives its position as "15h50m56.7s +58d56m06s"; Arp as "15/50/56.8 +58/56/04".

How many quasars (etc) are within 91' of NGC 5985?