tag:blogger.com,1999:blog-22973357.post1598980939104545961..comments2017-04-24T18:44:14.445-04:00Comments on Backreaction: The Bullet Cluster as Evidence against Dark MatterSabine Hossenfelderhttps://plus.google.com/111136225362929878171noreply@blogger.comBlogger104125tag:blogger.com,1999:blog-22973357.post-23075274619222472212017-01-18T07:35:13.034-05:002017-01-18T07:35:13.034-05:00"Also, check, eg, the discussion of the BC in...<i>"Also, check, eg, the discussion of the BC in this paper., page 5. That's exactly the kind of 'argument' I've heard countless times, repeated by literally thousands of people who work on particle dark matter."</i><br /><br />Actually, this is pretty tame. The BC is evidence that dark matter and baryons behave differently. They do. The whole idea of MOND is to Phillip Helbighttp://www.blogger.com/profile/12067585245603436809noreply@blogger.comtag:blogger.com,1999:blog-22973357.post-34852532322732044062017-01-18T07:19:53.720-05:002017-01-18T07:19:53.720-05:00"You are alleging a bias against generating, ...<i>"You are alleging a bias against generating, submitting, or accepting work that disfavors LCDM cosmology, a claim for which you have no evidence."</i><br /><br />Yes, maybe, probably not. Let's take them in reverse order.<br /><br />I don't think that there is a bias against non-LCDM results in journals, i.e. accepting. As to whether someone with a non-permanent position Phillip Helbighttp://www.blogger.com/profile/12067585245603436809noreply@blogger.comtag:blogger.com,1999:blog-22973357.post-54796433516072276472017-01-16T00:16:54.435-05:002017-01-16T00:16:54.435-05:00Peter,
You continue to accuse me of opinions I do...Peter,<br /><br />You continue to accuse me of opinions I do not hold which you then attempt to tear down. It's a classic straw-man argument. Look, I wrote this very post to point out that there *are* papers arguing that observations contradict LambdaCDM. Why do you think you have to tell me that? What I'm saying is that this evidence is widely ignored and quickly discussed away and I'Sabine Hossenfelderhttp://www.blogger.com/profile/06151209308084588985noreply@blogger.comtag:blogger.com,1999:blog-22973357.post-56715907276988709982017-01-15T20:10:29.102-05:002017-01-15T20:10:29.102-05:00Since you asked about this:
Where are the people ...Since you asked about this:<br /><br /><i>Where are the people trying to improve the numerical simulations for matter forming structures in modified gravity?</i><br /><br />Well, they're out there, and they're being published in the same journals as the LCDM simulations. For example:<br /><br /><a href="https://arxiv.org/abs/1104.5040" rel="nofollow">Angus & Diaferio (2011): The Peter Erwinhttp://www.blogger.com/profile/18415612458902079584noreply@blogger.comtag:blogger.com,1999:blog-22973357.post-18257460391508894912017-01-15T14:45:34.371-05:002017-01-15T14:45:34.371-05:00(Continuation of previous comment:)
You are alleg...(Continuation of previous comment:)<br /><br />You are alleging a bias against generating, submitting, or accepting work that disfavors LCDM cosmology, <i>a claim for which you have no evidence.</i> I have tried to point out that the actual publication record shows that people are <i>not</i> dissuaded from publishing such work -- in fact, some authors are perfectly comfortable with publishing Peter Erwinhttp://www.blogger.com/profile/18415612458902079584noreply@blogger.comtag:blogger.com,1999:blog-22973357.post-43289003229143648422017-01-15T14:03:24.360-05:002017-01-15T14:03:24.360-05:00Sabine,
From your earlier comment (since it seems...Sabine,<br /><br />From your earlier comment (since it seems to summarize your argument):<br /><br /><i>How many computer simulations have been made which did not increase the probability of a Bullet-Cluster like event and how many of these negative results did get published? How many people would publish these results? How many journals would publish them? What's the merit of fiddling with Peter Erwinhttp://www.blogger.com/profile/18415612458902079584noreply@blogger.comtag:blogger.com,1999:blog-22973357.post-29674168850285032362017-01-15T12:25:32.185-05:002017-01-15T12:25:32.185-05:00so I'm not sure what I've learned from tha...<i>so I'm not sure what I've learned from that exchange except that the probability I've quoted isn't particularly meaningful. </i><br /><br />Well, that <i>was</i> sort of the original point that I was trying to make: that you had misinterpreted the meaning of the probability, and thus misreported what the most recent papers were saying.Peter Erwinhttp://www.blogger.com/profile/18415612458902079584noreply@blogger.comtag:blogger.com,1999:blog-22973357.post-47488798338103378592017-01-15T00:20:11.916-05:002017-01-15T00:20:11.916-05:00Peter,
Well, I admit it isn't entirely useles...Peter,<br /><br />Well, I admit it isn't entirely useless knowledge. At least you know there shouldn't be more. But look, to use your example: You want to know how many diseases you'll be exposed to in your lifetime. A doctor tells you there are 20 billion diseases somewhere in the universe. The only thing you learn from that is at least you won't contract more than 20 billion. Sabine Hossenfelderhttp://www.blogger.com/profile/06151209308084588985noreply@blogger.comtag:blogger.com,1999:blog-22973357.post-52572713333958285682017-01-14T09:56:09.262-05:002017-01-14T09:56:09.262-05:00You mention a paper by, as you say 'basically ...<i>You mention a paper by, as you say 'basically the same authors' that used 'what they argued was a better and more sophisticated analysis' to arrive at a result more favorable to LamdaCDM to argue against my concern?</i><br /><br />Yes, because your "concern" was that papers like Thompson et al. (2015) had "adapted the model to fit observations". There's Peter Erwinhttp://www.blogger.com/profile/18415612458902079584noreply@blogger.comtag:blogger.com,1999:blog-22973357.post-85422746818797034902017-01-14T08:15:51.431-05:002017-01-14T08:15:51.431-05:00Thanks again. I don't understand why predictin...<i>Thanks again. I don't understand why predicting what we observe requires to model the observation. If we observe N cluster collisions, then can't we say how many of these are expected to have masses of m_1/2 and a relative velocity of x? That's what I was asking for. That doesn't require to predict N to begin with. Either way, the total number of such events in the observable Peter Erwinhttp://www.blogger.com/profile/18415612458902079584noreply@blogger.comtag:blogger.com,1999:blog-22973357.post-78911362415572035472017-01-14T00:40:45.839-05:002017-01-14T00:40:45.839-05:00Peter,
Thanks again. I don't understand why p...Peter,<br /><br />Thanks again. I don't understand why predicting what we observe requires to model the observation. If we observe N cluster collisions, then can't we say how many of these are expected to have masses of m_1/2 and a relative velocity of x? That's what I was asking for. That doesn't require to predict N to begin with. Either way, the total number of such events in Sabine Hossenfelderhttp://www.blogger.com/profile/06151209308084588985noreply@blogger.comtag:blogger.com,1999:blog-22973357.post-84240148291838060012017-01-13T15:25:08.224-05:002017-01-13T15:25:08.224-05:00You say nobody has adapted the model to fit observ...<i>You say nobody has adapted the model to fit observations. Are you telling me there's only one numerical code that you guys all agree on is undoubtedly the one that follows from LambdaCDM and that one numerical code hasn't changed since you learned of the Bullet Cluster?</i><br /><br />There's no such thing as a numerical code which "follows from LambdaCDM", and you wouldnPeter Erwinhttp://www.blogger.com/profile/18415612458902079584noreply@blogger.comtag:blogger.com,1999:blog-22973357.post-88597480454055260462017-01-13T15:18:14.358-05:002017-01-13T15:18:14.358-05:00Phillip,
Statistics with one object is a bit comp...Phillip,<br /><br /><i>Statistics with one object is a bit complicated. The Poisson error of 1 is 1</i><br /><br />Well, the Poisson "error" of 1 is the Poisson distribution for a rate of 1, which excludes any value <= 0. You're assuming the Gaussian approximation for Poisson statistics, which really breaks down for small numbers. (But, yes, statistics involving just one object Peter Erwinhttp://www.blogger.com/profile/18415612458902079584noreply@blogger.comtag:blogger.com,1999:blog-22973357.post-16350222085138014002017-01-13T13:30:27.124-05:002017-01-13T13:30:27.124-05:00Sabine,
I don't know why the number in the ob...Sabine,<br /><br /><i>I don't know why the number in the observable universe is relevant. What's relevant seems to me the number in the actually observed universe.</i><br /><br />In order to <i>predict</i> "the number in the actually observed universe" (for comparison to the actual observations), you would have to simulate the actual observing process to date -- which parts of Peter Erwinhttp://www.blogger.com/profile/18415612458902079584noreply@blogger.comtag:blogger.com,1999:blog-22973357.post-84206590484025404622017-01-13T12:27:04.265-05:002017-01-13T12:27:04.265-05:00Peter,
Thanks for the explanation, but I think yo...Peter,<br /><br />Thanks for the explanation, but I think you misunderstood my question. See Phillip's comment above, who understood it. I don't know why the number in the observable universe is relevant. What's relevant seems to me the number in the actually observed universe. Or else I might misunderstand what you mean by 'observable.' I take it to mean anything that's Sabine Hossenfelderhttp://www.blogger.com/profile/06151209308084588985noreply@blogger.comtag:blogger.com,1999:blog-22973357.post-23160342451881258432017-01-13T11:34:44.756-05:002017-01-13T11:34:44.756-05:00I still fail to see though how it counts as eviden...<i>I still fail to see though how it counts as evidence for LambdaCDM that it's possible to adapt the model to fit observations.</i><br /><br />But that's not what they've been doing; no one has been "adapting the model to fit the observations".Peter Erwinhttp://www.blogger.com/profile/18415612458902079584noreply@blogger.comtag:blogger.com,1999:blog-22973357.post-34300757644248367682017-01-13T11:24:33.429-05:002017-01-13T11:24:33.429-05:00I would agree that the particular type of probabil...I would agree that the particular type of probability these papers have been using is a slightly strange way to approach the problem, at least for people outside their sub-discipline. I think it's one of those cases where the first paper or papers adopted a particular approach for idiosyncratic reasons, and the subsequent papers use the same approach for purposes of comparison. The first Peter Erwinhttp://www.blogger.com/profile/18415612458902079584noreply@blogger.comtag:blogger.com,1999:blog-22973357.post-2255021117806417122017-01-13T11:20:09.815-05:002017-01-13T11:20:09.815-05:00Sabine,
Why, if I want to know the probability of...Sabine,<br /><br /><i>Why, if I want to know the probability of seeing the event, would I multiply it with the number expected in the universe</i><br /><br />Multiplying the probability of seeing event X in a single object of class A by the total number of objects in class A gives you the expected number. If the expected number is 1000 and you see 940, not a problem; if you only see 1 or 2, thatPeter Erwinhttp://www.blogger.com/profile/18415612458902079584noreply@blogger.comtag:blogger.com,1999:blog-22973357.post-14730055708251589862017-01-13T10:16:46.168-05:002017-01-13T10:16:46.168-05:00Phillip,
Well, I was thinking it's the probab...Phillip,<br /><br />Well, I was thinking it's the probability that a Bullet-Cluster like event has a relative velocity in the range that's observed. How many such events have we observed in total (disregarding the velocity)? I thought it's a handful or so. Hence, I was thinking, the two probabilities are pretty much the same, take or give a a factor accounting for the handful. Not Sabine Hossenfelderhttp://www.blogger.com/profile/06151209308084588985noreply@blogger.comtag:blogger.com,1999:blog-22973357.post-52273128880578446402017-01-13T09:00:32.715-05:002017-01-13T09:00:32.715-05:00I think it is clear that the low number is not the...I think it is clear that the low number is not the probability that we would observe one bullet cluster. Thus, multiplying the probability per "event" by the number of "events" makes sense. However, in one of the papers they state a probability of finding something like the bullet cluster as one such system per 10 billion cubic megaparsecs. So this is one in a cube a bit Phillip Helbighttp://www.blogger.com/profile/12067585245603436809noreply@blogger.comtag:blogger.com,1999:blog-22973357.post-89986692642882610512017-01-13T00:33:25.208-05:002017-01-13T00:33:25.208-05:00Peter,
Why, if I want to know the probability of ...Peter,<br /><br />Why, if I want to know the probability of seeing the event, would I multiply it with the number expected in the universe, given that we don't observe most of what is out there? That's how I interpreted the number. It seems that I got that wrong, but I would appreciate if you could clarify the relevance of the number you refer to. What do I learn from knowing there are Sabine Hossenfelderhttp://www.blogger.com/profile/06151209308084588985noreply@blogger.comtag:blogger.com,1999:blog-22973357.post-37995354867134883042017-01-12T18:17:22.376-05:002017-01-12T18:17:22.376-05:00... some inventive humanoids had optimized the dar...<i>... some inventive humanoids had optimized the dark-matter based computer simulations and arrived at a more optimistic estimate of a probability of 4.6×10-4 for seeing something like the Bullet-Cluster. Briefly later they revised the probability again to 6.4×10−6.</i><br /><br /><i>Either way, the Bullet Cluster remained a stunningly unlikely event to happen in the theory of particle dark Peter Erwinhttp://www.blogger.com/profile/18415612458902079584noreply@blogger.comtag:blogger.com,1999:blog-22973357.post-28953061298178113482017-01-12T13:37:32.619-05:002017-01-12T13:37:32.619-05:00Very well said, Sabine. Nailed it. The early unive...Very well said, Sabine. Nailed it. The early universe has a quite straight-forward explnation, it turns out...richard.jowseyhttp://www.blogger.com/profile/02508609964696300986noreply@blogger.comtag:blogger.com,1999:blog-22973357.post-31715036555140226162017-01-12T05:56:01.578-05:002017-01-12T05:56:01.578-05:00thanks!thanks!Nima Khosravihttp://www.blogger.com/profile/14179086668894768438noreply@blogger.comtag:blogger.com,1999:blog-22973357.post-60677453583891599042017-01-12T05:52:48.162-05:002017-01-12T05:52:48.162-05:00Hi Nima,
For references, please see section 8.3 o...Hi Nima,<br /><br />For references, please see <a href="https://arxiv.org/abs/1112.3960" rel="nofollow">section 8.3 of this review.</a> Best,<br /><br />B.Sabine Hossenfelderhttp://www.blogger.com/profile/06151209308084588985noreply@blogger.com