Talk:2011 OPERA faster-than-light neutrino anomaly/Archive 3

Proposed removal of "Discussions within the OPERA collaboration"
I propose to remove the section "Discussions within the OPERA collaboration". The reasons are the following: 1) any Collaboration is such because it has one voice. The section is deliberately violating the unity and integrity of the Collaboration. This is different from the right to sign a paper. The person who posted this contribution clearly ignores this. 2) What are reported are rumors, and as such they should come from inside OPERA. Then the source of the rumors has to be quoted (name/surname). Otherwise anybody could just spread rumors about "discussions inside a government" or "discussions inside a company". This is crazy. If the rumors are not coming from inside OPERA, they are just lacking any sense and should be removed. 3) It appears to me that this section is just an attempt to weaken the scientific result on "sociopolitical" basis. Kryssb (talk) 10:38, 29 November 2011 (UTC)
 * The secondary sources for those claims are given in the article. However, I think we can remove this section in the future, in case that those rumors are not mentioned any more in the next "ScienceMag" article. --D.H (talk) 10:59, 29 November 2011 (UTC)
 * I agree with Kryssb per all my discussions in the sections above. Additionally those secondary sources disagree on the details and as such the information is not reliable and should not be included, rather we should wait until all sources agree. Sources should not disagree on facts and if they do they are unreliable. They are based on rumours which are in disagreement with each other. There are times when sources offer different analyses, and they can of course be compared and contrasted. But facts are facts, if the sources can't agree on them then it's not reliable and we shouldn't be including it. Polyamorph (talk) 11:03, 29 November 2011 (UTC)
 * ... unless the disagreement between sources is sufficiently notable to mention just that as a controversy. - DVdm (talk) 11:21, 29 November 2011 (UTC)
 * Perhaps but I don't think that is the case here, unless you can explain why it is notable? Polyamorph (talk) 12:29, 29 November 2011 (UTC)
 * (One portion of irony on its way) — Perhaps the fact that there is so much fuss over it, even right here? - DVdm (talk) 13:45, 29 November 2011 (UTC)
 * Yes but unfortunately wikipedia is not a reliable source. Is there anyone off wikipedia that has reported the controversy.Polyamorph (talk) 15:59, 29 November 2011 (UTC)
 * We don't really have any WP:POLICY against discussions from inside "governments", "companies" or scientific collaborations. Wikipedia is not a public relations outlet for the entities involved.--Anders Feder (talk) 11:29, 29 November 2011 (UTC)
 * I hope this does not mean that anybody can mess up real information with his/her own fantasies. The text mentions some issues, but it deliberately stays vague. What does it mean "the dissenters consider this poor experimental procedure"? Following what is written, I can draw the conclusion that person A, B, C expressed exactly that statement. What is the source? This is typical of rumors: seemingly precise issues, but in practice vague indications. There will never be an agreement on that, unless the persons who did not sign give an explicit statement. It seems to me that what is reported is speculations about what is in somebody's mind, and the only (maybe ineffective) way I know to learn people's mind is to directly ask them. Kryssb (talk) 11:39, 29 November 2011 (UTC)
 * It's not a rumor, it's a statement made by a reliable source (ScienceInsider).--Anders Feder (talk) 12:20, 29 November 2011 (UTC)
 * If you are disagreeing with the content of Cartlidges article, then you have to provide other sources, or wait for the next "Science"-news article. Above, we came to an agreement that the "News"-sections of "Nature" and "Science" are the most reliable ones for this article (at least as long we are relying only on News-Sources...) --D.H (talk) 12:17, 29 November 2011 (UTC)
 * Yes, I disagree. The article quotes "a source within the Collaboration". Fine. I can claim I have a source in Mars inhabitants (if any, but another source at NASA tells me they exist). Either the source is quoted in name+surname or it's just mud. Kryssb (talk) 12:42, 29 November 2011 (UTC)
 * Then publish your findings on Science or Nature etc... Afterwards, we can use it as a secondary source and include it here. --D.H (talk) 12:49, 29 November 2011 (UTC)
 * Fine policy for a school journal. So, instead of focusing on the scientific subject, which requires hard work and time, and may be fruitful to strengthen or to disprove the result, you prefer to keep unverifiable statements based on the reputation of a news writer, without even warning the reader that the pretended "source in the Collaboration" may well not exist at all as far as you know. That's a good example of how the myth of supposedly objective reporting falls. — Preceding unsigned comment added by Kryssb (talk • contribs) 13:39, 29 November 2011 (UTC)
 * The source may well exist as far as you know. Also, this is not a matter of preference, it's a matter of policy.--Anders Feder (talk) 13:58, 29 November 2011 (UTC)
 * WP:VNT is an essay, not policy. WP:V from where it comes from is not a licence to include everything that is verifiable. Polyamorph (talk) 15:54, 29 November 2011 (UTC)
 * The source does not exist until it is proven to exist, and this is not a matter of policy. Otherwise, selecting what to publish and what not is just a way to control the opinions of the reader without taking the responsibility of expressing a judgement in first person. Kryssb (talk) 14:02, 29 November 2011 (UTC)
 * And why is your judgement better than everyone elses?--Anders Feder (talk) 14:09, 29 November 2011 (UTC)
 * The policy of publishing verifiable statements is not being respected. The fact that the source exists simply cannot be verified. Hence the content of the article is meaningless. It's not my judgement, it's just under everybody's eyes, unless one wants to look aside or shut them. And the fact that keeping a section in the article orients at least a part of the readers to disbelieve the result on a purely socio-political basis, rather than scientific, is also in plain sight. Kryssb (talk) 14:32, 29 November 2011 (UTC)
 * The policy of publishing verifiable statements defines verifiability as: "whether readers can check that material in Wikipedia has already been published by a reliable source, not whether editors think it is true." So, yes, it is being respected.--Anders Feder (talk) 14:57, 29 November 2011 (UTC)
 * Sorry but you are interpreting that policy wrong. It means that only information that is verifiable can be included in wikipedia. If an editor thinks that something is true but they can't prove it (i.e. they don't have a reliable source to back it up) then they can't include it in wikipedia. It doesn't mean, as your interpretation suggests, that if we see something is false or unreliable that we should include it simply because it is verifiable. As a matter of fact the wording of that policy currently under intense discussion. We have a number of sources which don't agree on the facts, no evidence is given by the sources so we have no reason to believe a word that they say. Hence we can use editors discretion and simply leave that out until some more robust sources come along. Polyamorph (talk) 15:54, 29 November 2011 (UTC)
 * I am not interpreting the policy wrong. It is as it stands and it is very clear: the threshold is whether material has been published by a reliable source, not whether you or someone else think it is false or "unreliable". If you want to have it taken out anyway, you should show that it conflicts with another policy or build consensus for its removal some other way. And there is no policy against reliable sources based on unnamed sources.--Anders Feder (talk) 16:33, 29 November 2011 (UTC)
 * As I said there is an intense discussion of this verifiability, not truth statement http://en.wikipedia.org/wiki/Wikipedia_talk:Verifiability/First_sentence . We write wikipedia, we have discretion as to whether it is suitable for inclusion or not. As I have said the sources don't disagree hence we have no reason to believe it is accurate. It is better to have nothing than something that is unreliable. Polyamorph (talk) 16:49, 29 November 2011 (UTC)
 * So your source and my source does not agree and somehow that defaults to your source being the only one that should be represented. How is that?--Anders Feder (talk) 16:59, 29 November 2011 (UTC)
 * No it's all bollocks. Leave it out, why bother including it. None of the sources agree so why include any of them. Until there is some consistency I just don't see the point. And we can't have a comparison of the different views reported by the different sources unless that comparison has been reported in a reliable source. Polyamorph (talk) 17:02, 29 November 2011 (UTC)
 * We absolutely can make that comparison. In fact, we are required to by WP:WEIGHT.--Anders Feder (talk) 17:12, 29 November 2011 (UTC)
 * We didn't compare or evaluate the claims in the section, we just stated them based on the sources. Also: Nobody said that we "must" include this section, but we certainly "can" include this section since the discussions are mentioned by several secondary sources. So as long as there are editors who want to include it, they are in a better position due to the existence of those sources. All I've heard thus far against its inclusion, are personal reasonings of some WP-editors, which doesn't count. So let's wait for the next News-publications - I'm pretty sure we can remove this section, when it is not mentioned any more by them...--D.H (talk) 17:06, 29 November 2011 (UTC)
 * Personal reasonings of some WP-editors does count, because we write wikipedia. There is nothing to say we have to include something just because it is sourceable. But yes, I'm finding it difficult to formulate my reasoning very well since you do have the trump card of having reliable sources. You'll notice I haven't attempted to remove the section myself so do respect your side. I'll concede for now because I'm not going to win this but I really hope some better sources are forthcoming and I do stand by my POV that the sources are unreliable and speculative, per reasoning given in this and other sections above. Polyamorph (talk) 17:14, 29 November 2011 (UTC)


 * We didn't really come to an agreement, though we can agree to disagree if you like? :) I would like a more robust source to be found and for all sources to be in agreement. Rather than pieceing together a story from several sources which disagree on the details. We have no idea if any of it is true. My instinct is that the picture Cartlidge is painting is not a valid representation of the group dynamics, of course I have no evidence for that and we don't go on instinct. But I'd rather Cartlidge actually provide some substantial evidence for their claims. I agree that if another source comes a long that backs up this story and it is widely reported and the facts are 100% reliable then we should include it. Until then I am apprehensive in keeping the content in and only removing it if we don't find any new sources to corroborate the story. Ultimately though, I think it's unimportant and irrelevant until the final peer reviewed paper is published. Polyamorph (talk) 12:29, 29 November 2011 (UTC)
 * I can only say: There are several News-sources which say that there were discussion before and possibly after the experiment. This makes it notable, at least for the time being. However, when no further reports (in the relevant sources) appear that mention this issue, then we should remove it. --D.H (talk) 12:49, 29 November 2011 (UTC)
 * I think the political aspects are relevant even in a scientific experiment. I too think we should keep it for a while longer, until reports on the JHEP publication come in (I am pretty sure the submission would be accepted after changes). I don't think we can be in the business of vetting secondary sources. Science Mag has face credibility. Others also have reported the OPERA team had trouble with getting all its people to sign the JHEP paper.
 * That said, I also think we should look at the frequency of a publication to figure out its credibility. Weeklies tend to be more credible than dailies simply because they have a longer time to edit and digest the information. Nature News, Scientific American, and the CERN bulletin are weeklies. Science Mag and NYTimes are dailies (the last a general news daily). Ajoykt (talk) 16:14, 29 November 2011 (UTC)
 * Surely the correct choice here is to note that there have been reports of disagreements and where to find those reports. That is not something that even needs to be deleted later; only stated as refuted if it should ever happen. The details of the alleged disagreements ought not be reported in the article given that the sources do not agree and therefore cannot themselves all separately be reliable reports. But since there is no question of the verifiability or reliability of the fact that disagreements have been reported, there ought also be no controversy in reporting it. Strebe (talk) 19:20, 29 November 2011 (UTC)
 * I thought we did capture three things: that there were disagreements within OPERA, that reports on these conflict, and that the reports of everyone-signing-on are later in time than the disagreement-reports. If those three facts don't show up in the sentences, I would say modify the sentences to make them explicit. Ajoykt (talk) 19:28, 29 November 2011 (UTC)

SN1987A section
I moved (with some changes) the comment on SN1987A from "Previous measurements" into a new section "OPERA versus SN1987A" of the Discussion part. --D.H (talk) 17:47, 1 December 2011 (UTC)
 * But it is no longer in the Discussion part now. It is under the "Reception by the Physics Community" heading. Ajoykt (talk) 22:58, 1 December 2011 (UTC)
 * Since it plays an important role in the reception/discussion of the result and its theoretical explanation, the reception section seems in my view to be the better place. Maybe in the future, when more papers have appeared, the whole "previous measurements" should be moved to the reception section. --D.H (talk) 16:24, 2 December 2011 (UTC)
 * I moved it before seeing your response. We can keep it in the reception section, but then we should logically link the results to how physicists consider the results a possible contradiction of OPERA. Right now the wording just flatly presents the SN1987A results. If it were to be in the reception section, then we would have to quote physicists (either specific ones, or use qualifers such as some/most if reliable secondary sources say so) saying what the SN1987A results are, and how they violate OPERA's results assuming certain things hold true. Such quotations do exist - will see if I can dig them up. Ajoykt (talk) 19:26, 2 December 2011 (UTC)

(Proposed new section) Measurement challenges
The result depends on errors in the measurement of both the distance and time of flight of the neutrinos being held to just over 3 parts per million. Since both ends of the experiment are deep underground on opposite sides of mountain ranges, this is not easy. Nor is it easy to verify that the numerous calibrations and corrections were all performed correctly.

While GPS surveying and common-view time transfer techniques can routinely achieve this level of accuracy on the surface, the OPERA experiment has numerous links to connect surface time and position references to the experiments in the underground caverns. Much of the effort put into confirming this result went into re-measuring the exact distance between the beam measurement points (731278.0±0.2 m)  and verifying the time measurement at both endpoints.

At this level of accuracy, even a surveyed distance is not constant. The experiment measures ±1 cm daily fluctuations due to tides,  1 cm per year changes due to continental drift,  and a 7 cm abrupt change due to the 2009 L'Aquila earthquake. Fortunately, all of these effects are too small to affect the result.

Time is even more difficult, because it is continuously changing. The time delays caused by each metre of wire or fibre optic must be measured an accounted for, as well as concerns that the delays may change with temperature.

Time measurement at CERN
At the CERN end, the standard GPS-synchronized "general machine timing" signal is fed from the CERN control room to a control timing receiver near the point where the proton beam is extracted from the SPS. This entails a 10085±2 ns delay. (Due to the length of this delay, it was re-measured multiple times to ensure that it was consistent and stable. ) The signal from the control timing receiver is sent to a high-speed waveform digitizer, where it arrives after an additional delay of 30±1 ns. Another input to the waveform digitizer is fed from a beam current transformer which detects the passing of the proton beam to the neutrino-generating target and the detector. This signal arrives at the waveform digitizer 580±5 ns after the passing of the proton bunch. By comparing the two channels of the waveform digitizer (which is itself started slightly in advance of the expected proton arrival ), the time difference between the proton pulse and the GMT signal can be established with high accuracy.

The protons (and the pions they are transformed into before decaying into neutrinos) do not travel at quite the speed of light, but the difference is too small (approximately 0.02 ns ) to affect the results. A slightly larger effect is produced by the neutrinos absorbed by rock just in front of the OPERA detector and producing muons which are detected, but this is also too small (2 ns ) to significantly affect the result.

The CERN GMT system is not accurate enough for this measurement. It is not practical to upgrade the entire CERN timing system to the level of accuracy required by this one experiment, so instead its errors are measured and subtracted out of the results. A pair of identical high-resolution timing systems were installed at both CERN and LNGS. Using a cesium atomic clock and a common-view GPS receiver, the CERN GMT timing is measured each second to 0.1 ns resolution so that its error may be subtracted from the results.

The high-resolution timing systems were calibrated aginst each other by the Swiss federal office of metrology (METAS) before installation, and independently by the German national standards laboratory (PTB) after installation. The measured difference (2.3±0.9 ns) was subtracted from the results.

Time measurement at LNGS
The basic structure of the system at LNGS is similar to that at CERN, although the details differ. The laboratory is synchronized to a conventional GPS receiver on the surface, which is not itself accurate enough, but it is continuously measured against a high-resolution GPS and atomic clock combination so that the errors may be calibrated out.

The time signal is transmitted (in the form of 1000 pulses per second) to the underground laboratory over an 8.3 km long optical fibre, introducing a delay of 40996±1 ns. The underground laboratory is itself operated from a 20 MHz high-stability master oscillator. This oscillator is not actively synchronized to the external time signal, but rather its time errors (which drift by about 2 ps/s according to its 1 s Allan deviation) are also continuously measured and subtracted out of the results. Because of the 50 ns clock frequency, any single measurement may vary by ±25 ns, but this can be reduced by averaging.

The 20 MHz clock is used to generate a 100 MHz clock for a series of FPGAs which timestamp the signals coming from the neutrino detector. There is a 4263±1 ns propagation delay within the laboratory. The FPGA introduces an average 24.5±1 ns time delay in the measurement. (Because of the 10 ns granularity of the FPGA clock, this is 24.5±5 ns for any single measurement, but as the neutrino arrival times are not correlated with the clock, the error is reduced by averaging.)

The neutrino detector consists of numerous scintillators feeding photomultiplier tubes which are connected to the FPGAs. The delay varies depending on the position in the scintillator, but can be averaged over the measurements to produce a total delay of 59.6±3.8 ns from particle impact to FPGA timestamping. [FIXME: What about the 59.6±7.3 ns figure mentioned on p.15 of the paper?] (This was measured using a picosecond laser to simulate particle impacts.)

Additional corrections
The experimenters considered a number of minor corrections caused by the experiment's location on a planet rather than floating in empty space.

The Sagnac effect due to the motion of the two endpoints on the rotating Earth increases the expected time of flight by 2.2 ns. (After the neutrino beam's 2.4 ms trip from CERN, the Earth's rotation has carried LNGS approximately 1 m east of where it was originally. Because the beam is travelling south-east, this delays its arrival.)  The effects of the Earth's orbital velocity around the Sun and the centre of the galaxy are negligible, as are relativistic effects due to the gravity of the  Moon, and Sun, and Milky Way. The effect of the Earth's gravity is approximately 10−8 (0.01 ppm) and likewise negligible. The clock rate difference due to the different altitudes of CERN and LNGS is even smaller.

Bunch structure
Because the primary goal of the OPERA experiment requires measuring as many neutrinos as possible, they are generated in large 10.5 μs (10500 ns) long bunches. The timing measurement requires accurately aligning the detected bunches with the generated bunches. This alignment was itself carefully tested with simulated data and various subsets of the data, but critics felt that even a slight tendency for the detector to be more sensitive to the beginning of a bunch than the end could produce a misalignment large enough to explain the observed effect.

For example, the protons hit graphite rod targets to produce pions which are magnetically focused before decaying into neutrinos. The rods are heated by the proton pulse. If the graphite's thermal expansion were enough (this was computed and it is not), that could reduce the pion (and thus neutrino) production rate during the later part of each pulse.

To address such concerns, the OPERA experiment had CERN generate much shorter bunches, 3 ns long at 524 ns intervals, for 2 weeks. Although this only produced neutrinos at 1/60 the normal rate, it allowed the timing to be confirmed with no concerns about bunch shape.

Blind analysis
The discrepancy of 57,8 ns was estimated as follows: First, a blind analysis was carried out to avoid observer bias. Therefore, superseded and incomplete values for distances and delays were used from the year 2006. The data analysis of the measurement under those "blind" conditions gave an early neutrino arrival of 1043,4 ns. Afterwards, the data were analyzed again under consideration of the complete and actual sources of errors. If neutrino and light speed were the same, a correction value of -1043,4 ns should have been obtained. However, the actual value only amounted to -985,6 ns.

End of draft; start of discussion about it

 * A couple of issues - you need to find secondary sources to support the stuff (look around on the CERN/LNGS site). You also need to integrate this with what is already there in the article on distance and time measurement (the METAS/PTB calibration, common-view GPS, cesium clock, etc are already there, sourced from a CERN article). Some of your stuff relates to the first run, and some to the second run (I think the Sagnac effect, for example, was considered only in the second analysis). You need to compress the beginning part - just a sentence or so on how difficult the measurements are should be good. The central issue, from my point of view, is the secondary source meant for a non-physics audience. The BBC source, Discover:Neutrinos, for example, could be cited for some of the stuff regarding using laser beams for cable chain timing at LNGS. Another secondary source (granted not really for a general-purpose audience, but seems understandable to the layperson): https://indico.fnal.gov/getFile.py/access?resId=0&materialId=slides&contribId=43&sessionId=9&subContId=0&confId=4887 Ajoykt (talk) 01:14, 22 November 2011 (UTC)


 * I think you're over-fixated on the word "secondary". I certainly agree that a secondary source generally has more synthesis and analysis, and provides a better-balanced overview of a subject.  Especially in a field like history.  But here, all we have are a variety of press releases and news articles and blog posts that, with varying degrees of fidelity, directly quote the OPERA group's paper.  The Wikipedia concept of a secondary source, meaning one that synthesizes information from multiple primary sources, simply doesn't exist here.  (When we want analysis, I fully agree that we need to find an analyst to quote.)
 * Remember, careful choice and attribution to sources is most important for controversial material, where there are multiple conflicting points of view and "the truth" is hard to discern. While I have no objection to more detailed citation, I'm wondering what the point is here.  Absolutely nobody (except maybe some crackpots attracted by the media attention) is questioning the basic accuracy of the OPERA group's description of their experiment.  This is not material that is likely to be challenged.
 * "Please make it easier for me to track what you wrote back to the reliable sources you used" I can understand. "This is unsuitable for Wikipedia as-is" seems a bit overly fussy.  Is there anything in there that you would like to challenge as being of dubious veracity?  The difference is one between "could be better" and "fails to pass the minimum quality threshold".  I hope you can see how the former is a bit more motivating.  If that's what you mean to say, please say so more clearly!
 * "You also need to integrate this with what is already there..." Yes, I can tweak it, but I think that removing redundancy is most easily done on the live article (I'll add the new, then remove the unnecessary), and I don't want to start until I know the new is going in.  If someone is going to delete it again, it would be bad.  I'm not going to paste the wole article into the Talk page here to develop a draft, only to have to merge the results with the edits made to the live article in the meantime.  For a whole new section, it's possible.
 * As for the difference between the runs (a misleading phrase in any case, as the "first" run has never stopped, except for an interruption to do the 3 ns bunch experiment), there is no difference other than bunch structure. The geodesy and timing setups are identical.
 * "You need to compress the beginning part - just a sentence or so on how difficult the measurements are should be good." I have a better idea: WP:Be bold and edit in better wording yourself!  The first section is currently 4 paragraphs, 3+2+3+2 = 10 sentences.  If you have a better idea, go for it!
 * I understand how it's easier for me to provide citations than for you, since I did the research in the first place. For everything else, why are you kibitzing rather than just doing it?  The whole reason for posting the draft here is so that people can edit it!
 * 71.41.210.146 (talk) 16:12, 22 November 2011 (UTC)
 * P.S. You're quite right, the Sagnac effect was only considered in v2 of the paper (the second analysis), but it was applied to both data sets. While I am giving some history of the data collection (to explain the two bunch sizes), I'm not bothering to explain the history of the analysis; I'm using only the best/latest.  It didn't change the results much, anyway.
 * My idea behind asking you to post it here was to see if anybody else objected. Nobody has, so I would say add it to the article, and we will edit there. Ajoykt (talk) 19:56, 22 November 2011 (UTC)


 * I've included a "blind analysis" section above. But the question remains: Where should all of this be placed in the article? --D.H (talk) 11:30, 2 December 2011 (UTC)


 * Hmm, any other comments? --D.H (talk) 18:12, 7 December 2011 (UTC)

Why anti-neutrinos from SN 1987A?
I am confused about why SN 1987A would be expected to emit anti-neutrinos. I thought that the main reaction taking place in stars (including supernova which generate neutron stars) was
 * $$p + e^- \to n + \nu \,$$

which yields a normal neutrino, not an anti-neutrino. Or it is a question of how the particles are detected? JRSpriggs (talk) 01:31, 8 December 2011 (UTC)
 * The detectors detected antineutrinos - I assume they weren't set up to detect neutrinos. Ajoykt (talk) 03:03, 8 December 2011 (UTC)

Order
The article focuses too much on the superseded September results, while the current values (internal replication) are mentioned too late. I think the order should be:
 * First results
 * Internal replication
 * The measurement
 * The analysis

This is because the measurement and analysis sections rely on the numbers of the refined analysis of November.--D.H (talk) 19:46, 7 December 2011 (UTC)
 * As of now, the Internal replication starts with a reference to the GPS-sync and proton-spill issues. That will have to change with the new order. We could either write this chronologically, like a story starting from the beginning, or in order of importance - the second result much earlier. Ajoykt (talk) 20:11, 7 December 2011 (UTC)


 * Right. Presently the article’s sectioning is incoherent. We have something called “The second run” down at the bottom of the section already discussion “Internal replication” in depth. I imagine everyone would consider the first run to be largely obsolete, so it makes sense to note the salient details of the first run but then discuss the second run exclusively. Strebe (talk) 20:40, 7 December 2011 (UTC)
 * Note that the current value of 57.8 ns (refined analysis of November) is based on the main analysis, while the "second run" (62.1) was only to exclude some possible statistical errors.--D.H (talk) 20:56, 7 December 2011 (UTC)
 * Hi, I've added "An alternative analysis of this internal replication in November" to the image description to differentiate this analysis result (62.1) from the main analysis result (57.8). Thanks,  vis  uall  13:51, 8 December 2011 (UTC)
 * The image description has been changed many times:
 * "Result of the OPERA experiment rerun in November"(Ajoykt) --> "An alternative analysis of this internal replication in November"(visuall) --> "Analysis of this internal replication in November"(Ajoykt) --> "Another analysis of this internal replication in November"(visuall) --> "Analysis of this internal replication in November"(129.210.115.6). Thanks,  vis  uall  17:14, 9 December 2011 (UTC)

About the math regarding one in forty thousand.
This was not appropriate. Given that (v-c)/c = (2.48 ± 0.28 (stat.) ± 0.30 (sys.))/100000 you must take only the significant digits given and the error which gives a range of values between 1 in (1/2.76)*100000 = 36,200 and 1 in (1/2.20)*100000 = 44,500. Also (1/2.48)*100000 for the three significant digits stated gives, at best, a rounded three digit precision of 1 in 40300, which lies between range of 1 in 36200 and 1 in 44500. I see that there is now a different statistic that does not seem as meaningful as the original. --Modocc (talk) 20:35, 8 December 2011 (UTC)
 * I added the new wording to make the result more understandable, rather than more precise (1/40,000 is exactly 25 parts per million). I am ok with somebody changing that back to 1/40,000 (which was my original version). I don't think it a good idea to add 1/4xxxx.yy, partly because the implied precision is not accurate, and partly because the non-math-oriented would like round numbers where the difference from an exact calculation is not significant. Ajoykt (talk) 20:41, 8 December 2011 (UTC)
 * I'm somewhat ambivalent regarding which of the two statistics to use, but larger numbers are usually more often harder for people to grasp, thus I prefer the one in forty thousand. In any case, I see that even with your change, unfortunately, there was a reversion to the overly precise number, but now that this matter is brought to the talk page, maybe the edit-warring will stop. --Modocc (talk) 21:26, 8 December 2011 (UTC)
 * This form is required by Ajoykt, but it's just like a repeat of 2.48, which has already been stated in that paragraph.
 * I've seen the form 1/40322.58 is being used on some related wiki pages, such as "approximately 1 in 40,322.58" ("en:Faster-than-light"), "およそ40322.58分の1" ("ja:光速") and etc. Thanks,  vis  uall  21:57, 8 December 2011 (UTC)
 * Ajoykt explicitly said the 25 in a million is not required. Also see wp:otherstuffexists regarding the other entries. Third, I had other things to do earlier or otherwise I would have brought this to this talkpage sooner for discussion, which is what should have happened anyway instead of wp:editwars which can result in warnings and blocks. Thus, I'll wait until there has been adequate discussion here before I would restate the figure to "one part per 40,000" instead of "25 parts per million". --Modocc (talk) 22:24, 8 December 2011 (UTC)
 * I am not quite sure what "required by [me]" means? I prefer 25 ppm, since the ppm measure is widely used for small quantities, but I am fine with 1/40,000 (if somebody changes, I won't revert). I don't think 2.48 x 10^-5 is understandable to most people. We need to remember while those who edit this page largely have strong connections to math, those who read this page mostly don't. Ajoykt (talk) 22:59, 8 December 2011 (UTC)
 * I just thought that "1 in 40,322.58", "1 in 40,322" or "1 in 40,000" is more straightforward than "25 in 1,000,000", but you always "required it as the latter" rather than either of the former while editing several hours ago. Since you've agreed to change it to the former, I'm fine with either of them (the former). Thanks,  vis  uall  23:27, 8 December 2011 (UTC)
 * I've changed the value from "25 parts per million" to "one part per 40,000", which is preferred by Modocc. Thanks,  vis  uall  00:31, 9 December 2011 (UTC)
 * Not to keep harping on what is a closed issue, but isn't this more or less what we started with? When somebody reverts your "new" change, please take it to the talk page, instead of just insisting the change stay. 129.210.115.6 (talk) 01:41, 9 December 2011 (UTC)

Overview of results
I've removed all detailed numbers from the first two sections, and included a separate one called "Overview of results". I think this is better understandable for the general reader. --D.H (talk) 14:58, 9 December 2011 (UTC)
 * Hi, the phrase "second run" is used on that table, but there's no "second run", it's just performed with a dedicated CNGS beam as a part of this run.
 * I just think a "second run" means a restart of the whole detector, but in this case, it just deploys an additional test to the current run. If this definition is incorrect, please point it out.
 * Anyway thanks for your great work, it now looks more clear than before.
 * I often see some significant improvement you made... Thanks,  vis  uall  16:23, 9 December 2011 (UTC)
 * They write "a test was performed with a dedicated CNGS beam" and "Running with the CNGS bunched beam" (p. 26). So I think we can call it a "second run". However, I've changed the name into "CNGS bunched beam" which they actually used. --D.H (talk) 17:32, 9 December 2011 (UTC)


 * Thanks, D.H. This is much clearer. Strebe (talk) 19:37, 9 December 2011 (UTC)

File:CNGS_layout.jpg slated for deletion based on the wrong premises
The argument is the image is just a map, and could be reproduced in an hour with good drawing software. For one thing, the image is of underground structures, details of which are known just to the architects of CERN (the public is not allowed in there). For another, no, I don't think the image is all that easy to reproduce from scratch. Note that the copyright arguments are moot, since we are asserting the fair-use rationale (a small excerpt, needed for our article, does not hinder any future commercial use by owner, already published elsewhere, and no replacement available). Would be good to have more people chime in: Ajoykt (talk) 18:19, 8 December 2011 (UTC)
 * Carnildo has acknowledged that the image could not be recreated without official support, therefore there's no cause for deletion.. Thanks,  vis  uall  23:32, 8 December 2011 (UTC)
 * Is that private communication? I don't see anything on the deletion page itself. Ajoykt (talk) 01:54, 9 December 2011 (UTC)
 * Firstly, Carnildo said that an official map was needed for reference yesterday. Didn't you see that?
 * Secondly, why did you insert your reply into the front of my reply? Usually those replies are ordered by post time. Thanks,  vis  uall  16:47, 9 December 2011 (UTC)
 * Carnildo asked for an "overhead" map, not an official map. I don't think Carnildo's reply indicates agreement with retaining the image. As to the formatting, go ahead and rearrange it to how it is supposed to be. Ajoykt (talk) 19:39, 9 December 2011 (UTC)
 * But nobody has the map except for official people. Thanks,  vis  uall  19:48, 9 December 2011 (UTC)
 * Yes, but these are things you need to mention on that page. Please don't assume the discussion there is done - Carnildo's last post certainly doesn't mean Carnildo takes your inference for granted. Ajoykt (talk) 19:53, 9 December 2011 (UTC)
 * Hi Ajoykt, for another image - "File:OPERA analysis.png" - I'm not sure why I still got this message "This image does not have a copyright tag". - I've added both "Non-free with permission" tag and "Non-free image data" tag... Thanks,  vis  uall  20:33, 9 December 2011 (UTC)
 * Hi Ajoykt, both File:CNGS_layout.jpg and File:OPERA_experiment.png have been deleted by Fastily at 0:08, with this ill-founded cause. Thanks,  vis  uall  01:00, 10 December 2011
 * You should try the process in WP:RFC/U. It is ridiculous to claim underground structures can be reproduced from overhead maps. Ajoykt (talk) 01:59, 10 December 2011 (UTC)
 * Hi Ajoykt, both File:CNGS.jpg and File:OPERA.png have been deleted by Carnildo with the cause "Recreation of deleted material". I've changed File:CNGS.jpg to File:Cern1.jpg. Thanks,  vis  uall  09:27, 11 December 2011 (UTC)
 * Can you go and update my complaint about the image deletion: http://en.wikipedia.org/wiki/Wikipedia:Requests_for_comment/Ajoykt. You need to endorse it somewhere or the other, and also add excerpts from the deletion talk page. 00:38, 12 December 2011 (UTC) — Preceding unsigned comment added by Ajoykt (talk • contribs)
 * The actual place you want to be is Deletion review. WP:RFC/U is aimed more at malicious misconduct, not disagreements over whether actions are appropriate.  Deletion Review will get a result one way or another in about a week; Request for Comment, at best, will take at least a month, and in this case I can guarantee you that the result will be "his actions were defensible, take it up at deletion review".  (As an aside, the convention at RFC/U is to name the page after the user the complaint is about&mdash;Requests for comment/Ajoykt will suggest misconduct on your part, not Carnildo's, to those reading it.) 74.74.150.139 (talk) 00:50, 12 December 2011 (UTC)
 * Hi, I've created the "deletion review" for the above CNGS images. Thanks,  vis  uall  07:27, 12 December 2011 (UTC)
 * Can you post a link? Ajoykt (talk) 16:54, 12 December 2011 (UTC)
 * Hi Ajoykt, the page link is Wikipedia:Deletion_review. Thanks,  vis  uall  03:16, 13 December 2011 (UTC)

About images
Aligning all of them on the right doesn't look very well, because the images appear in sections where they don't belong, at least on my monitor. I also think that we shouldn't include more than one image per section. --D.H (talk) 09:30, 10 December 2011 (UTC)
 * Hi Ajoykt, the pdf image looks somewhat blurry, could you upload a clear image instead? The original image in the paper is clear enough. Thanks,  vis  uall  00:53, 11 December 2011 (UTC)
 * Hi Ajoykt, I've uploaded a more clear version of this image. Thanks,  vis  uall  02:36, 11 December 2011 (UTC)

Chodos's statement on MINOS hint of CPT violation
The statement was made in December, and clearly, he would have known of the August MINOS results on the neutrino/anti-neutrino oscillation-mass-difference signal. Is he referring to other unpublished results from MINOS? How exactly do we link his statement to this specific MINOS finding? Ajoykt (talk) 17:38, 14 December 2011 (UTC)
 * Hmm.. Unfortunately he gives no references, and note that this stuff is extremely speculative and not accepted at all. So let's try to formulate it as general as possible. --D.H (talk) 17:55, 14 December 2011 (UTC)
 * BTW, their is no contradiction with Chodos's statement, since the mass difference is still not fully explained, though it is now much smaller than before, which reduces its probability.... --D.H (talk) 19:09, 14 December 2011 (UTC)

Connection to Superluminal Quantum Tunnelling and the 10^{-19} Figure.
I notice that there is no discussion in the article (nor in the OPERA paper, itself) comparing and connecting the OPERA results to the existing literature on superluminal quantum tunneling (c.f., the references under the Quantum tunnelling article and ). Nor is there any further elaboration on the origin of the 10^{-19} bound mentioned in the article (nor any to be found in the paper). This would be especially useful to see in light of superluminal quantum tunneling results, where no such bound related to finite rest mass is involved. — Preceding unsigned comment added by 64.136.26.17 (talk) 22:30, 14 December 2011 (UTC)

Bunched Beam Rerun Distribution Unimodal or Bimodal?
There should be some comment under the "Internal replication" section relating to the poor goodness of fit the N = 20 distribution depicted to a unimodal distribution. What's shown there (based on a careful analysis of the graph, itself) shows a small 3rd Cumulant and large negative 4th Cumulant -- the signature of a Bimodal distribution with a 50-50 mixing ratio. Depending on the placement of the lower mode (based on the 4th Cumulant the mode separations are about 15-20 ns away from the 62.1 ns mean cited), the distribution cited for the follow-up test might actually be decreasing the significance of the OPERA results, while increasing the likelihood of a two-fluid dynamics characteristic of particle oscillation. There should be some mention of the issue, along with pointers to related cites in the literature. Note the warning under the Bimodal distribution article: &ldquo;Bimodal distributions are a commonly-used example of how summary statistics such as the mean, median, and standard deviation can be deceptive when used on an arbitrary distribution.&rdquo; — Preceding unsigned comment added by 64.136.26.17 (talk) 22:53, 14 December 2011 (UTC)
 * These are issues you can discuss at arXiv (www.arxiv.org). Here we collect information, we don't analyze it. Ajoykt (talk) 23:09, 14 December 2011 (UTC)

Back in 2010, CNGS was scheduled to be shut down the whole of 2012
Quote from last year []:
 * The cuts will delay the entire Cern research programme, director general Rolf Heuer told staff on 31 August. Funding for general R&D will be reduced and the 74 experiments outside Cern’s flagship Large Hadron Collider delayed, he said. For example, experiments such as the CNGS project, a neutrino beam fired from Cern to Gran Sasso in Italy, will collect no data in 2012 when all accelerators will shut down and efforts be redirected to the LHC upgrade.

Looks like now it is going to reopen in March 2012; they are taking the result really seriously. Ajoykt (talk) 06:03, 20 December 2011 (UTC)


 * What maintenance work are they omitting? Does it affect the search for the Higgs boson, or what? JRSpriggs (talk) 17:59, 20 December 2011 (UTC)
 * Hmmm, looks like the LHC upgrade was moved from 2012 to 2013 pre-OPERA. The LHC will get to full-power (7 TeV per beam) only in 2014, http://press.web.cern.ch/press/PressReleases/Releases2011/PR01.11E.html, many years behind schedule. I thought the 3.5 TeV beams were sufficient for the Higgs boson search-hence the decision to keep going with the same beam energy-but am not sure. Note that on the OPERA replication front, things are kind of stuck, nobody seems to know how to avoid using GPS to time the neutrinos. So, end-2012 we might still be in not-confirmed land, even if MINOS repeats using common-view GPS time transfer. I have a hunch CERN will have to put off its LHC upgrade again, keeping CNGS going. Ajoykt (talk) 01:41, 21 December 2011 (UTC)

IceCube results
Another analysis has become available regarding FTL neutrinos. --Eleassar my talk 17:08, 23 December 2011 (UTC)
 * PRL is bent on publishing "theoretical refutations" of an experiment. Oh well. This objection has been around for a while; I guess the authors decided to give it publicity. Since PRL is no easy journal to get published in, I think a separate subsection is called for (this doesn't fit into the C-G subsection). We will have to wait for somebody to point out the obvious problem - this is another case of theoreticians claiming an observed fact can't be because it doesn't fit their theory.

All three papers on OPERA in PRL are theoetical refutations: Cohen-Glashow, this paper, and Xiao-Jun Bi et al. All of them basically say just observing the energy of the neutrino tells us how fast it can go. I think PRL's editorial board has a strong stance on the OPERA results; none of the refutations are presented as what would have to change if the OPERA results hold good. Ajoykt (talk) 19:11, 23 December 2011 (UTC)


 * I don’t think the authors of these papers consider them to be “refutations”; nor do I think PRL’s editorial board has any “stance” on OPERA results a priori. These papers point out theoretical difficulties in taking the OPERA results at face value. That is important because:
 * They point out which theories will have to be modified if OPERA is confirmed unambiguously;
 * Since the effects these theories describe are not “just” theories, but represent phenomena already verified or are rigorous extrapolations of theory that describe phenomena already verified (as opposed to hypothetical blathering), in the presence of constrained resources they can help guide efforts to figure out what’s wrong with OPERA (if something is wrong).
 * Similarly, once OPERA results are confirmed to some degree of satisfaction (if that were to happen), these papers can help guide guide and prune theories needed to explain the results.
 * Theoretical objections are important; they help funnel efforts into directions most likely to pay off. Obviously no theory trumps fact, but such either-or thinking rarely means much useful. The OPERA experience deploys a bogglingly complex apparatus and a bogglingly complex sequence of inferences to reach the conclusion that neutrinos have traveled faster than light. The results are no Michelson-Morley moment until the accuracy of the apparatus itself is confirmed to as high a degree as a simple interferometer… and that will be a long time in coming. There simply isn’t any one person who has the expertise to evaluate the entire experiment. Where is the massive team going to come from? Strebe (talk) 21:47, 23 December 2011 (UTC)


 * In addition, those papers were not written in defense of standard physics. On the contrary, they all predict Lorentz violations, i.e. breakdowns of special relativity, and by mainstream standards (at least a short time ago) those models have been considered "exotic". Now, when even those exotic models disagree with the OPERA result, one can imagine how extreme the Lorentz violating framework must be, that possibly might explain the result (some of them have been already published on arxiv etc.) For instance, current bounds for superluminality of other particles lie between $$10^{-15}$$ and $$10^{-25}$$, while the OPERA result gives a comparably high deviation of $${10^-5}$$. --D.H (talk) 22:32, 23 December 2011 (UTC)
 * I am not saying they are defending standard physics; they are defending their own old theories of LIV and what tests to conduct to check those theories. That, of course, is a bias practically everyone has. Ajoykt (talk) 22:54, 23 December 2011 (UTC)
 * Maybe. Actually, putting together an international team to explicitly cross-check the measurements, etc., may be financially viable. Physicists aren't paid all that much, and when traveling do not get CEO-style accommodations.

These theories extrapolate already-verified-phenomena to realms where they may or may not apply. The initial press releases of the scientists omit that fact. Then, slowly, they back off from the initial imperative statements. Note how far C&G have come from their first "We thus refute OPERA" statement. Cowsik now essentially starts from where they started; he doesn't mention any possible loophole, though one of his coauthors already has presented one. The papers don't point to any path forward&mdash;they just plainly state neutrino-velocity can be figured out simply by looking at the energy, and that is that. We already have neutrinos in the TeV range, so per these papers the MINOS replication is meaningless. That is the core practical part&mdash;by printing papers only of this view, PRL is influencing the budgetary allocation away from extra funding to replicate OPERA, and toward programs which continue the existing theoretical framework. This may not be intentional; it probably represents inertia, and a bias among theoretical physicists who have dabbled in LIV toward their earlier work, but the consequences are still real. Ajoykt (talk) 22:56, 23 December 2011 (UTC)
 * Whatever, here (this talk page) is not the place to discuss such issues. We are only reporting.... --D.H (talk) 00:25, 24 December 2011 (UTC)
 * Happy Christmas to all of you. --D.H (talk) 00:31, 24 December 2011 (UTC)

Not inconsistent with Lorentz Invariance
At the end of section OPERA neutrino anomaly, it says "... since the higher-energy OPERA neutrinos are apparently faster than the lower-energy SN 1987A ones, they are not conventionally tachyonic.". This does not take into consideration the different flavors of the neutrinos. If the (anti-)electron neutrinos from the supernova have zero (or much closer to zero) mass and the muon neutrinos have an imaginary mass, then the former would travel at light-speed and the latter would travel faster than light-speed. A mass difference among flavors is expected due to the apparent oscillations between neutrino flavors being studied in the experiment. Thus the result is not inconsistent with Lorentz invariance, right? JRSpriggs (talk) 04:59, 27 December 2011 (UTC)
 * We are quoting a peer-reviewed summary there, so from a encyc. POV, we are ok. But, yes, it does have its assumptions. I think there are other papers which purport to show if one kind of neutrino is superluminal all of them are, but since none of this is tested or observed, I guess largely we have just debates on every front. Nothing can be thought settled until a replication and testing of predictions of the theories. Ajoykt (talk) 05:47, 27 December 2011 (UTC)

Some peer-reviewed sources: Alexandre et al. argue that "neutrino oscillation phenomenology severely constrains differences in the propagation speeds of different neutrino flavours". Also Cohen/Glashow said "Lorentz-violating velocity differences as large as 10^−20 between neutrinos of different species would have been readily detected and are excluded." But who knows how this develops, we have to wait for additional sources. --D.H (talk) 15:50, 27 December 2011 (UTC)

Names
In my view, links to existing articles, especially names, should be highlighted in the main text, not somewhere in the "see also" section. That is the strength of an encyclopedia like WP. --D.H (talk) 17:02, 25 December 2011 (UTC)
 * I think too many links make the article hard to read, with words popping out of the sentence. From WP:MOS, the "Links" section:
 * Make links only where they are relevant and helpful in the context: Excessive use of hyperlinks can be distracting, and may slow the reader down. Redundant links (like the one in the tallest people on Earth) clutter the page and make future maintenance harder. High-value links that are worth pursuing should stand out clearly [emphasis in original].
 * I don't think the biographies of scientists are relevant to understanding the article. Their presence makes it impossible for readers to figure out what the high-value links are, the ones related directly to the article content. Those who really are interested can look at the 'see also' part. Ajoykt (talk) 18:09, 25 December 2011 (UTC)


 * I have not the impression that the current article is overladen with links - especially links to Nobel laureate should be included. Of course, I'm not going to start an edit war because of style issues, see WP:UNDERLINK. Other opinions? --D.H (talk) 18:23, 25 December 2011 (UTC)
 * IMO, it may be useful to try to picture a cross-section of one's target readership, and ask oneself who might find what useful or otherwise. So what's the readership likely to be here? Quite broad, I'd guess... Including a variety of people looking for more systematically presented information than what they get in the media. That was why I felt "geodesy" might conceivably be a helpful wl (unlike "La Sapienza University of Rome", which doesn't potentially advance comprehension, I fully agree). Just my 2 linklets, MistyMorn (talk) 18:57, 25 December 2011 (UTC)


 * I want to see links to anything directly relevant in Wikipedia articles. I’m not that interested in someone else deciding for me what is highly relevant and what isn’t. I have had to fight frivolous linking in a lot of articles, but that wouldn’t be this article. For me, links to Wikipedia pages of scientists engaged in a project are directly relevant. I don’t scan pages looking for links to follow; I follow them in the course of reading the text if I care to, and they don’t distract me if I don’t care to. Strebe (talk) 21:40, 25 December 2011 (UTC)


 * Wikipedia physics articles, in general, have many links. But WP:MOSLINK is clear in its guidelines: links are supposed to "increase reader's understanding of the topic", and "deepen understanding of a topic", and should "[be] needed to aid the understanding of the article", and "[be] relevant to the article in question". The MOS also advises considering placing links in tables, figures and section openings.

Looking at other online encyclopedias helps. Britannica is speckled with links, but their articles are typically just a few paras. The Stanford Encyclopedia of Philosophy rarely uses links (it has always been online-only, and is highly credible). See, for example, http://plato.stanford.edu/entries/logic-ai/ on Logic and AI. The Columbia encyclopedia uses some links, typically directly related to the article. See http://www.encyclopedia.com/topic/elementary_particles_%28universe%29.aspx#1-1E1:elementr-p-full for an article on "elementary particles". The CIA world factbook rarely has links (an example: https://www.cia.gov/library/publications/the-world-factbook/geos/us.html). The Canadian Encyclopedia uses a few scattered links (http://www.thecanadianencyclopedia.com/index.cfm?PgNm=TCE&Params=A1ARTA0007784). Our article is fairly long and likely to get longer. People would want to read it end to end. Links distract from that. Note that somebody like George Smoot III is not linked to OPERA or related experiments in any way; we include him because he is a Nobel winner physicist (cosmologist). How his biography helps a reader understand the "neutrino opera anomaly" better is questionable. That applies to most biographies. I think we who are used to reading Wikipedia articles find nothing wrong with text having a-word-a-sentence in embossed blue. Most others would. Newspaper and science magazine articles published online have such words only sparingly. -- Ajoykt (Dec 28, 2011, 12:05 PST)


 * Compared to other articles, it isn't overladen with links, see for instance the current featured article Peace dollar, or other Featured articles like 7 World Trade Center, Millennium Park etc. As far as I can see, all of them link to the biographies (and much more) in the main text. We should look at our featured articles, not other encyclopedias, websites, or newspapers which have a very different article structure..
 * PS: The guidelines of WP:MOSLINK don't tell us which links are important or not for this article, this must be decided by us. I think the biographies are important because they tell us something about the credibility of the person and thus his opinion. --D.H (talk) 21:39, 28 December 2011 (UTC)


 * Not only that, reading the WP:OVERLINK section, it is clear that the links in question come nowhere near the examples advised against, and reading the WP:UNDERLINK section shows examples that seem very much in line with the proposed links. My opinion is that Ajoykt interprets WP:MOSLINK excessively conservatively. Strebe (talk)

Article Title
Currently a Google search for "faster than light neutrinos", after clearing all history, returns, on the second page, a link to the WP "Neutrino" article. This article isn't in the picture. I think the problem is nobody refers to this as the 'anomaly', most people think of it as the 'OPERA faster than light neutrinos'. Ajoykt (talk) 21:59, 28 December 2011 (UTC)

Images
The diagrams seem nice and illustrative, but as a nitpick: I want to recommend against using outline, emboss, curve and other MS Publisher font effects - they make the otherwise good image look like a birthday invitation.--Anders Feder (talk) 22:42, 28 December 2011 (UTC)

Reactions?
The article does not make entirely clear which subatomic reactions produce and detect the muon neutrinos. My guess is
 * $$ \pi^{+} \to \mu^{+} + \nu \,$$
 * $$ \nu + n \to p + \mu^{-} \,.$$

Is this correct? JRSpriggs (talk) 20:11, 28 December 2011 (UTC)
 * For the first, we do mention in the lead-image-caption and later in the text that the CNGS beam is produced by pions decaying into neutrinos (some kaons do too, but that doesn't seem material). For the reception side, we mention the detection is that of light converted to charge (The measurement:Overview), though the reference is highly indirect. The details should really be in the OPERA experiment article, since the neutrino-velocity paper (arXiv 1109.4897) refers to the original OPERA experiment paper for this information. Ajoykt (talk)


 * The OPERA experiment article does not seem to provide this information either.
 * Are you saying that the detection (even for purposes of observing FTL rather than oscillation) is given by
 * $$ \nu + n \to p + \tau^{-} \,?$$
 * JRSpriggs (talk) 05:43, 29 December 2011 (UTC)
 * The OPERA detector is set up to detect all kinds of neutrinos. They then use a complex software algorithm to filter out only tau events based on the exact track left in the brick and so on (http://arxiv.org/abs/1006.1623, section 4). For the velocity measurement, I guess they didn't care; the CNGS beam is almost completely muon-neutrinos anyway, and a few electron neutrinos wouldn't matter (tau-n would be unlikely). So the detection is given by both your equations there, plus a similar equation for the electron neutrino. Yes, this stuff should be in the OPERA experiment, which is right now just a stub. Documenting the work of almost 200 physicists over 3 years requires at least two articles. Ajoykt (talk) 06:52, 29 December 2011 (UTC)

Requested move
<div class="boilerplate" style="background-color: #efe; margin: 2em 0 0 0; padding: 0 10px 0 10px; border: 1px dotted #aaa;">
 * The following discussion is an archived discussion of a requested move. Please do not modify it. Subsequent comments should be made in a new section on the talk page. No further edits should be made to this section. 

The result of the move request was: page moved. Vegaswikian (talk) 03:28, 5 January 2012 (UTC)

OPERA neutrino anomaly → OPERA faster-than-light neutrinos – Most people refer to this finding as the "faster-than-light neutrino" finding. Google searches for that term right now show up another WP article ("Neutrino"), which is less relevant than this one. A link does not help; the Google crawler is ignoring it. Our references (Nature News, Science, Scientific American, New Scientist, The Guardian, NY Times, and even the more-for-physicists EuCARD) all title their articles "faster-than-light neutrinos." The word "anomaly" is used only by us, the original CERN press release, and by some published physics papers including the OPERA arXiv eprint. The word has not caught on, unlike the "pioneer anomaly". Ajoykt (talk) 00:08, 29 December 2011 (UTC)


 * I would be for the new title if it were accurate. It’s not. Part of the reason Google Search doesn’t like the article is because of its rapid pace of change. I would favor a redirect to this page from the proposed title, though. That would also offer a cue to Google Search. Strebe (talk) 00:54, 29 December 2011 (UTC)
 * The redirect already exists and does nothing. The reason Google search doesn't work is because the phrase searched for isn't there in the article, at least nowhere prominent. The question is how people refer to the term; most, including the references we cite, use that phrase. We can't possibly claim to be more accurate than our sources. Here are our first 6 sources (excluding pre-OPERA ones):


 * Eugenie Samuel Reich (November 18, 2011). "Neutrino experiment replicates faster-than-light finding". Nature News.
 * Adrian Cho (September 30, 2011). "From Geneva to Italy Faster than a Speeding Photon?". Science.
 * Charles Q. Choi (October 13, 2011). "Leading Light: What Would Faster-Than-Light Neutrinos Mean for Physics". Scientific American.
 * Eugenie Samuel Reich. "Finding puts brakes on faster-than-light neutrinos". Nature News.
 * Edwin Cartlidge (November 17, 2011). "Faster-Than-Light Neutrinos: OPERA Confirms and Submits Results, But Unease Remains". Science.
 * Alok Jha (November 17, 2011). "Neutrinos still faster than light in latest version of experiment". The Guardian.


 * That is not a biased sample - all outlets use faster-than-light neutrinos, dropping the "apparently" part. It is common usage now. Nobody uses the word "anomaly". It is too formal, stilted, and not widely recognized. The title doesn't mean OPERA really has seen faster-than-light neutrinos; it just means what OPERA reported is now widely referred to as faster-light-neutrinos. More often than not, the reference is in skeptical tones, but the phrase is still what is tied to the concept in both the public mind and in the media world. Ajoykt (talk) 01:38, 29 December 2011 (UTC)


 * What do you mean “does nothing”? The “faster than light neutrinos” redirect works. Do you mean the redirect does not affect PageRank? Also, we do not have a “faster-than-light neutrinos” redirect. Bing search, altavista, Yandex search yields this article before the Neutrino article. DuckDuckGo lists this article as its first result. Baidu search turns up this article as its 7th entry on the front page, before any other Wikipedia articles. Catering article titles for specific search engines doesn’t seem defensible.


 * We can't possibly claim to be more accurate than our sources. Some of our lay literature sources say “faster-than-light neutrinos” in the title? Not persuasive. Headlines aren’t reliable sources; they’re sales devices. The articles themselves recognize the finding as very tentative. Titling this article as proposed gives the impression of scientific fact to the results of the experiment. Strebe (talk) 01:49, 29 December 2011 (UTC)
 * Titles refer to things. Most people refer to this as the "OPERA faster-than-light neutrino" with the "apparent" understood. Nature News isn't exactly a lay-literature source; science-news outlets do not use the word "OPERA neutrino anomaly". I don't think titles, just like headlines, are meant to be a synopsis of the article; they are meant to lead readers/searchers to what they are looking for. Our current title doesn't serve that basic purpose. Ajoykt (talk) 01:57, 29 December 2011 (UTC)


 * You have no evidence that the title doesn’t serve its purpose. You have Google search results for a search query that you have hypothesized is a (the most?) likely search query. Strebe (talk) 02:10, 29 December 2011 (UTC)
 * That hypothesis is based on how the media, both general and science-news mags, are reporting this. There are other indications: check the page-view statistics for the "neutrino" WP article. It used to be less than 1k before Sep 22; on Sep 23, it jumped to 102.9k. On November 18, they jumped from 2.7k to 17.5k; on the very same day ours jumped to 3.4k (that was the day the OPERA second, rerun, results came out). So 11k people were interested in reading about neutrinos but not about the result itself? Check the patterns in http://stats.grok.se/en/201111/OPERA_neutrino_anomaly and http://stats.grok.se/en/201112/Neutrino. Way too many people are winding up in the "Neutrino" article on the days OPERA-related news comes out. Their average hit rates have gone up way higher than ours since Sep. 22.
 * I don't think "OPERA faster-than-light neutrinos" should be read as a sentence meaning OPERA saw faster-than-light neutrinos. It is now a noun phrase lent a specific meaning by the media coverage, and that meaning is just "The OPERA results of Sep. 23". Going through our own references shows how much more prevalent that phrase is than the clunkier "OPERA neutrino anomaly". Ajoykt (talk) 02:36, 29 December 2011 (UTC)


 * Your assumptions and theses:
 * People most commonly search on "faster than light neutrinos".
 * People most commonly search on Google.
 * Retitling the article would convince Google to rank the article higher.
 * We should retitle articles in order to conform to Google specifically.
 * People will choose to visit the article if Google ranks it more highly.
 * People don’t want to know about neutrinos when they read a lay article on the anomaly; they want more information about the anomaly itself because the article they read doesn’t suffice.
 * Driving traffic to the article is worth changing the title from one that is unambiguously accurate to one that allows people to interpret it as meaning the OPERA findings are accepted.


 * I think (1) is just one common search query. Others would be "opera neutrinos” (still doesn’t rank on first page despite your theory about the importance of title), "superluminal neutrinos" (no worse than "faster than light neutrinos), "relativity neutrinos" (fail), "CERN neutrinos" (no worse) &c. (2), granted. (3) Doubtful. (4) I’m very much opposed. (5) Probably, but the problem with that theory is that Google likes a bazillion articles on this topic, and that’s the real problem. (6) I’m skeptical. (7) I’m opposed. Strebe (talk) 03:26, 29 December 2011 (UTC)
 * The title matters for Google searches all right: we rank first for "OPERA neutrino anomaly", second for "OPERA neutrino", second for "opera anomaly", second for "neutrino anomaly", first for "anomaly opera" and second for "anomaly neutrino". We should title articles to allow readers to find the article. I don't quite see how the phrase "OPERA faster-than-light neutrinos" means the results are accepted. Everybody else is using that phrase to mean something else. The accepted usage by now is different from your interpretation. Ajoykt (talk) 04:19, 29 December 2011 (UTC)


 * No, that still doesn’t follow. This article ranks highly on a Google search that includes “anomaly” because few other articles use it, not because the title has necessarily awed Google into ranking the article highly. The article uses the term “faster-than-light” in close proximity to “neutrino” 17 times. It directly uses the term “faster-than-light neutrino”(s) 12 times. The idea that Google doesn’t “realize” the article is about faster-than-light neutrinos doesn’t hold water. Google ranks the article poorly even for phrases that it contains directly that are uncommon and unimportant in other articles that it ranks more highly, such “Lorentz-violating models”. It ranks it highly for others, such as “OPERA collaboration results” and “Cohen-Glashow effect”. I think you are pinning way too much on the assumption that changing the title will drive traffic to the article. Too many assumptions. Strebe (talk) 04:51, 29 December 2011 (UTC)
 * We rank second for "OPERA neutrino" which doesn't have 'anomaly' in it. Actually "neutrino anomaly" is supposed to refer to something else: the solar neutrino anomaly. Also, we don't rank high for variants like "neutri anomaly". Yes, the article uses "faster-than-light" in the text, and my whole argument is that it is not enough, we need it in the title. Ajoykt (talk) 06:02, 29 December 2011 (UTC)


 * Oppose The proposed title looks like OPERA produced the neutrinos, instead of detecting them. Perhaps OPERA faster-than-light neutrino anomaly would solve your problem? 76.65.128.132 (talk) 04:17, 29 December 2011 (UTC)
 * Yes, I believe so. Thanks for the suggestion. Ajoykt (talk) 04:20, 29 December 2011 (UTC)


 * I considered that, but it’s very long and ungainly, so I didn’t propose it. Plus, I have no faith it will increase page views. Strebe (talk) 04:52, 29 December 2011 (UTC)


 * Support inserting 'faster-than-light' (or a suitable equivalent) somewhere in the title. To my knowledge, titles are important for SEO as they are weighted in algorithms like Google's, based on content assumptions. We've got 'neutrino' in the title but lack any overt reference to the speed of light. IMO, we need that. I wouldn't be altogether comfortable with OPERA faster-than-light neutrinos for the reasons stated above. OPERA faster-than-light neutrino anomaly is better conceptually, but I agree that it's a bit of a mouthful. Detection of faster-than-light neutrinos by OPERA?
 * Support. "Faster-than-light neutrino" is the idiom used by our sources and by our readers. We summarize our sources as opposed to judging them. Kauffner (talk) 10:58, 29 December 2011 (UTC)
 * Suggest Faster-than-light neutrino -- I am not sure why OPERA needs to be in the title. The article anticipates experiments to replicate the result.  Once these have taken place, it will no longer be an OPERA result.  Peterkingiron (talk) 15:40, 29 December 2011 (UTC)
 * Comment: OPERA faster-than-light neutrino anomaly seems acceptable. Faster-than-light neutrino is not so good, because as of yet this is only an OPERA result awaiting independent confirmation - if confirmed Faster-than-light neutrino should be the name of a new article, summarizing all experiments and theoretical considerations. --D.H (talk) 16:29, 29 December 2011 (UTC)
 * Support I have always seen sources talking about FTL neutrinos, not about "anomalies". And please let's not worry so much about tweaking titles to fit search engine preferences. --Enric Naval (talk) 23:04, 29 December 2011 (UTC)
 * Oppose – the "faster-than-light" thing is still an anomalous finding that pretty much nobody believes is going to be correct in the long run. For now, it's an anomaly, and the best sources say so.  The more popular sources often say do, too, just not in their headlines.  Alternatively, consider Faster-than-light neutrino anomaly. Dicklyon (talk) 05:31, 30 December 2011 (UTC)
 * Comment:I agree with "Faster-than-light neutrino anomaly." The main issue is the article is really about the FTL part&mdash;right, wrong, real, apparent, whatever, the frame of reference is still the FTL part; the notability is not from OPERA, neutrino or anomaly, it is from FTL. Note that physics papers all use the equivalent "superluminal neutrinos". (To D.H.: if confirmed we could change the name here to faster-than-light OPERA experiment or something like that, and create whatever other articles need to be then; since the anomaly part would have to be removed, renaming would be needed anyway). Ajoykt (talk) 04:01, 31 December 2011 (UTC)
 * I get 49 Google scholars results for "faster-than-light neutrinos". There is an article entitled "Testing the OPERA superluminal neutrino anomaly". But otherwise the word "anomaly" does not seem to be part of RS descriptions of this experiment. I think it injects pro-relativity POV without conveying additional information. Kauffner (talk) 10:46, 31 December 2011 (UTC)
 * OPERA themselves call it an anomaly in both their paper and the press release]. GoogleScholar gives over 70 hits for anomaly + superluminal + opera (limited to 2011) . 16 hits of "OPERA anomaly", 19 hits for "velocity anomaly" + opera. So the word anomaly is scientifically justified in the title (at least as of yet). There are also other neutrino anomalies such as "LSND anomaly", "MINOS anomaly", "Miniboone anomaly" etc. --D.H (talk) 11:22, 31 December 2011 (UTC)


 * There should be redirects from all of the redlinks brought up here as possible titles, of course. 74.74.150.139 (talk) 00:02, 1 January 2012 (UTC)

Votes
Some discussed choices (please sign below the chosen entry using #):

Faster-than-light neutrinos

 * 1) Kauffner (talk) 02:01, 1 January 2012 (UTC)

Faster-than-light neutrino anomaly

 * 1) --D.H (talk) 12:19, 31 December 2011 (UTC)
 * 2) Ajoykt (talk) 16:08, 31 December 2011 (UTC)
 * 3) Strebe (talk) 23:50, 31 December 2011 (UTC) (With stipulation that article’s scope is more than just OPERA)
 * 4) Dicklyon (talk) 02:08, 1 January 2012 (UTC)
 * 5)  MistyMorn (talk) 10:47, 1 January 2012 (UTC) (with the understanding that the scope issue will probably need to be addressed later, as appropriate)
 * The above discussion is preserved as an archive of a requested move. Please do not modify it. Subsequent comments should be made in a new section on this talk page. No further edits should be made to this section.

Typo To Correct
The article quotes 2.43 microseconds as the commute time, ie 2,430 nanoseconds. It should be 2.43 milliseconds, ie 2.43 million nanoseconds. Light travels about 0.3 m per ns, so 2.43 M ns is about 730 km, which is about right for the commute distance. — Preceding unsigned comment added by Trilithium (talk • contribs) 18:59, 1 January 2012 (UTC)
 * Thanks for spotting it. Feel free to fix it; I don't think this requires a discussion (Fig 1. has the right value - 2.439 msec). Also, typically the convention here has been to add new "talk" topics at the end, though I have no logical explanation as to why. Ajoykt (talk) 19:24, 1 January 2012 (UTC)

Lost archives
When moved this article and talk page from "OPERA neutrino anomaly" to "Faster-than-light neutrino anomaly", it appears that he failed to also move the archives of this talk page. JRSpriggs (talk) 07:11, 5 January 2012 (UTC)
 * Corrected. --D.H (talk) 09:43, 5 January 2012 (UTC)
 * User:Fastily has gone ahead and deleted our archives 1 and 2. Not sure why. Ajoykt (talk) 04:06, 14 January 2012 (UTC)

Any confirmation of an independent replication by summer?
http://www.runcornandwidnesweeklynews.co.uk/runcorn-widnes-news/runcorn-widnes-local-news/2012/02/09/large-hadron-collider-boss-predicts-faster-than-light-neutrino-myster-to-be-solved-by-end-of-summer-55368-30291218/

Ex-LHC boss: "I’m working on an independent experiment that should start in a few months and we should have the results by the end of summer." Ajoykt (talk) 00:54, 10 February 2012 (UTC)

ETRF2000 is not inertial
In the subsection Faster-than-light neutrino anomaly, it says that the distance is measured relative to the European Terrestrial Reference System 1989 (ETRF2000). However, ETRF2000 is not an inertial frame of reference since it is not in free-fall and it is rotating. Should the article not explain why this is not a problem (or is it?), since special relativity is supposed to apply only to inertial frames of reference? JRSpriggs (talk) 04:36, 10 January 2012 (UTC)
 * The article doesn't go into the details of what corrections the distance survey used, largely because neither secondary sources nor even the arXiv paper report them, and that level of detail just may not be appropriate for an encyclopedic article. The details are there in the references, the geodesic survey reference in particular. For your question, I think they used ETRF2000 to eliminate problems with worldwide coordinates. Worldwide coordinates change with continental drift. In this case, since both Gran Sasso and CERN are on the same continental plate, drift is moot. So they used the Eurasian-plate centered system (ETRF2000). Gran Sasso and CERN are relatively static, not rotating, so no correction is required for distance contraction. GPS signals do require both SR and GR corrections, and these seem to be applied directly at the receiver, and are hence buried deep in a chain of references (arXiv paper -> metrology reference -> GPS receiver maker's docs -> GPS standards). That is probably one reason why there are so many questions about those corrections. Ajoykt (talk) 03:41, 14 January 2012 (UTC)


 * The rotation to which I was referring was mainly the daily rotation of the Earth around its axis, not just rotation of the line between Gran Sasso & CERN relative to the Eurasian Plate. JRSpriggs (talk) 07:21, 15 January 2012 (UTC)


 * I just noticed Claudio Germana. "Are OPERA neutrinos faster than light because of non-inertial reference frames?" in the section of peer reviewed papers. This appears to be the sort of explanation for which I was hoping, if it holds up. JRSpriggs (talk) 19:47, 18 February 2012 (UTC)
 * I think Germana's paper hinges on OPERA not collecting data during peak winter months; if data were collected through the year, per Germana the deviations would cancel out. There is the problem that both repeats of OPERA showed the same results&mdash;one spread over the non two-or-three winter months, and the other just in October. Hard to believe a seasonal offset worked to introduce the same systematic error for both cases (one over a period of 9 months; another just one month). There is the additional issue there isn't going to be any way to check this in the near future. 68.126.183.202 (talk) 01:16, 19 February 2012 (UTC)

A report on another repeat of the results by OPERA
Live science reports that "OPERA physicists conducted the experiment again, using only atomic clocks to measure timing, and still came up with the same result." I haven't seen anybody else report this (and various other reports suggest they will check clock synchronization only in March), so not including it. http://www.livescience.com/18555-faster-light-neutrinos-experimental-confirmation.html Ajoykt (talk) 19:21, 18 February 2012 (UTC)

Fiber optic cable connection as possible cause of error
Currently suspected, but unconfirmed hypothesis suggests that the 60 ns error came from a loose connection from the GPS unit to a computer may be to blame. After tightening it and rerunning the test on the connection the result yielded a 60 ns difference; but the experiment in question has not been run again. Until further tests are done I propose keeping the unconfirmed, but suspected cause of the error. ChrisGualtieri (talk) 20:53, 22 February 2012 (UTC)
 * The only source is Cartlidge, and his tone seems rather final. As to the test-with-only-atomic clocks, how credible is the LiveScience report (see the section in the talk page just above)? Nobody else has reported it, and I thought the CERN beam was actually shut down for the winter? Ajoykt (talk) 21:24, 22 February 2012 (UTC)
 * For the time being, we should only state the content of those reports without comment. Also LiveScience now repeats the Cartligde report - so let's wait some time. (http://www.livescience.com/18603-error-faster-light-neutrinos.html) --D.H (talk) 22:40, 22 February 2012 (UTC)
 * The source seems to be James Gillies, and tomorrow a CERN press release should be published
 * http://www.science20.com/quantum_diaries_survivor/opera_result_affected_instrumental_error-87192
 * http://www.cbsnews.com/8301-205_162-57383079/flaw-in-gps-setup-found-in-faster-than-light-neutrino-test/.
 * When they don't mention the "atomic clock versus GPS" issue in the press release, then the former "LiveScience" report from February 18 should be removed. --D.H (talk) 22:59, 22 February 2012 (UTC)


 * Now, also nature (http://www.nature.com/news/flaws-found-in-faster-than-light-neutrino-measurement-1.10099) is on board. --D.H (talk) 23:44, 22 February 2012 (UTC)
 * See http://www.repubblica.it/scienze/2012/02/22/news/neutrini_pi_veloci_della_luce_c_era_anomalia_in_strumenti-30349960/ (Google translate unless you read Italian). If the translation is right, looks like the connector came loose while they were detecting neutrinos, not during the initial calibration. In that case, I don't see how anybody could be sure what delays it added - it is practically impossible to replicate the effects of a loose connector. Cartlidge is based in Rome and being from ScienceMag probably has contacts at INFN, but it is a bit irritating he relies so much on anonymous sources. Ajoykt (talk) 04:25, 23 February 2012 (UTC)

"The Telegraph" also has it. JRSpriggs (talk) 10:09, 23 February 2012 (UTC)


 * CERN updated their press release (http://press.web.cern.ch/press/PressReleases/Releases2011/PR19.11E.html). --D.H (talk) 11:09, 23 February 2012 (UTC)
 * + BBC news (http://www.bbc.co.uk/news/science-environment-17139635) --D.H (talk) 11:39, 23 February 2012 (UTC)


 * OPERA-member Luca Stanco (http://www.newscientist.com/article/dn21510-was-speeding-neutrino-claim-a-human-error.html) also expressed some of his insights. --D.H (talk) 15:23, 23 February 2012 (UTC)
 * I think the crucial point in Cartlidge's article is the statement that the loose connector accounts precisely for a 60 ns difference. If true, that would pretty much seal the case; a coincidence that strong is possible, but not at all probable. But none of the other sources - CERN, OPERA, Luca, precisely say exactly how much error the fiber connector contributed. Also, DH: the CERN press release you mention has nothing on the issue? Ajoykt (talk) 16:21, 23 February 2012 (UTC)
 * Well, we won't find out before OPERA publishes more details. The Cartlidge paper is probably not correct, but it is at least cited by many other papers. --D.H (talk) 16:33, 23 February 2012 (UTC)


 * Whatever, I think we should at first rely on the CERN press release, containing the most recent information (I cannot imagine that OPERA isn't aware about this release). --D.H (talk) 17:25, 23 February 2012 (UTC)

Style and grammar
On a different tack, the grammar of this section could really use improvement.

"The OPERA collaboration has identified two possible sources of error: A loose connection between the fiber link from a GPS receiver to a computer, tightening the connection makes the delay through the fiber decrease; and a second error with the crystal oscillator is expected to have lengthened the reported flight-time of neutrinos."

^This^ seems to be worth at least two sentences (maybe three), and then there is "The two errors affect the result in opposite ways." which now seems redundant. Hopefully someone here can work out something better. Chris857 (talk) 16:52, 23 February 2012 (UTC)


 * Seems much improved, though possibly and like words are a bit frequent. At least it now has real sentences. Chris857 (talk) 03:23, 25 February 2012 (UTC)