Are unofficial Higgs Combinations “Valid” ?

The Unofficial Higgs Combination Tool has now been updated with all the new Higgs plots released in the last few days, including the Tevatron updates and the new 8 TeV data from the LHC. There will probably be more to add on 7th and 9th July from ICHEP. Feel free to play around with it.

At the CERN press conference yesterday the Director General Dr Rolf Heuer warned journalists about unofficial combinations. What he said exactly was at follows (It is 26:50 in if you are looking on the recording):

The fact that they [CMS and ATLAS] have not yet combined their results today is that they did not have enough time. We should have shifted the Melbourne conference by 2 weeks or 3 weeks or 4 weeks but that was not possible. You have to stay tuned until at some time they combine their results. Whatever combination you get beforehand is unauthorised and is certainly not valid because you have to take into account the different correlations, one has to be very careful.”

I agree with what he says. The unofficial combinations you find on this blog are approximate and unofficial and should be used with caution. I have always made that clear. It is not just the correlations that are neglected. The quick combination method assumes that the statistical errors have a flat normal distribution and that is not quite correct. The detector collaborations don’t provide detailed likelihood data to outsiders so this is the best I can do. Luckily all statistical errors tend towards the normal Gaussian as the quantity of data increases (central limit theorem) and in most cases there is enough data for the results to be good, with a few exceptions.

Whether the combinations are “valid” or not depends on what you are using them for. I don’t consider them valid for writing up published results of any kind, but they are good enough as a rough guide to theorists looking for possible signals in the data and there is nothing wrong with showing them at conferences as some eminent theorists have already done, provided they come with appropriate caveats.

I have previously shown some comparisons between official combinations and my unofficial ones to show how accurate they can be (or not). I think it is worth doing a few more now using some of the recent results where the amount of data has increased. In all the plots below the red line is the official result and the black is the unofficial. First up is the latest version of the Tevatron combination compared with an unofficial combination of the updated Dzero and the latest CDF plot that was updated in March. You can click on the plots to get a larger version.

The combinations across all channels have always worked quite well because they use lots of data. The last time that the LHC provided an official combination for ATLAS + CMS was in November when there was only 2.3/fb. here is how it looked next to the unofficial combination that I had done 10 weeks earlier.

Notice here how the accuracy gets worse at higher energies where there is less data available. Heuer seemed to be implying that there should be another combination due out soon. If so it will be interesting to see if the comparison improves as I would expect.

The combinations for single channels have been less successful in the past, but now they are improving. Here is a reconstruction of the ATLAS combination for 7 TeV + 8 TeV data in the diphoton channel

But the results don’t always come out so well even now. The 4 lepton channel uses very few events in both the signal and the background. Here is the result of a similar combination (Update: There was an error in the digitisation that I now fixed and it is not so bad now)

The combination across ATLAS and CMS should be better because it involves twice as much data. They should also have twice as much again by the end of the year so by then combination should work OK even in this channel.

If you want to try more the Higgs combination tool is easy to use and free.

Update: I said that I dont think these combination methods should be used in published papers but other theorists are apparently not as reticent. arXiv:1207.1347 is one example of  paper showing a combined signal plot as well as combined channel values and other fits. There conclusion is that everything fits the standard model except that the diphoton rate is 2.5 sigma too high, in agreement with my figure.

40 Responses to Are unofficial Higgs Combinations “Valid” ?

  1. Lubos Motl says:

    Sometimes it’s better than in other contexts.

    BTW I think that all viXra blog readers may want to hear what Mathematica’s and ANKOS’ Stephen Wolfram has to say about the Higgs boson and everything he’s been experiencing in particle physics for decades:

    • Congratulations for this contributions! Did Stephen asked you, after complicated analysis to decide which was the better place to do a post ;-)

      • Lubos Motl says:

        Dear Alejandro, Dr Wolfram sent me a link to the text somewhere on his website saying that I could be interested – and I was – and I offered him to circulate it to a new broader audience via TRF which he happily agreed with.

  2. [...] ¿Realmente podemos confiar en las combinaciones oficiosas de Philip Gibbs? Su hipótesis sobre el comportamiento de los errores (seguir una distribución normal o gaussiana). Cuanto más colisiones se consideran, gracias a la ley de los grandes números, más confianza tienen estas combinaciones. Abajo tenéis la única combinación oficial ATLAS+CMS publicada hasta el momento (finales de agosto de 2011) y cómo se compara con la combinación oficiosa (curva roja). Sobran más comentarios (pero si queréis más podéis consultar esta entrada de viXra). [...]

  3. [...] also: Are unofficial Higgs Combinations “Valid” ? Share this:ShareTwitterEmailFacebookLike this:LikeBe the first to like this. Tagged: HIGGS Posted [...]

  4. Lubos Motl says:

    Dear Phil, I’ve looked at all the global combinations for the individual channels – changing only the channels options.

    The diphoton is 70% above the Standard Model which is not really too high – arguably below 2 sigma – and otherwise I would say that all the other channels including ditau are really “right on the money”.

    As far as I can, there’s pretty much no experimental reason to escalate speculations on particular deviations from the Standard Model. Do you agree?

    • Philip Gibbs says:

      On the data bubble the quantity xsigm tells you how many sigma it is outisde the standard model. Diphoton is 2.5 sigma, nothing to get excited about but it is good to see it going in the right direction. ditau is 2 sigma but this is very early and I dont trust it.

      I think it is always interesting to talk about what could cause these things if they are real. We can then look for signals elsewhere that might match up. However I agree that we should not blow it out of proportion.

      • Ervin Goldfain says:

        Is it then fair to say that all the discrepancies seen in di-gamma, di- tau, WW to ZZ are artifacts of under-sampling? Same conclusion on about anomalies seen at BaBar in B-decays?

      • Philip Gibbs says:

        I don’t think we can really say that either. We can only say that so far there is nothing outside the standard model that is significant enough to take very seriously yet.

      • Ervin Goldfain says:

        I agree.

      • iya says:

        Isn’t diphoton xsigma = 1.95 at 125.5 GeV?
        It’s still missing the 8 TeV data from CMS, which showed only a very small excess above SM (page 45 of the Incandela presentation), so it will pull down the combination a little more.

      • Philip Gibbs says:

        Iya, I can’t get the data from the p-value plot and CMS did not give separate exlusion plots for 8 TeV. You have to use the officially combined 7 Tev + 8 TeV data by leaving the energy selector on the “official” setting

      • iya says:

        Ah, so that’s where the new data is.
        So, we can already say that the excess above SM is significant at 95% CL? Can’t wait for the end of year…

      • sika says:

        iya says:
        So, we can already say that the excess above SM is significant at 95% CL?

        To quote Lubos Motl:
        “The 2.5-sigma deviation is equivalent to the nearly 99-percent confidence level that something unusual is going on over there.”

      • Philip Gibbs says:

        The individual plots for diphoton at 7 TeV and 8 TeV are now available and have been uploaded to the applet. The consisitency of the diphoton excess is impressive. Statistically it is quite clear. The only question in my mind is whether it is due to real BSM physics or is just a combination of systematic and theoretical errors. Such a big error is unexpected but these things happen. Look at how bad early measurements of Hubbles constant were for example.

        There will be some improvements with better top mass estimates and more data and work that could help calibrate the detectors, but we may need to wait for a linear collider to pin it down.

        I don’t find anomalies in the other channels convincing at all yet, but that is just my gut feeling which does not count for much.

    • Philip Gibbs says:

      If the diphoton signal is real it could easily reach 4 sigma by the end if the year. Then it will be very exciting.

  5. mr.nothing says:

    Congratulations, Philips, you’re a good scientist, and honest. Your observation that the anomaly on the decay in the channel of the diphotons may be owed ​​to the existence of bosons with charge + – is correct.
    Soon I send an article in vixra, which lodges the load of the rest mass of the Higgs bosons, as well as the scale of supersymmetry. Observed Excess around 250-300 GeV, is due precisely to the existence of all other Higss bosons, whose masses are mA = 261.8 Gev
    mH0 = 262, 3 and mH + – = 273.87 Gev
    The variance between these masses gives precisely the fluctuation about 320 Gev.

  6. Higg’s boson godly particle can be tuned up mentally to produce any matter out of fast gravity as carried out by Shri Sathyasai Bhaba who materialized things out of his hands.This is nothing but true .What scientific explanation you give for this output.He has materialized idols under the earth and they were very hot.It is the love and affection.Iis the love and affection simulated by brain and atman produces Higg’s boson results by gravity at a very faster rate.What benefits the the ordinary instruments will brings about even with such a discovery.Few things are beyond scientific explanations.Electron particle acting as interference wave in the system of entering a barrier along with electron particle that act as a guide wave by analogy could be used in a new propulsion system. The ring radiation could be confined with John Pendry’s invisible cloaking screen the exiting pulse is strongly chirped: short wavelengths are propagating faster than the long wavelengths.
    The idea came to me when magician David Copperfield throw the card through the window glass on the other side that penetrated the glass without any damage that promoted my line of thought .Saint Thiru Gnanasanbhandar disappeared in space along with his new bride on his marriage day along with some 16000 disciples in fireball quantum black hole.This was not a story.It really happened.He brought back one woman who died already from her ashes.Her name was Bhoompavai and this happened at Mylapore temple ,chennai.,Tmailnadu.

  7. Ravi K says:

    Hi Phil, Thanks for the unofficial Higgs combination tool. When I try it there seems to be no data in CMS 8 TeV channel for digamma. I am using all default settings except that I set: Collider: LHC, Expt Custom CMS, Energy Custom 8 TeV, Channel Custom H to gamma gamma. Sub-channel is left official as in default.


    • Philip Gibbs says:

      CMS have not yet provided separate 8 TeV data but I have digitised the 7 TeV + 8 TeV data. If you leave the energy selector on official you can get that and combine it across CMS+ATLAS

  8. Lubos Motl says:

    You’re right, Phil, I had to look carelessly; it’s 2.5 for diphoton and around -2 for ditau.In total, that’s more than one would expect. You have 5 channels (Z-gamma is not reported here) so the sum of xsigm^2 over those channels should be 5 in average, right? But 2.5^2 + 2^2 = 10 or so which is about 2 sigma above 5, if you know what kind of chi distribution I am talking about.

    So the deviation from the SM over all channels is at the 2-sigma level now, marginally enough to be interested in it.

  9. vmguptaphy says:

    Higgs Boson has become physical reality. Since Pico-Physics encompasses nature without classification of different area, it is natural for us to have a serious look at new discovery. The subject matter of this discussion also available at

    The essential difference between mainstream physics and picophysics is how property of inertia and gravity is assigned. PicoPhysics do not distinguish between mass and energy. PicoPhysics assigns these properties along with energy to Knergy. Mainstream is looking for answers in elementary particles. In picophysics, there are very few elementray particles – UCO unit conserved object and neutrino. Integer number of UCOs are confined into main stream particles. These particles have the limitation on observed speed being less that light speed due to vortex like flow of UCO inside the particle. So particles with mass have the speed of light as limit.

    The appearance of Higgs Boson is the apparent image of colliding particle. The decay of Higgs boson into two photons represent the gravitational force of attraction never exceeds the level to confine them into a particle. (Expanding spiral on loss of confinement with separation from mother particle, preceding decay into photons).

    The proof of this hypothesis is feasible, if the measurements are repeated at different energy of the incident particle. The pattern in terms of peaks and valleys shall remain essentially same, with shift in energy axis. The prominence of 125 Gev peak shall be reduced if experiment is conducted with particles accelerated to different energy levels.

    CERN experiment may make the internal structure of mainstream particles visible.

    In PicoPhysics Higgs Boson will be a virtual particle, an image at the time UCOs tear apart overcoming gravitational force of attraction between them.

  10. Lubos Motl says:

    Dear Phil, if you want to verify my chi-squared calculation

    concluding that the overall disagreement from the Standard Model is at the 2+ sigma level now, it could be useful. I use your xsigm values for 5 channels.

  11. [...] 7.7.: here are the vixra unofficial Higgs combinations of both detectors (1) and (2). Also of interest might be a guest post of Stephen Wolfram on TRF. Share this:EmailPrintLike [...]

  12. anonymous says:

    Lubos—People have discussed a light stop for a long time. I can’t believe that Higgs branching ratios in the light stop MSSM haven’t been calculated in many, many papers before. I’m curious as to what Buckley and Hooper have done that hasn’t been done before.

    • Lubos Motl says:

      That would be normally my guess, too. However, can you actually tell me the coordinates to a paper that has done this simple thing before?

    • Lubos Motl says:

      Let me mention one more thing. One can perhaps find papers calculating much more general decays as functions of lots of parameters, or just describing how to do it, or presenting a program to do so.

      If that were so, I would still say that such a thing does *not* include the actual observation what a simple light stop, perhaps stripped of very light gluinos and other things that would be implicitly meant, does.

      The previous papers may include tools how to do such things etc. but if no one has said that a light stop squark may modify the branching ratios, then – assuming that the claim is right – Hooper and Buckley should be credited for this observation and it could be the real essence of a breakthrough. Important physics isn’t just about writing general things etc. but also about isolating the important things in the sea of irrelevancies. Disagreement?

      • anonymous says:

        I agree. A quick inspire perusal found which does the calculation. But it is pretty old and might not be totally relevant. If it hasn’t been done in the more recent context though, then more power to Buckley and Hooper.

      • Lubos Motl says:

        Very relevant paper! It’s hard to find out the relevance for today’s LHC from it, however. Moreover, it talks about a light stop of mass 200 TeV and 300 TeV several times – is it a repeated typo?

  13. Anonymous says:

    Dorigo has an interesting article about Higgs aftermath. Basically, the LHC analysis might have not been as blind as claimed. The experiments might have stopped the data collection in a situation where the data had a positive fluctuation in order to reach 5 sigma discovery level. This would neatly explain the above SM cross-section for several channels.

    • Philip Gibbs says:

      His “keep-looking” effect may have some validity, but the excess in diphoton was there at the CCM in December and only grew a little when they made changes. It could be that it is there as a combination of this plus theory errors plus systematic errors plus statistical fluctuations, or it could simply be new physics.

      You can’t really account for the deficit in WW and ditau this way. I am less convinced about those as about diphoton because the analysis is not as clean.

      I’m putting the odds at 50/50 for a BSM explanation and looking forward to the next update to at least reduce the statistical uncetainty. Note that the independent measurements of the diphoton and 4 lepton spectrum aside from the Higgs search helps to eliminate some other causes of systematic error. Some of that was also shown at ICHEP and looked OK but better statistics will help there too.

      Not sure how they can reduce theoretical uncertainty apart from a small improvement from the mass of the top, but if theory error is the only thing left to account for it may be smaller than the observed excesses and deficits.

    • Ulla says:

      The experiments might have stopped the data collection in a situation where the data had a positive fluctuation in order to reach 5 sigma discovery level.

      So this result is manipulated? Maybe they have selected away results too? This was no good news.

      • Philip Gibbs says:

        This does not mean its manipulated, just that part of the process was not blinded so this could introduce some bias. I doubt the effect is large if there at all.

      • vmguptaphy says:

        For some recording errors, I think provision may have been made by statisticians who decided on minimum difference between noise and signal to establish results positively.

  14. [...] también nos indica que estamos ante el bosón de Higgs predicho por el modelo estándar [ver esto, esto y esto]. La situación está tan clara que hasta Matt Strassler lo tiene claro, como nos cuenta en [...]

  15. [...] Buckley & Hooper, STFC, Brown und Univ. of Buffalo Releases und Artikel hier, hier, hier, hier, hier, hier, hier, hier, hier, hier, hier, hier, hier, hier, hier, hier, hier, hier, hier, hier, hier, [...]


Get every new post delivered to your Inbox.

Join 270 other followers

%d bloggers like this: