Higgs Combination and Fits Revisited

Now that the Tevatron Higgs combination is public and I can see how much error there is in the direct combination process, it seems like a good idea to redo my earlier combinations. I know lots of people are interested to see these now to give information about where we stand.

This first plot is the LHC Combination with a grey band to show the uncertainty in the combination process. This is based just on the observation that the Tevatron combination was up to half a sigma out in places and I am assuming that similar size errors can be expected for the LHC combination. In fact my best estimate is that the combination will closely follow the upper limit of the grey region.  Up to you to decide whether this is “NONSENSE“. 🙂

The Tevatron results are still best at the lowest masses so let’s combine the new Tevatron combination with this one. there is most uncertainty in the regions where all experiments have similar limits.

What this is showing is that an excess around the 140 GeV area is possible but it is not likely to be consistent with a standard model Higgs because it is below or near the red line. If the excess is at the higher limit as I expect then it will have at least 3 sigma significance.

We can strengthen this by doing the global fit with the combination uncertainty shown. The electroweak precision tests reduce the likelihood that a Standard Model Higgs Boson is at this mass. (I should point out that there is an important different between the way I have combined these plots and what the gfitter group have done. They adjust the plot so that it always has a minimum of zero. This is because they are making a prior assumption that the Standard Model Higgs boson exists at some mass which I am not. If you have doubts about the validity then do not go beyond the combined exclusion plot above)

If you compare this with my previous Standard Model Killer plot you will see that the black line is slightly lower at the minimum point because of the marginally less restrictive Tevatron combination. The combination uncertainty now added in grey shows that the Δχ2 could go as low as 2.5. Although this is not as dangerous for the Standard Model as before it still corresponds to a 90% or better exclusion for all Standard Model Higgs masses.

Some of the updated SUSY model fits only manage an 85% exclusion and other less restricted supersymmetry models would surely have a better chance. I think it is therefore reasonable to claim on this basis that Supersymmetry is in better shape than the Standard Model Higgs. This is contrary to the slant from the media and some other blogs who suggest that the excesses at 140 GeV are hints of the Higgs Boson while supersymmetry is in more trouble.

Of course many possibilities are still open and more data will certainly make a difference.

Update 29-July-2011: To be clear about what this does and does not rule out.

If we accept the combination uncertainty estimate and the statistical validity of combining all direct searches with electroweak fits :

  •  We indirectly rule out a lone standard model Higgs boson of any mass with no additional BSM physics at 90% confidence, i.e. a fair bit short of conclusively.
  •  We directly rule out any standard model Higgs boson at 95% confidence except in the mass ranges 114GeV to 144GeV or 240 GeV to 265 GeV or above 480 GeV
  •  We do not say anything about other BSM Higgs-type mechanisms including composite Higgs, technicolor Higgs, Higgs doublets, SUSY Higgs, Fermiophobic Higgs etc. These would require a separate analysis.
  • We do not rule out high-mass Higgs bosons above 480 GeV in combination with other BSM physics that could explain electroweak fits and cure theoretical limitations of the SM at higher energies.
  • We see excesses at around 130 GeV to around 160 GeV that could be over three sigma level. It might suggest some new physics such as some kind of Higgs particle(s) in this region. However, these are spread wide and are near the exclusion limit. Perhaps a different Higgs model would fit better than a lone Standard Model Higgs boson.
See also Motl’s analysis which complements mine and predicts that the highest excess will be at 144 GeV

41 Responses to Higgs Combination and Fits Revisited

  1. Luboš Motl says:

    Well, 109-132 GeV is still allowed by you. It’s too early to say and I am impartial and blah blah blah but I surely like your results. 🙂

    There’s one Higgs below 120-130 GeV, one Higgs around 144 GeV, and tan(beta) is of order 10. 🙂

  2. Wilhelmus de Wilde says:

    Hi Philip,
    I am just a (very) interested layman, but I read an article :
    that is very positive about at last finding the Higgs boson, the Birmingham particle physicists (3000 strong team of scientists !!!) are trawling through the data from the Atlas and are very positive, they expect to have a definitive answer at the end of 2012. Is this the same source ?
    best regards

    • Philip Gibbs says:

      There are indeed hints of a Higgs Boson around this region but the global fit shows that it is not likely to be a lone Standard Model Higgs. It could be part of a SUSY Higgs multiplet or some other system of particles.

      The collaborations are saying that my plots are nonsense and everyone should wait for the official combination plots and fits before drawing conclusions. That is why the media story is a little different. In reality the official results cannot make a substantial difference to the conclusions unless more data is added in. That is shown by the uncertainty regions I have plotted here.

      All the signals are weak by the standards of particle physics so we have to wait for more data to get a better indication of what is really going on. The CERN DG said that he is confident we will have a definite answer about the existence of the Higgs by the end of 2012, but he is playing safe and we are really likely to have much more information even by the end of 2011.

  3. Alex says:

    “I think it is therefore reasonable to claim on this basis that Supersymmetry is in better shape than the Standard Model Higgs.”

    Let’s say a simple two higgs doublet model is in better shape than the Standard Model Higgs – absent any other observations, that’s the logical minimal assumption…

    • Philip Gibbs says:

      Good, so a Higgs doublet model would be a good place to start to look for something that can fit the present results a bit better. I suppose it must be all those negative missing energy searches that hurt the SUSY fits most.

      If the Higgs multiplet parts can exist on their own people should be fitting to those models too. With any luck some people are and we will hear about it soon.

    • Alex says:

      They can, absolutely. The MSSM type higgs sector is what is called the Two Higgs Doublet Model Type II (THDM II), in which the two doublets have opposite hypercharge and one only gives masses to the up -type, the other only to the down-type fermions. If SUSY is not imposed, both can give masses to up and down type fermions, so THDM models can actually be a little more general. There is a tan beta as in the MSSM of course, and charged and peudoscalar higgs (if CP is approximately preserved). I just found out that these people (http://www.isv.uu.se/thep/MC/2HDMC/) provide some software for that for example…

    • Luboš Motl says:

      “Let’s say a simple two higgs doublet model is in better shape than the Standard Model Higgs…”

      Is that some kind of a censorship? Did Phil write something that violated your taboos or holy cows? Did he use a word that has to be suppressed?

      Phil’s original sentence is not only true but more informative and accurate than yours, Alex. First, supersymmetry is the only known deeper explanation why the number of the Higgs doublets could be two – i.e. non-minimal.

      Second, the LHC data seem to agree not only with a generic two-Higgs-doublet model but with its special form, a supersymmetric two-Higgs-doublet model – especially because the constraints on masses imposed by SUSY are nontrivially compatible with the observations so far.

      That’s why the new experimental evidence supporting Phil’s statement is stronger than the new experimental evidence supporting your statement.

  4. Alex says:

    Yes Luboš, censorship. Because obviously I hacked Phil’s blog in order to remove all mention of supersymmetry, and my cybersquadron goes after anyone who dares mention it … *sigh*. Projection much? I think by mentioning the actual minimal model that can fit the data, I rather insulted your holy cow, not the reverse. That being said, my bet is on color charged superpartners or superpartner like states starting at around 1500 GeV, as some indirect flavor constraints suggest to the desperate.
    That doesn’t change the fact that Senor Occam is still the scientist’s best friend.

    • Luboš Motl says:

      Dear Alex,

      you have apparently been sleeping at least for 35 years so that you have missed the last 35 years of developments in theoretical particle physics. This observation is made very explicit by your statement that “the actual minimal model” is “Two Higgs Doublet Model Type II”.

      This is, of course, wrong. The Two Higgs Doublet Model Type II is surely not any kind of a minimal model, by any stretch of imagination. That’s also why it’s not called a “minimal model”. It doesn’t have the adjective “minimal” in its name. In fact, the name contains two constructs – “Two Higgs” and “type II” – and both of them are associated with two features that make the model non-minimal.

      On the other hand, the MSSM *is* a minimal model. That’s also why it’s called minimal. Let me kindly inform you that the acronym stands for the Minimal Supersymmetric Standard Model.

      In fact, even the (non-supersymmetric) “Standard Model” is less minimal from the top-down perspective as any realization in string theory makes more than clear. So you may be using Occam’s razor but you use it upside down, hurting your flesh instead of the pork. You use the term “minimal” for the things that are not minimal and vice versa.

      Cheers, LM

      • Alex says:

        Thank you for providing your perspective on minimality of models. I think we can let these gladly be the closing words of this particular discussion.

      • Dr. Guest says:

        “You have apparently been sleeping at least for 35 years so that you have missed the last 35 years of developments in theoretical particle physics.”

        No, he is only differentiating from “nonsense”

  5. chris bolger says:

    Man, you all should take a chill pill. This is not the Debt Ceiling Debate.

    • Alex says:

      And we’re not American 🙂

    • Luboš Motl says:

      The debt ceiling debate is a lot of fun and it’s great that the Tea Party has already won it – now it just has to decide in which form the bounty should be served. 😉

      • Lawrence B. Crowell says:

        The tea party guys are insane; they are true believer know-nothings who have control of Congress in the same way that bank robbers have control of a bank for a while. This whole thing is manufactured, for there is no real immediate crisis over the debt, but rather a long term question which needs attention by cooler heads. The tea party has basically set its sight on demolishing Obama, and they do not care what they bring down with that. This will lead to a global economic depression — coming soon now.

        About ever century the human race enters into some sort of period of collective insanity. The last one was from 1914-45, and prior to that 1788-1815 and so forth. We are about due for the next one, and it seems at least plausible that we are seeing the dress rehearsal for the next episode of ghastliness. Each one of these episodes has a body count that is about an order of magnitude larger than the previous one. As seen in the past, a global economic meltdown will bring about politics of desperation, and the wanna-be Hitlers, Mussolinis and Stalins are waiting in the wings and chomping at the bit to take the reins. It looks as if we are all going to tango one more time — this time with nuclear bombs.

        It is interesting that there was in the last period of global chaos a “revolution” in physics as well. Well folks, maybe this is deja vue all over again.

      • Philip Gibbs says:

        OK but let’s end the off topic discussion about economics at this point 🙂

      • Lawrence B. Crowell says:

        It is a curious age we live in though. The INTEGRAL data and now the EP presentation with 95% Higgs exclusion seems to suggest we are in a Michelson-Morely intellectual phase, where a vast edifice of scientific theory is on the verge of collapse. This age today is mirrored at a time when the 19th century national/political arrangements were around 1890-1910 beginning to tremble and crazy ideologies began to take hold. This was followed up with WWI, revolutions, brown shirts and … . I listened to Weisskopf once tell about how a debate over quantum mechanics had to acoustically compete with a Nazi rally down the street. His only mental refuge of peace during that time Nazism was on the rise was listening to Mozart and thinking about quantum mechanics.

  6. David says:

    Hi Philip,

    Thanks a lot for all your posts on Higgs searches. I have a layman question:

    In some of these exclusion plots, we all have seen the experimental curve (black) showing an excess of events w.r.t. the theory predictions (Brazil-flag colored), e.g. around 140 GeV etc.

    Even if they are not statistically significant, I do understand why people could get excited about it. But what I do not understand is why they should be interpreted as “hints of a Higgs boson” (or, at least, hints of *the* Higgs boson).

    As far as I can see, any excess indicates more events than expected… according to the SM! (Since the Brazil-flag colored region comes from a computation whose input is the SM). So any real excess would point to a failure of the SM. Sure it is exciting, but, why should it be a hint that the SM Higgs hides there? It could be anything: a non-SM Higgs, SUSY, or a honeybee. But if the SM Higgs exists, it could never be produced at a higher rate than the one given by the “upper limit in the cross-section” predicted by the SM… which is what I thought that the dashed curve in all these plots indicates…

    I know the reasoning above is surely wrong, but I would like to know what I am misinterpreting in all these plots. Thanks!

    • Alex says:

      That was my interpretation, too. The bump at around 115, if real, is likely something with an enhanced production rate compared to the SM higgs.

      • Mark says:

        Alex, why do you say that? Looking at the ATLAS/CMS unofficial plot, the bump at 115 is about 3.3 times the SM, and the background is about 1.8 times the SM. The background has an uncertainty at 1 sigma which can bring it up to 2.3, easily. The difference would then be 1.0 times the SM, which is fine.

      • Alex says:

        Correct me if I’m wrong, but that’s not quite how it works i thought… At the risk of saying something stupid, I don’t think “the background is 1.8 the SM” is what you can read off of that. From theory, we expected to be able to exclude Higgs bosons at that mass with about 1.8 the SM cross section, and failed to do so, instead only excluding Higgs bosons at this mass with more than 3 times the SM cross section, roughly. Isn’t that right?

      • Alex says:

        Ahm, it occurs to me we might be talking about different plots….

  7. Ervin Goldfain says:


    What’s the point of publishing combo plots now if Higgs searches will go on with up to 5 fb^(-1) (or more) by year’s end, when data analysis will hopefully become more reliable? I am not saying that combining preliminary results are not useful, but they can also be misleading…



    • Philip Gibbs says:

      Ervin, I find that quite a surprising question. What I have done here is likely to be a tiny effort compared to what the professionals in this business are looking at now. A huge amount of data has been produced. It may look mostly negative but if someone analyses it the right way they may be able to tease out a clue as to what is happening. The reason its worth doing is that they may then be able to make a prediction before the next lot of data becomes available.

      I think a good thing about the way these experiments are being conducted is that the data is being made available very quickly. With a lot of other big-science experiments the people running them keep the data to themselves for years so that they can analyse it in detail before anyone else has a chance (c.f. Planck). The collaborations here are making extraordinary efforts to get the data out as soon as possible giving the whole physics community a chance to go through it. Science will progress faster that way so we should applaud them and hope that the theorists take full advantage of the opportunity being offered.

      • ervin goldfain says:


        Sorry, but I do not quite see it your way. You say:

        “A huge amount of data has been produced. It may look mostly negative but if someone analyses it the right way they may be able to tease out a clue as to what is happening. The reason its worth doing is that they may then be able to make a prediction before the next lot of data becomes available.”

        There are three “may” in your three-sentence comment which suggests a good dose of wishful thinking. In general, how strong can be a prediction drawn from combining undersampled datasets or datasets with low signal-to-noise ratio?

        You add:

        “I think a good thing about the way these experiments are being conducted is that the data is being made available very quickly…
        The collaborations here are making extraordinary efforts to get the data out as soon as possible giving the whole physics community a chance to go through it. Science will progress faster that way…”

        Are we trading speed for due dilligence when reporting critically important results? Driving in the fast lane does not mean driving in the right direction…

        Feel free to disagree, this is only my subjective take on it.



      • Kea says:

        Yeah, right, so now the stringers load up the arxiv with non superpartner and non fairy papers? Er, and what methods exactly are they going to use to do that? Of course, they could take Schreiber et al’s tack and claim to have known all along that global susy was unnecessary, but the non mathematicians, whose toolbox is limited, can’t really get away with this excuse.

      • Philip Gibbs says:

        Uncertainty is not the same thing as wishful thinking. There is always uncertainty in research until you reach the end. it is important to get as much information as soon as possible for a variety of reasons. Guessing what models could work best now is an important part of that. It will help prioritize the searches that the collaborations look at so that they get best results for their efforts.

        One particular reason for wanting as much information as possible as soon as possible is that someone has to make decisions about what collider options to go for next. The plans for the next 10-20 years include ILC, CLIC, HE-LHC, eLHC, and Muon colliders. It will not be possible to do them all but they must start the research and planning now. The decisions will depend on what physics there is. If there is not much below 600 GeV then there is little point building the ILC for example.

        For me it is just for fun and curiosity, a bit like watching a thriller and trying to guess the final twist before the end. If you prefer just to be hit by the surprise at the end then I hope the rest of us don’t spoil it for you.

        Remember the story of how DNA was discovered. While Rosalind Franklin worked methodically through the analysis to get the chemical structure of DNA, Crick and Wilson decided that there could already be enough information. They just had to use some detective work and some good guesses to get there. I think the current position in HEP could be similar.

      • Lawrence B. Crowell says:

        It also needs to be pointed out that some basic theories deal with structures that are very compelling. Supersymmetry provides a sort of “extra dimension” whereby one can loop around the Coleman-Mandula theorem. This is a theory which deals with “structure” and seems as if it should manifest itself in some way. An emergent form of supersymmetry has turned up in standard nuclear physics, and nature has a way of “recycling” basic structures. What is problematic are all the numbers and fiddle factors people put on it with phenomenology. I have similar, though more reserved, opinions about string theory, where the basic stringy conjecture seems reasonable. As for the Higgs theory, that straddles basic theory and phenomenology. Things look pretty dicey now, a 95% dice roll against it so far. A lot of stuff people took as nearly phys-gospel is in trouble.

        The big problem IMO is that we have a little problem with this idea of the vacuum. We pack lots of crap into it, energy densities, degrees of freedom, lots of ZPE stuff and so forth. I have thought that a lot of the QFT stuff with the vacuum might be excess baggage that needs to be cast off. Supersymmetry is probably some part of the answer, for the Fermi and Boson do cancel.

      • Philip Gibbs says:

        The reason why supersymmetry has so many fiddle factors is because we dont know how it is broken at (possibly) GUT scale. If we did it would become much more constrained. If supersymmetry is right then perhaps we will be able to work out how the breaking works once we have the numbers from experiment.

  8. I believe that it does not take much time to get rid of Higgs hypothesis also in SUSY sector. Internal consistency might already now exclude it because of the limits on squark and gluino masses and the strong constraints between fermion, sfermion, sboson masses from SUSY breaking. Some-one might still remember that SUSY should also stabilize Higgs mass against radiative corrections. And explain also the anomaly in muonic g-2. SUSY should be the healer, not the patient!

    I would not be surprised if technicolor models would experience a brief renaissance but fail experimentally since very many new hadronlike states and new elementary particles is implied by the extension of the color gauge group. Sooner or later the simple p-adic scaling of ordinary hadron physics probably turns out to be the only realistic option. If technicolor becomes in fashion, the hadrons of M_89 hadron physics will be however found as a side product and this is good.

    This will gradually make people ready to make real questions concerning the microscopic description of massivation.

    *p-Adic thermodynamics for conformal weight (to which mass squared is proportional) is part of the answer and extremely natural in the framework involving super-conformal invariance since it does not require the questionable QFT limit but applies in the fundamental theory.

    *How massive particles receive their additional helicities? Could it be that all components of Higgs, of its super partners, and of its higher spin generalizations are eaten in a process in which massless multiplets with various spins combine to form only massive multiplets. Here twistor approach might provide the guideline since its applicability requires that massive particles should allow interpretation as bound states of massless ones. Perhaps the simple observation that spin one bound states of massless fermion and anti-fermion are automatically massive might help to get to the deeper waters.

  9. A little addition. Accepting p-adic thermodynamics in the description of massivation requires giving up the Planck length scale reductionism as the basic philosophy and replacing it with p-adic fractality. This would be the basic guiding principle tying together physics at very short and at very long length scales making possible the long sough for ultraviolet completion of known physics.

    Indeed, the attempt to achieve UV completion- a funny name for what string and M-theorists have tried to do- led to the landscape catastrophe in M-theory since very many physics in long length scales had the same UV completion. Some general principle fixing the long range physics is obviously missing and p-adic smoothness for which infinite in real sense is infinitesimal selects the unique long length scale physics among infinitely many alternatives.

    The real problems are really much much deeper than finding proper parameters for SUSY and it would be a high time for theoreticians to finally realize this!

  10. anomynous says:

    John Ellis showed your combinations at Higgs Hunting 2011 in Orsay : http://indico2.lal.in2p3.fr/indico/conferenceDisplay.py?confId=1507

  11. ervin goldfain says:


    I am not sure I buy your arguments. You say:

    “One particular reason for wanting as much information as possible as soon as possible is that someone has to make decisions about what collider options to go for next. The plans for the next 10-20 years include ILC, CLIC, HE-LHC, eLHC, and Muon colliders. It will not be possible to do them all but they must start the research and planning now. The decisions will depend on what physics there is. If there is not much below 600 GeV then there is little point building the ILC for example.”

    Decisions on future collider plans are critical to the future of HEP. In my view, it would be irresponsible to base these crucial decisions on incomplete knowledge and guesswork. Again, I am not suggesting that preliminary combo plots do not provide useful guidance. But they should be always taken with a healthy dose of skepticism. Even more so when preliminary results are deemed inconclusive.



  12. Jin He says:

    What Phil did is right. In the history of human society and science, there are always some dictators who hijack Truth. The great example is Mach Principle.

    Einstein recognized the importance of Mach Principle. But he could not explain the Principle. He probably knew that his Field Equation is against the Principle. He must have known that the Equation is a guess at the truth of Mother Nature and might not be true. That resembles the way some people made a guess at the explanation of cosmic redshift. Human’s conscience might have realized that those guesses may be wrong. However, those dictators wanted to take an advantage of the fame of Einstein and Hubble. Dictators always treat common people as stupid pigs. However, dictators’ lives and fame are based on the “pigs”. You can imagine: If there were no humans (“pigs”) on the Earth, the dictators would be nothing!!!!

    The dictators made Einstein Field Equation and Cosmic Redshift holy, and lived a grandious life on the “holy” equation and redshift.

    It is possible that the LHC results might kill the theories of Big Bang and Expansion Redshift. However, we “pigs” should be careful of that the “elitei” dictators might once again hijack the TRUTH!!!!

  13. Janko Kokosar says:

    I wrote in the paper ((http://vixra.org/abs/1103.0025):

    “The idea about the pion as an elementary hadron disagrees with measurements.
    Maybe the Higgs boson and a new less mysterious interpretation of quarks will seal
    this disagreement.”

    But preliminary measurements in CERN gives that Higgs boson maybe does not exists:

    If the CERN measurements will rule out Higgs boson, this will not be a full confirmation of my theory, but something a little it will be.

    Do anyone knows to shortly explain, why Higgs boson is so urgent for physics and maybe, where my theory come in conradiction with quantum field theory? I think that Higgs boson follows from quantum field theory.

  14. carla says:

    When Bill Murray says your plots are nonsense, by how much? At the 95% confidence level or a lot lower? 🙂

    • Philip Gibbs says:

      LOL, but seriously, let’s be fair to Bill. It is difficult to know how accurate these plots will work out to be. I think they are worth looking at with the proviso that we understand their limitations. They are not quite nonsense but we dont know how much sense they contain. I have tried to estimate the uncertainty based on comparisons with the Tevatron combination but this may not be valid. I never intended that these should be trusted as anything more than a rough indication.

      At the Higgs Hunting 2011 workshop the gfitter group said they would not try to combine the individual limits because they are now “too complicated” to combine reliably in that way. I certainly would not dispute that view.

      Having said that it was amusing to watch Murray’s talk again, He looks at several bumps on the ATLAS and CMS plots and claims that a coincidence of one pair is potentially significant then a few seconds later says that my plots are complete nonsense based on absolutely nothing. The method worked surprisingly well for the Tevatron if it is based on nothing I think.

  15. […] plots got a surprising amount of attention. At a plenary session during EPS Bill Murray referred to my plots as “nonsense based on absolutely nothing” (which is not too far from the truth). Then […]

  16. […] es obvio, ha habido críticas. Bill Murray ser refirió a estas figuras de Gibbs como “un sinsentido sin ninguna base científica” (el propio Gibbs reconoce que […]

%d bloggers like this: