Friday, June 28, 2013

Jacques Distler on his lost new physics bet

Jacques Distler of Austin lost a $750 bet to Tommaso Dorigo and wrote an article about it:
Guest Post: Jacques Distler, Why I Lost $750 On New Physics At The LHC (original text)
They ultimately agreed that Distler wins if a new-particle-like 5-sigma discrepancy from the Standard Model is announced within 12 months after the moment when the LHC has had accumulated 10 inverse femtobarns of collisions (the energy was allowed to be just 8 TeV, it seems). Gordon Watts supported Distler's bet by another $250 so if he concedes as well, and I think he should, Dorigo should be $1,000 richer because the conditions were fulfilled several weeks or months ago.

It was a risky bet for both sides. Would Distler make a similar bet on the 2015 run?
The answer, I think, is: not unless you were willing to give me some substantial odds (at least 5–1; if I think about it, maybe even higher).
Well, my bet against Jester turned out to be effectively a bet on the early 2015 run because we're talking about 30 inverse femtobarns which haven't been accumulated yet (we're around 27) and our bet is 100-to-1 which means that I may win $10,000 but will lose $100 only. You see that your humble correspondent was expecting better conditions and when it came to the assertions, he was more cautious, but just by a factor of three, than Distler and Watts.

I think that my odds of winning remain substantial because the 13 TeV collision energy of events that will start to be produced in April 2015 opens a whole new game.

If you keep your energy constant, it's pretty much necessary to have at least 3-sigma excesses if you want to see 5-sigma excesses at a doubled or tripled total luminosity. After all, the results are changing quasi-continuously if you're adding new collisions to your dataset. When the luminosity jumps 6 times or so, you get "completely new data" because 5-sigma deviations in the new data may easily come from bumps that were expected to be smaller than 2 sigma, and therefore invisible, in the 6 times smaller dataset.

However, with that doubling in the energy, the rules are completely different because particles of certain masses would be almost impossible to discover at 8 TeV but they may be instantly discovered at 13 TeV, perhaps after an inverse femtobarn of data if not earlier. So Jester of Resonaances cares about his $10,000, he shouldn't sleep well until mid 2015 or so. ;-)

Yes, I surely think that even with all the facts we know, the probability that new physics will be discovered in the early 2015 run is substantially greater than 1%. In other words, my position in the bet is worth jealousy. The new particles that may be found may still be much much lighter than 1 TeV. Tons of scenarios with the LSP at 130 GeV (like in the Fermi hints) or even 8.6 GeV (like in the dark matter direct search experiments) remain viable.

Jacques starts his article with some memories of Steven Weinberg on C-SPAN talking about the SSC two decades ago; and about the way how energy is divided between the partons inside the proton so that the actual energy scale you may easily probe is smaller than the proton energy (except for an ever smaller fraction of the collisions in which the proton energy is increasingly more concentrated in one parton).

Distler ends up by saying that the probability that the LHC will ever see new physics has dropped significantly; and the conditional probability that the new physics, assuming that it will be found, will be supersymmetry has increased because the other types of new physics were disfavored much more rapidly. I agree with those statements assuming that the word "significantly" is understood in my way. It's significant but surely not totally qualitative. Maybe there would be a disagreement between Jacques and me if the meaning of the words were clarified. The disagreement could boil to this statement by Jacques:
Still, there are (or were) lots of scenarios with new physics, accessible to the LHC. And theorists, being perennial optimists, put a lot of effort into exploring those scenarios.
I disagree and I have always disagreed with this definition of optimism of a theorist; it is a bias in the literature, not legitimate optimism. A theorist thinking like myself is equally pleased if Nature obeys nice laws with a new particle waiting at 150 GeV; or nice laws with a new particle waiting at 3,000 GeV. If you're equally pleased by both possibilities, you can't say that believing in one of them is "optimism" and believing in the other is "pessimism". Preferring the former possibility – to the extent of selectively writing papers about the first possibility – is just wrong and if the bulk of phenomenologists are acting in this way, it is (and actually was) a case of group think. An experimenter dreaming about his own discovery – assuming that the experimenter's job is to maximize the probability (times importance) of a discovery – may call the belief in low-lying new physics fruits "optimism". But a theorist's job is to find the truth so he simply can't afford the asymmetric perception of different, equally justifiable or likely scenarios. There are lots of "big desert-like" scenarios where – up to a possible exception of SUSY – nothing happens between the electroweak scale and the Planck scale (or at least the GUT scale). I think they're pretty in their characteristic way so the belief that they're true can't be called it a "pessimistic belief".

"Hopes" in the new physics around the corner were always (mostly) motivated by some phenomenologists' desire to increase their odds to get famous quickly (so claiming that this bias was due to their "virtue of optimism" is completely obscuring the true motivations) and the impact of this desire on the literature may be classified as a distortion of the facts because the composition of the literature reflects their desires rather than available facts and it's always wrong for theory literature to be skewed by similar non-fact-based pressures. So this bias describing "new physics around the corner" was surely wrong and I always thought it was wrong but this wrongness doesn't imply that the LHC will never see new physics. Jacques Distler may switch to an opposite extreme but this won't make new physics at the LHC in 2015 impossible just like his "optimism" didn't guarantee early LHC discoveries of BSM physics.

I believe that as the LHC is increasing the total luminosity and/or energy kind of exponentially, the probability of a new discovery per unit time is staying pretty much constant because a sensible distribution of physical phenomena between the low-energy scales and the Planck scale is pretty much uniform on the log axis (you may even estimate the density of particles per decade from the particles we already know). The idea that if the LHC finds something new, it has to happen immediately, is unwarranted. Well, the LHC made a qualitative leap at the very beginning because it had more energy than any earlier experiment. It only found the Higgs boson as the quasi-new physics. But that doesn't mean that almost the whole space of possibilities has been exhausted. Now, the LHC is moving the frontiers of science more gradually.

It's plausible that LUX will find the sub-10-GeV dark matter particle before the end of 2013 i.e. years before the LHC will say anything about it. The two years' vacation at the LHC may substantially shrink the collider's competitiveness in discovering new physics.


  1. Hello Lubos. I would like to send you an email to ask a math/physics question, but I don't know your email address.

    Can you please send me an email so I can email you back? Thanks. Of course, I would keep your email address confidential.

  2. Dear Andrew, done. [Lots of communication occurred here.] End.

  3. The really bad thing about biases towards new physics around the corner is that it gave sourpusses a too large point of attack ... :-/

    Dear Lumo, I heartily wish you more success with your bet than Jaques Distler had :-)

  4. Haha, don't consider a few critical bloggers more important that the actual science :) The bad thing about "searching under the lamppost" biases is if actual working physicists get led astray.

    I think a lot of people will pop a bottle of bubbly stuff if Lubos wins his bet. I might. :)

  5. Θ³Σx² - ∂³Σx² - ΘΣJun 28, 2013, 6:41:00 PM

    I disagree with your opinion about that we will find DM particles in the new LHC rund by 2015, Lubos. Aging hava turned me in a very conservative yet radical man. I really thought that the SUSY god-opportunity was in the first run...If the "standard" and "minimal" (simpler) supersymmetric scenarios were true, we should have obtained some hints from B-factories, LHCb, and of course at ATLAS or CMS. The SM (even when we know it is not a complete theory) rules yet. I suppose you do remember all those books (not only divulgation) and talks some years ago trying to push forward and say that we would be finding Extra Dimensions/Superstrings and or effective gravity scale at around the TeV scale. Of course, it COULD be true, but the odds and probabilities have decreased A LOT. You can not deny that ;). We can tune of course (yet) the MSSM and all his more than 100 free parameters to fit the actual data, but the SM does it even better! Some books long ago were written where it was said "Only a new boring fit to the SM" And we (sadly) are in (almost) in the same place than 40 years ago, but now with a Higgs sector to understand (is it a single or is it a multiplet around 127GeV?). My only left hope is that Higgs measurements plus neutrino experiments (nu-factories, short and long baseline exeperiments, neutrino astronomy) and dark matter experiments say something new, or we will experience a decrease in the HEP community. LUX and some ideas concerning WISPs and ultralight axion-like particles are very interesting too, I agree there. Optical magnetometry seems to be a new promising tool...Moreover, we have to detect (some day) gravitational waves...

  6. "Tons of scenarios with the LSP at 130 GeV (like in the Fermi hints) or even 8.6 GeV (like in the dark matter direct search experiments) remain viable."

    Can the LSP be as light as 8.6 GeV if it is a neutralino, or would it have to be a sneutrino?