From: FNALD::SMTP%"neuberg@ekp.physik.uni-karlsruhe.de" 23-DEC-1997 16:00:49.32 To: TOBACK CC: Subj: Karlsruhe comments on Diphoton paper Date: Tue, 23 Dec 1997 22:59:55 +0100 (MET) From: Dirk Neuberger X-Sender: neuberg@ekpux18.physik.uni-karlsruhe.de To: toback@fnald.fnal.gov, frisch@fnald.fnal.gov, REBCDF@fnald.fnal.gov Cc: kuhlmann@fnald.fnal.gov, mshapiro@fnald.fnal.gov, yagil@fnald.fnal.gov Subject: Karlsruhe comments on Diphoton paper Message-Id: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Dear Authors, we congratulate you on this well-written paper covering our most (in)famous event. We had some controversial discussion over the fact that the paper is written based on the measurement of ONE more or less dubious event. The general consensus was, however, that the paper is a nice description of important results and should be sent out for publication soon. ** Thanks. Here are our comments: Abstract: Replace "...set limits using .." by "...discuss...". The interesting part of the paper (related to SUSY) is not to produce some numbers/limits, but to discuss the findings with two exemplary SUSY scenarios. ** We agree that we have, at some level, discussed the our results within the context of two SUSY scenarios. However, this discussion is really only two pronged. First, we show that these models would give events which we should see, and we don't. Second we set limits on them. In any case, for simplicity, we chose to use the phrase "set limits" in the abstract. It really isn't a full discussion like in a theory paper and we didn't want to mislead any readers, especially theorists. Page 4, Line 9 Replace "...an eggMET candidate event .." by something more descriptive, like in the abstract, e.g. "..."an event with two photon candidates and unusually large MET ..." ** An interesting idea. Unfortunately, this paper is still over the line limit, so we simply don't have room to discuss the event more here. It is discussed in more detail later the meaning becomes clear shortly. Furthermore, we wanted to let the reader know right off which event we are referring to by using the nomenclature in the reference. Page 4, Line 11 "anomalous gamma_gamma events ..." : in this context, anomalous to what? What you mean is 'non-SM' events or those 'with a small SM cross-section'. ** Both are true. We did a counting experiment, so we looked for anomalous cross sections. We also looked at the events for evidence that they were non-SM in production. The eeggmett event appears to be anomalous in both respects, and at some level we are looking for others like it, albeit in a much more sophisticated and systematic manner. Page 4, Line 19 '.... SUSY models ... have arisen to explain it'. That sounds too strong; as if the SUSY models came up to describe THAT event. A alternative way would be to drop the part starting with "to set limits ..." and simply write "to discuss two SUSY models". ** I think the history is mixed. Clearly SUSY was not developed to describe that event, however a new (or different depending on your perspective) breed of models were studied and written about because of the observation of the event. Also, many older theories (e.g. the gauge mediated models, or Gordy's N2->GammaN1) regained favor because of the event. I think the distinction here between model and theory is important. SUSY is a theory. The models often refer to the parameter space, and people have studied new parameter space models to explain the event. Furthermore, I think it would be disingenuous to say we are doing all of SUSY, we are only looking at models which specifically refer to the event in the text of the papers (many of which cite it as a motivation or natural consequence). Page 4, Line 2 Replace "relevant aspects of the CDF detector" by "parts of the CDF detector relevant to this analysis" (otherwise, there wouldn't have been the necessity to build the irrelevant parts of the detector :-) ) ** While you are quite correct, we think that the sentence is clear especially in context. Also, we don't want to add more words if we can at all help it. Page 5, last Paragraph What do you mean by "resolution on .. the x or y component of MET...". Which coordinate system do you refer to? Why do you cite only the x-component? It may not be very clear to the reader why you stress the parameterization of the MET resolution in x and y. ** Met cannot be parameterized as a gaussian resolution. We have already defined the coordinate system in Reference 3. We only cite the x-component because the resolution is gaussian in only linear directions. We chose the direction to be x. It could just as easily have been y. This is exactly what is done in previous papers (e.g. W-mass, R, W'->enu etc). The resolution is given for two reasons, 1) it tells the reader how well the detector measures the MET and 2) it is the parameterization we use to estimate the MET for the sample in a later paragraph. Page 5, last Paragraph Replace "discovery" by "analyses" ** The references are the discovery papers. Page 7 Include the c in 49+-2 Gev/c, to be consistent. ** Done Page 8, Line 10 You use 1009 events to give a precise (0.3%) estimate for the phi mismatch. With this few events it seems impossible to obtain such a result from a non-gaussian tail. ** This is described in CDF Note 4304. The idea is that we have 1009 event, none of which have a |Deltaphi|>30mrad. We don't extrapolate using a tail, it's a straight counting experiment. 0 event observed -> 3 at 95%CL. 3/1009 = 0.3%. This is an upper limit. Page 8, 2nd Paragraph You certainly didn't check ALL possible scenarios which could lead to such a signature due to mismeasured/fake events. One has to be careful with such tiny rates in the oder of 10^-8 - 10^-6; other kinds of mismeasured SM events might contribute. More honest would be to start with something like, "We have investigated various sources which could lead to ..." ** We checked all the scenarios we could think of very carefully. This is described in great gory detail in CDF Note 3892. We looked at every kind of mismeasurement we could think of. We even looked at lots of sources with very tiny contributions. I think the paragraph uses the right word, we have "estimated" the rate and we list the types of sources we looked at as well as the rates of the dominant ones. We think the paragraph is very honest. Page 8, Last Paragraph Replace '...to set limits ..." by "...to discuss.." (see comments above) ** Again we disagree. (See comments above) Table I ------- Where do these SM expectations come from? It would add useful information to the table if you'd say in a sentence or two about which kinds of 'diphoton' analyses/physics (in refs. 15,16) lead to these expectations. ** I think every number is covered in the references or in the text. The MET is from the resolution (in text), the NJET is from the exponential extrapolation (in text), the b-tag is the Method-one background estimate (in the references), the gamma is very similar to the purity estimate, the dominant sources of e's and mu's are in the text as Z\gamma, and the dominant background for taus is the same as in the tau-dilepton PRL (i.e, fakes) as in the reference. Merry Xmas! ** Thanks, and same to you! Dirk Neuberger for the Karlsruhe Group