The transparencies presented at the workshop are available on the web.

Paolo gave a summary of what was discussed in Lisbon w.r.t. the measurement of the LEP beam energy using radiative return events. Details can be found on the transparencies from the Lisbon workshop. In short, the 4 LEP experiments have done the analysis similar to ALEPH using qq(g ) events and found consistent low values for the LEP beam energy (A: -153 ± 40 MeV, D: -158 ± 49 MeV, L: -45 ± 61 MeV, O: -167 ± 43 MeV ) where the errors quoted are only statistical Þ LEP: -143± 23 MeV !!!

Most of the discussion in Lisbon was devoted to understand the systematic errors in this measurement. It's clear that an important source of systematic error is the uncertainty on the jet direction in absolute terms. In this sense, the knowledge of the ratio between the length and the radius of the detector seems to be a particular concern in DELPHI and dominates what they consider as systematic error (» 94 MeV). In ALEPH this systematic error has been found to be small, but we also quote a large systematic error (20-50 MeV) from the bias in the jet angular measurement. On the other hand, it's difficult to imagine that these sources of error are highly correlated between experiments!!!

In summary, the preliminary evaluation of systematic errors quoted range between 50 MeV to 94 MeV (DELPHI being the most conservative).

ALEPH, DELPHI and OPAL have had a look also to the m m (g ) channel, (OPAL even look at the t t (g ) channel), and the results are: (A:-288 ± 147 MeV, D: :+80 ± 102 MeV, O:+51 ± 94 MeV) where the errors quoted are only statistics. It needs to be understood why ALEPH has a much larger error in this channel!!!. If combined, LEP: 0 ± 63 MeV , compatible with the LEPECAL value but also compatible with the determination from the qq(g ) channel.

Coming back to the jet angular uncertainty, Andrea Venturi has tried to evaluate what would be the effect on the determination of the LEP beam energy if we were correcting for the discrepancies observed and not asigning an uncertainty of 20 to 50 MeV to it. From his studies, if corrections were applied to the MC we will determine an even lower LEP beam energy by 20 to 50 MeV!!!

Paolo has also look at non radiative dimuon and Bhabha events and reconstructed their invariant mass. Comparison with MC predictions shows lower values for the position of the peak of the distribution in data than in MC.

Jason gave a summary of the discussions about the W mass measurement in Lisbon. Details can be found on the transparencies from the Lisbon workshop. Some ideas come up from the Workshop and Jason introduced them for further discussion in ALEPH:

• Produce MC events with the 4 sets of tuning used at LEP and compare distributions. Are the experiments consistent on the QCD part of the MC?… In the discussion it wasn't clear what can be learnt from this study that we don't know already from the studies from Stan Thomson and Gerald Rudolph, but as DELPHI has volunteered to do the work…
• Fast simulation studies of the 4 LEP detectors: What part of the fragmentation error is due to detector response? And how is correlated between experiments?… Ann Moutoussi is working on this comparing ALEPH and OPAL.
• High statistics comparisons of different MC models… ARIADNE vs JETSET (promised by OPAL), HERWIG vs JETSET (ALEPH & OPAL).
• What is the role of the jet correction parameterisations in what we call fragmentation error? The proposal is to compile this numbers from the four LEP experiments to compare them… again from the discussion it wasn't clear what can be learn.

Oliver gave a summary of the discussions about the FSI measurements in Lisbon. Details can be found on the transparencies from the Lisbon workshop. The workshop was particularly useful due to the interaction with the theoretical community (mainly with Lonbland). From these discussions it become clear the need to tune MC models for BE and CR at the Z peak, requiring to tune also the QCD parameters, if this models may have any meaning at all. This phenomenological models should work at the Z if they have to work for WW physics.

For the BE models this is what the four LEP experiments are trying to do, trying to find a global tuning including QCD parameters that works for all experiments. Although, conceptually this is the right way to go, it may take a long time and it's not clear this procedure will work (see past experience at LEP1).

For the CR models, in particular for the JETSET models, the situation is very different. The JETSET-SK1 model that gives the largest effect for MW cannot be tuned at the Z peak, because does not work for Z physics!!! Lonbland recommended to use a different implementation of a similar concept, that it works for Z physics (see http://www3.tsl.uu.se/~rathsman/gal/ ). We should try this model in ALEPH.

So… when saying things like:" L3 discard BE between W's at 5 sigmas" one needs to be aware that what is discarded is the particular implementation of LUBOEI with the L3 tunning wich is quite different from ALEPH and DELPHI, for instance the values of Lambda for the BE3 implementation are ( A: 2.1± 0.1 , D: 2.0± 0.1 and L:2.9± 0.1 )…

Some progress has been achieved toward the possibility to have a combination of the Particle Flow analysis. It has been shown that there are correlations between the bins in the rescaled angle distribution, and this has been neglected so far by the other 3 LEP collaborations. The most outstanding problem is, however, the incompatibility of the results obtained using the implementation of the SK1 model implemented in ALEPH w.r.t. the one implemented in the other LEP experiments. See the transparencies for all the checks performed …

Andrea gave a summary of the discussions on the new O(a ) MC generators comparisons. OPAL has been able to generate events with YFSWW and compare the effect on the full analysis, while DELPHI has been working at generator level comparing RacoonWW and YFSWW with our standard MC generator for WW physics, KORALWW.

 Roberto Chierici, from DELPHI, show how at generator level YFSWW and RacoonWW agree very well, not only for the total cross-section (better than 0.4%) but also on the differential cross-section (predicting large corrections w.r.t. KORALW at cos(q w-) ~ 1). This has a large impact on the determination of the TGCs as shown by Eric Torrence, from OPAL, who compared the effect of using KORALW or YFSWW at fully simulated level. The biases found are significative and comparable to OPAL total error!!!, (D k g ~ 12 ± 4 %, D g1 ~ 3± 1 %, l ~ 4 ± 1 %). The small remaining differences between YFSWW and RacoonWW are related to the spectra of the photon emission.

Another very interesting result shown by Eric Torrence, was the effect on the CC03 selection efficiency. The effects are found to be at the level of +0.2± 0.1 %, increasing the overall efficiency with the new MC predictions, but no large effects seen (th. Uncertainty ~ 0.4% and exp. Systematics ~ 0.8%). The effect on Mw is measured to be very small at generator level (~ 5 MeV), and no significative bias at reconstructed level due to the MC statistics uncertainty (~20 MeV). Placzek announced that he's working on a reweighting routine that can work on KORALW events to include most of the effect of the O(a ) corrections, and allowing us to reuse most of the MC we have already generated.

Jorgen first described the status of the TGC paper up to 189 GeV energies. Draft #3 will be released end of November, and the Ed. Board is being scheduled mid of December.

Jorgen, then , gave a short summary of the discussions on TGCs at Lisbon. New results have been shown for Quartic Gauge Couplings and Neutral Gauge Couplings. The main discussion on these channels is related with the proper definition. Not much worries about systematics as they are statistically limited. Most of the discussion, though, was devoted to the Charged TGCs as systematics are now at the same level as the statistics uncertainty at LEP. The systematic error at LEP is dominated by the comparison between HERWIG and JETSET for the fragmentation error, assumed to be 100% correlated. On the other hand, it has been shown that the method used to combine the LEP experiments is not appropiate and bias the measurements by significative amounts (~ statistical uncertainty). The ALEPH method, considered to be the most rigorous, combining the results at the level of Optimal Observables will be tried to combine ALEPH and OPAL, as both experiments use Optimal Observables.

Moreover, the large effects seen due to the new MC available has not yet been taken into account in the results shown by LEP. If this corrections were applied naively now, we will have a 3s "observation" of anomalous couplings at LEP.

Renaud had a first look at the Y2k data. The results shown are based on 80.1 pb-1 at 205 GeV and 133.7 pb-1 at 207 GeV. The tau analysis still uses the old reconstruction, and the NNs used for semileptonic and hadronic channels have been trained at 196 GeV. Results are in agreemet with SM expectations. A first study on the dependence with the C.o.M. energy of the reference sample indicates that may not be negligible. Further studies are in progress.