The thrust distribution, which is obtained in the 2f analysis, shows good agreement with MC in the region dominated by W’s. This is in contrast with WW and in general other 4f analyses. Marie-Noelle has tracked down the difference as being related to the use of LCAL-SICAL energy flow objects. If these objects are not used in the thrust calculation the data are enhanced in the low thrust region. The events which moves from high to low thrust when LCAL-SICAL objects are not used have been scanned by Marie-Noelle and they seem to be Z return. It is envisaged to repeat the standard cross-section analysis without using LCAL-SICAL to see if this has any effect on the final number, or if this discrepancy is visible only at pre-selection level.

Franco has succeeded in reducing considerably the cosmic background in class 17. He is now cutting on the HCAL total energy and on ECAL wires. He has still few very spectacular events which are escaping his cuts. A routine based on the ECAL clusters distribution used by Marie-Noelle at Lep 1 times could be useful to kill the remaining contamination. In any case he presently gets only 5 events more (at 189 GeV) when using his present cuts. They are all clear cosmics by eyes. The cross section at 189 GeV is in good agreement with the energy-flow / NN based analysis. The value at higher energies (1999 data) have been somewhat increased after the reprocessing. Detailed checks under way.

Draft 2 is being released. The systematic error on 4q will include the recent work of Julian on bkg reweighting based on data/mc comparison at the Z.

The analytical tool developed by Andrea has been used to understand the kappa value in the DM/M vs DE/E relation to be used to computed the systematic due to uncertainty on the lep energy. By giving the same weight to every event a kappa value of 0.92 is obtained. A more realistic distribution, weighted by 1/(kinfit event error)**2 gives .84 . There is clear evidence that events in the tails (i.e. with large errors) have a rather spread value of kappa, indicating that the presently rather scattered values obtained by the various analyses could be affected by statistical fluctuations. For now kappa=1 (which seems anyway conservative) will be used.

Several checks performed by Helenka on the 3 classes, most systematic errors have been computed, others are under way. Numbers are final and we seem in a good shape for the paper. Please have a look to the transparencies for the details. We have to decide if we are going to describe the silver case (where the e/mu mass is used) in the e/mu or tau section. At present the second option is followed.

Again the numbers are final, a few final syst. errors still to be computed. It has been pointed out by Oliver that, as explained in the aleph note, the difference between 1d and 3d fits is statistically compatible at 1.7 sigma level.

7) Lvqq width (Barbara)

Barbara is doing a really detailed work on detector systematics related to leptons, this is going to be used in all semileptonic analyses in the paper. She is/will add also jet related systematics. Most errors are small and some seem dominated by a statistical component. For the LEP energy both shifts and resolutions have been added. Fragmentation systematics under evaluation (this is a worrying point for the width, see below).

Also Roger is close to final numbers for the systematics. As metioned above the largest is the error from jetset-herwig comparison, yielding -159+/- 89 MeV at present. Any effort to reduce the statistical error on the evaluation of the jetset-herwig is welcome (need much more MC ? Error could be reduced with different treatment ?). As far as CR systematics are concerned, contrary to the mass, the SK2 model is giving the largest shift (70 MeV).

Also here very close to final . Detector systematics are done. Fragmentation (jetset/herwig) is now 30 MeV, was 18 MeV very recently, new MC statistics was added … checks ongoing.

The expected statistical error has been computed with 41 MC samples same size as data, yielding a result compatible with the error found in the data analysis. Mean is 0 and pull is 1. Possible jet angular biases have been tested by using preliminary angular shifts based on the photon-tracks and photon-hadrons comparisons (see Roberto at last meeting). The largest shift observed is of 35 MeV, an order of magnitude smaller than the present disagreement with the LEPECAL w.g. value. For the mass the same angular shifts are giving negligible shift in the central value. A final proposal for angular shifts is being worked out by Roberto, using a function inspired by the larger statistics 1994 data analysis. Work in progress stay tuned …… Eugeni is computing the possible effect of differences in the barion jet content. The analysis is very close to final, the feeling at the meeting is that a section on this work would fit in the forthcoming W mass at 189 GeV paper.

The 3 methods have been reviewed by Djamel (see minutes of W meeting of 24/11). The methods based on fixed reference still suffers from the large lever arm between the fitted value and the reference, even if a new reference at 82.85 has been tried. The 3^rd method (MLP/Neural Network based) has problems realted to the boundaries, again due to large numerical values (and errors…) being fitted. It has been decided to concentrate on the 2^nd techinque which is at present the more reliable. 183 GeV data have been included in the analysis. Systematics have been computed. 183 and 189 together will go to the paper.

Good news : a draft 0 (incomplete) exists ! We encourage the editors to continue this effort and the various contributors to submit their parts. DEALINE for giving final numbers from the various analyses to Frederic (for the combination) is 7^th of January (definition of final :"what is going to be presented at the Thursday meeting of 13^th of January"). Small last minute shifts, additions for the paper will be accepted, but the combination needs some time, so …. match the deadline !

From Monte Carlo studies the individual likelihoods from semileptonics e and mu seems in good shape….. however when several likelihoods are added together a shift is observed in the resulting likelihood, both for electrons and muons. The shift disappears when e and mu are treated together. Something need to be understood ! The comparison with the O1 analysis works well. Still some work is needed.

Results on CP-violating TGC’s very interesting. Some concern about one coupling (g^z_5 which is not cp-violating….) which seems sensitive to acceptance effects related to missing momentum cuts. An improved treatment is being worked out, but the backup solution of using an analysis as presented in Siena, and without g^z_5, exists. Otherwise the analysis is pretty final.

The final systematic errors at 189 GeV for the cross section are under evaluation. The effect of the ECAL calib seems pretty large, to be understood. This part should be final early in January. TGCFITTER is much more reliable now and very promising for the future. A bug in the DKg evaluation given at Tampere has been found. Luckily enough we were conservative ….. unfortunately even if the present value gives a better lower limit, this is dominated by the presence of WW bkg…….

Well, a draft 0 exists. However the analysis work is not really completed (see above) so the decision is to submit the paper to the collaboration not on the 16^th of Dec, but on the 13^th of January (this is going to be a loooong Thursday meeting …….)

A new software tool has been for WW analyses has been presented by Anne. It would produce in a single structure 4q, semilept and lvlv n-tples for all analyses (x-section, w mass, tgc). It is felt that this unified code will save a lot of time and improve the reliability of our analyses, therefore everybody is encouraged to help Anne (a long list of collaborator is already there !!) in order to accomplish this project. For further details see Software tool proposal