The different decay possibilities are identified in a similar way as in the exercise of the branching ratios of the Z0 particle. Unlike the Z0 decays, it is not one but two vector bosons that decays. This creates problems since it can be difficult to distinguish the decay particles from the two W particles from each other. For example, an electron from one W particle can end up going in the same direction a jet from the other W particle. In this case it will look like that the electron is part of the jet and we will have no clue of what happened to the first W particle. We would like to see single electrons, that is, signals in the EM-calorimeter that are isolated from all other particles.
The tau decays are especially complicated since the tau particle itself can decay into an electron or a muon (plus neutrinos). In these cases it is very difficult to deduce if the W particle decayed directly into e.g. an electron or if the decay passed through a tau.
In an easier variation of this exercise, only the two branching ratios for hadronic and leptonic decays are measured. That is, one does not distinguish the three different types of leptonic decays. In this way the problem of identifying the complicated tau decays are avoided.
Now do select a number of WW collisions and count the number of different decays. The calculations of the branching ratios are then simpler than was the with the Z0 decays (since we do not have any invisible neutrino decays). The branching ratios are simply the number of identified decays of a certain type divided by the total number of decays that have been analysed.
When you have calculated your branching ratios you can compare these with the predictions of the Standard Model, see the link below.