Background modeling in bbWW

Idea: use similar approach to HH-4b.

Nominal just uses MC

Instead use DvT fit in the sidebands.

  • Better background model: Account for MC mismoelling (eg: missing toponium)
  • smaller(?) simpler systematics: no need for MC systematics on ttbar MC
  • No need to skim looser than pt cut in ttbar MC

Can validate with synthetic data OR mW vs mW* cut

The 2d mWW vs mbb will only work for W->lnu, we use mWW constraint for W*->ln

Backlinks