Bhabha rejection not used in the following: just to show the origin of the problem, at the final cut decrease of efficiency without important gain in background rejection.

Borrowed from ZZ analysis: background rejection (including non ZZ eeqq events) with a correlated cut on Mee and Mjet (4C fit) instead of Mjet(5C fit) cut.
Not as good for hee, but still useful ( plot ).
Comparison with old mass cut using optimization samples (except 70 GeV):
 

 
H 70 GeV
H 85 GeV
H 100 GeV
e+e-q qbar
q qbar
WWEX
Old cut
49.2%
56.6%
53.9%
4.86
0.37
0.53
+btag
41.3%
47.1%
46.2%
2.07
0.21
0.13
New cut
51.0%
56.1%
52.3%
4.38
0.32
0.40
+btag
43.0%
46.7%
44.7%
1.90
0.21
0.13

Efficiency on test sample:
 

 
70 GeV
75 GeV
80 GeV
85 GeV
90 GeV
92.5 GeV
95 GeV
97.5 GeV
100 GeV
Presel.
76.26
74.67
75.68
77.04
77.39
76.59
78.08
78.55
77.38
El. id.
63.90
63.81
63.96
63.15
65.08
63.51
66.40
67.15
64.92
El. en.
61.89
61.56
61.26
60.91
62.71
61.14
63.76
63.94
61.56
Fit prob.
56.13
56.91
57.01
56.54
58.24
57.51
59.73
59.37
56.10
Isol.
54.38
54.70
54.80
54.54
56.00
55.30
57.62
56.47
53.40
Corr.mass
50.98
52.85
53.35
53.54
55.17
54.70
57.06
55.27
51.67
B tag.
42.96
44.29
45.20
44.63
46.13
45.63
47.78
46.46
44.20

What about data and background?
 
 
Data Tot. bkg. q qbar l+l- WWEX e+e-q qbar llqq gam gam
Corr.mass 1 5.38 0.19 0 0.39 4.76 0.05 0
B tag. 0 1.79 0.03 0 0.13 1.63 0 0

 next page