Optimising a Cut Off

If you sort by score, you can calculate how many of each type are removed for any given threshold:

	A	B	C	D	E	F	G	H
1					26	9		E1: =COUNTIF($B$2:$B$37, E2)
2	ID	Type	Score		A	B
3	id23	A	-227		-25	-9		E3: =COUNTIF($B$2:$B3, E$2) - E$1
4	id34	A	-207		-24	-9
5	id1	A	-195		-23	-9
6	id30	B	-187		-23	-8
7	id31	A	-178		-22	-8
8	id15	A	-111		-21	-8
9	id21	B	-84		-21	-7
10	id28	B	-60		-21	-6
11	id3	A	-59		-20	-6
12	id9	B	-27		-20	-5
13	id26	A	6		-19	-5
14	id6	A	44		-18	-5
15	id13	A	73		-17	-5
16	id29	A	90		-16	-5
17	id12	A	130		-15	-5
18	id7	A	147		-14	-5
19	id16	B	152		-14	-4
20	id24	A	180		-13	-4
21	id10	A	184		-12	-4
22	id32	A	190		-11	-4
23	id19	A	196		-10	-4
24	id22	B	212		-10	-3
25	id27	A	214		-9	-3
26	id4	B	219		-9	-2
27	id11	A	258		-8	-2
28	id18	A	262		-7	-2
29	id20	B	263		-7	-1
30	id8	A	346		-6	-1
31	id2	A	387		-5	-1
32	id17	A	398		-4	-1
33	id14	A	436		-3	-1
34	id25	A	453		-2	-1
35	id33	A	457		-1	-1
36	id35	B	475		-1	0
37	id5	A	481		0	0

At 180, for example, you cut 13 A's and 4 B's. I don't understand your figure of merit, so I don't know what you do with that.