But my question was, why isn't 100(anti-body) :200(buffer) not a ratio of 1:2, why is it 1:3? When we do higher dilutions example, 1: 2000, in a 200ul buffer i am gonna put in 20ul to achieve that dilution.
The ratio of mixing of 1 part added to 2 parts gives you a final dilution of 1 part in 3 parts. If you had added 1 part to 3 parts you would end up with a final dilution of 1 part in 4 parts. I think there is some ambiguity in the way you are using symbols. Note I am assuming when you use a colon ":" as in x:y that you mean x parts added to y parts. When I use the slash character "/" as in x/y I'm using it in the sense of to divide x by y.
In your example above if you added 20 parts to 200 parts (20:200) then that would be a final dilution of 20 parts in 220 parts (20/220) which is a final dilution of 1 in 11! Clearly that is a long way from your desired 1:2000 i.e. you are at least two orders of magnitude out in your calculation.
If I wanted a dilution of 1 in 2000 then I would add 0.1 ul to 199.9 ul . What you could certainly argue is that the error in trying to pipette 199.9 with accuracy might be similar to just using 200 it's only going to result in a small error.
However you need to be cautious about accepting such errors. For simple biological experiments it doesn't usually make a significant difference. But if for example you were preparing serial dilutions, then the error becomes an exponential error with each additional step adding to the exponent.