Your theory doesn't even pass the Exercise test, if it doesn't work under 'laboratory conditions' it's not going to work in the real world. Your example uses units values that match workers for A (50 def, despite what tool tip says, and 100 life) and transports match B (10 attack). So 100 Bs would do 1000 damage which should kill 6 workers but actually kill 9.

When you say you can't post the link, does that mean your entire theory is based upon a single battle?

Just to make sure it's not me who's misinterpreting, is below how your theory works?

(Attack x Number of attackers) / (Def + Life) so AN/(D+L) = kills

example: 12500 warriors attacking cavalry

(50 x 12500) / (180 + 500) = 919 dead cavalry

Exercise result = 1025.

I've made my own calculator that incorporates tech & hero values, when these are set to zero the results match the results in the Exercise, they don't quite match real battles (for now!) but any theory has to agree with the Exercise otherwise it'll have no hope of working outside of the lab.