If it was the four ball test you were referring to, it likes higher viscosity and film strength for "better results". It is a hydrodynamic test, mocking a plain bearing con rod under severe loads.
As I think I understand it, the four balls go into continuous contact under a variable (but increasing) load until the film strength collapses at some known point. Since the balls are in continous contact, ZDDP has no chance to replace its boundary layer and as such does worst in this test than in others.
ZDDP for example works best on tappets, which are occasionally loaded to film failure with time to recover between the load induced failures. This is barrier film activity and really has no test rig other than lots & lots of miles inside an in-use engine.
Since you get to pick the 4 ball loading top point, you can stop the loading increase at a level where your Amisol does well and the others start to fail. You simply avoid testing oils that have a higher yeild point than your oil does and you don't mention the test you are using commonly gets used in testing heavy gear oils and that the rotational loading you are using are ridiculously high compared to anything oil inside a roller bearing Savage crank assembly will ever see.
Above inferences sourced from BITOG's information on 4 ball testing. Note that simple old Shaffers dino does well on this test, more so than most synthetics. Why, I haven't got a clue. Note that BITOG refers to it as a "snake oil test", once again the why isn't completely known. Thick viscosity additives (Lucas, STP) like this test as they do very well on it, showing a clear before and after difference.
http://www.bobistheoilguy.com/oilshear.htmNow, apart from Amisol's "snake oil" 4 ball testing, this is the way the oil industry tests shear down in the real tests that get published by them.
The shear stability of an oil is measured by using both ASTM test methods D445 and D5275. First, the viscosity of an engine oil is measured. Then, the oil is exposed to severe shearing conditions by repeatedly pumping it through a specially-sized diesel fuel injection nozzle at high pressure. After shearing the oil, its viscosity is measured again. The percentage of viscosity lost is determined by comparing the second viscosity measurement with the original viscosity measurement. I find these differences in test methods to be educational, don't you?