@The Stranger, I must say you got very interesting results with your tests. You even got some cost-effectiveness out of spartans. I tested them as well, and if anything I found them to be one of the most overpriced, cost-inefficient units in the game.

If you really want a laugh, take a unit of elite dacian skirmishers and throw them at a single unit of triari. Seeing the price of the elite dacian skirmishers for the very first time caused me to literally spray coffee on my monitor, but at any rate a single unit of triari, not even adjusted for cost (triari are significantly cheaper) beats an elite dacian skirmisher unit. Or, I should say, with my testing this happened. With your testing I wouldn't be surprised if you came back and said it was a slaughterfest the other way around.

I'm not sure why you are getting different results. Your methodology seem appropriate enough. I wonder if unit sizes has anything to do with it (my testing was always on 'normal' sizes)? For the most part, I was never challenged by others on the cost-ineffectiveness of elite units. Everyone seemed to agree on that, more or less. The debate always seemed to be about other things. For instance, folks said it had to be set up that way to keep the player from spamming elites and playing ahistorically.

At any rate, if you can get different results and better cost-effectiveness out of elites than I can, then good for you, and more power to you. Maybe you just have a magic touch.

@Kull, of course 1 levy will always lose to 1 elite. You have to test "cost for cost," i.e. throw the same amount of cost at the elites, which means you might have to throw 2, 3, 4, or more units of levies at one elite unit.

Whether that proves it's "worth it" to tech up is debateable since good tactical play will almost always offset unit quality differences (as the EB quote says, Army of sheep led by a lion beats an army of lions led by a sheep
Of course good tactical play with a crappier unit can offset the advantage of a superior unit. The question is whether you can improve your good tactical play even more by using cost-effective vs. cost-ineffective units (i.e. good tactical play with cost-effective units should trump good tactical play with cost-ineffective units).

Of course while testing this stuff you don't want to use good tactical play to offset anything as that will skew the result. You pretty much just want to throw the units at each other and see what happens. The way The Stranger has been testing it seems fine, as he just does a generic surround and then "hands off" for one test, and then does an additional chokepoint test.

Anyway, thanks for testing, Stranger. And remember, whatever doesn't kill you makes you... stranger, heh.