That one is the correct one. The opposite, or pure adamantine, is the common misconception.
I tested this one some in arena mode.
Steel > Adamantine > Iron
It doesn't seem to matter what kind of weapon the testing dwarves are armed with. Silver warhammers, iron battleaxes, silver whips - all about the same. Everyone got grandmaster skills, a candy chain shirt, a copper shield, and then either steel or candy hard armor consisting of two boots, two gauntlets, a helm, greaves and a breastplate.
I tried it again a few times with zero skill.
Adamantine ≈ Steel
What I think is going on here is that high armor user helps offset the weight of the steel. Remove that and the deathmatches between squads of ten are much closer and less predictable. I'd need to do a bunch more trials and make spreadsheets to prove it but I suspect that adamantine might actually be very slightly better than steel until they get to high levels.
So don't cry if you made a ton of either adamantine or steel armor.
This is incredibly strange and I want to begin investigations on how this is even a possibility. Steel's highest stats are impact and compressive fracture at 2,520,000. Its lowest strain at yield stats where lower is better is shear, torsion, and bending at 215. All of adamantine's yield and fracture stats are 5,000,000 twice that of steel's highest stats, most of steel's stats are lower. Then all of adamantine's strain at yield stats are at 0, which is definitely lower than 215. Now I could be wrong, but I had thought that armor effectiveness is based on all those shear, torsion, and whatever stats so how steel is possibly better than adamantine baffles me.