What! You mean Bill isn't the unbeatable player of legend who can shrug off such inconsequential things as the loss of a 400 point unit with ease?
To add a little more of substance to this post, I'd like to go back to what
GK Luke said a few posts back - what is the definition of a tier?
I believe that player skill on the tabletop is more important than how powerful a codex is, or even, how min-maxxed their list is. JWolf's gladiator list, Blackmoor's Foot Eldar, DarthDiggler's 'ard boy marines - lists that constantly draw comments like "you won with that" or "you couldn't have played anyone competitive" - show that there's more to it than taking the best mathhammer units.
So, how important are these theoretical tiers? I think that they really only come into play when comparing games played between equally skilled opponents.
I also don't think that it is necessarily correct to label Tiers by codex alone. I think it is more appropriate to consider list archetypes than just what book they're taken from.
Take C:
CSM as an example. Dual-Lash is still a top performer. But it's only one archtype from that codex. Most of the rest of that codex seems underwhelming in the face of subsequent releases. Does that mean that the whole codex is top-tier, because of one build? I don't think so. But does that mean you can't make a tier-one army out of C:
CSM - clearly you still can.
Some codexes have multiple tough builds. Space Wolves, for example, can field an extremely shooty mech list, or the thunderwolf cav heavy-assault version. Does their ability to field more than one competitive build mean that they're a higher tier?
There's also the
RPS effect to consider. I've seen Dark Eldar played very well in 5th ed, and they've got good matchups against several armies. But leafblower guard has their number.
Maybe that's a good baseline for deciding what tiers something is. First, assume equal player competence.
A top tier army is going to have no bad matchups, and at least one (or maybe more) easy wins from lower tier armies.
A 2nd tier army might have a bad matchup against a 1st tier army, but no outright bad matchups against other 2nd tier armies. And so on.
I still think that this is far more relevant when discussing army archtypes though, rather than codexes.