New interim spreadsheet: here.

Has the Censure tweak. Slightly improved autoreforge (Hitcap first, Expcap first, picks the superior). Still never going to reforge as well as a dedicated tool.

Looking at the top ret on Shannox 25H, who happens to be in 390 gear: World of Logs - Real Time Raid Analysis

Glyph of Exorcism gave him 53,749 extra damage.

He did 725,737 damage with Crusader Strike and 114,631 damage with the T12 2pc bonus, for a combined total of 840,368 damage. 31% of his Crusader Strikes were crits.

For the sake of simplicity, let's assume that 0% of his Crusader Strikes were crits. In this case, a 5% bonus to crit chance equates to a 5% bonus to damage done (when considered over a sufficiently long sample size, of course). Glyph of CS would have given him an extra 5% damage, which is 42,018 damage. This bonus damage is less than the 53,749 damage that Glyph of Exorcism gave him: Glyph of Exorcism wins. Since in reality he crit 31% of the time and not 0%, Glyph of CS would have given him considerably less than 42,018 damage, so Glyph of Exorcism wins by even more.

The logs for the #2, 3, 4, and 5 rets favor Glyph of Exorcism over Glyph of CS even more than this.

Correct me if I am wrong, but would it be better to assess the glyph of CS in this manner:

Take the total number of CS (hits + crits) then take 5% of that total number, subtract it from hits and add it to crits, then take his original mean crit damage and then multiply it by the adjusted number of crits performed.

Would that provide a better estimation of the CS glyph?

Thank you all for your time.

You also have to account the bonus damage of our mastery if you want to compare GoCS with GoExo.

T13 set bonuses:
Retribution, 2P -- Your Judgement ability has a 50% chance to generate 1 Holy Power.
Retribution, 4P -- While Zealotry is active your abilities deal 12% more damage.

2p doesn't sound amazing but 4p is very nice.

The wording is somewhat odd on the 4p. Is there any precedent as to whether an effect like that would affect auto attack and passive damage as well as "abilities", or is it really just for active abilities?

You do have to consider what 4.3 changes they are planning to make, specifically the blue post that mentioned swapping damage from HoW into other abilities. Looking at Prot's 2pc bonus (Your Judgement ability now also grants a physical absorb shield equal to 30% of the damage it dealt), it is possible that Judgement will see an increase in damage.

Assumptions aside, it is currently unlikely that it will chance our rotation much, except pushing Judgement up in priority possibly. We will have to wait on the 4.3 class changes before trying to crunch out any new priorities.

No; the more random elements there are, the less variance there is in the whole. Think about it in terms of coinflips. If you have two coins, your chance of getting 0 heads or 2 heads is 1 in 4 each way and you have a 1 in 2 chance of getting one head. Add a third coin, and your chance of getting 0 or 3 heads is 1 in 8 each way while you have a 6:8 chance to get one or two heads. At four coins, the odds of hitting the extremes become 1 in 16 each way and you have a 14 in 16 chance of getting one to three heads.

Adding more chances to proc will increase the number of steps from low to high on our scale, but make it more likely that we'll be somewhere in the centre of the range. Our spikiest damage will occur less frequently, but it will also become less likely that we do pathetic damage because we are starved for procs.

Evaluating the bonus in terms of value: we seem to use Judgement 4-5 times per minute when not under Zealotry, so we can expect the bonus to be worth 2/3 of a TV per minute under typical circumstances - 1 TV per minute if we elect to push Judgement only for CS or TV.

Yes.

On a really large scale you are correct, but on a really large scale DP and AoW balance out with the same arguments you are using for T13 2p. The only difference is that 2p 50% chance, AoW is a 20% chance and DP is a 15% chance. Anything that has a chance to proc is by definition RNG.

Thinking about it a bit more the new 4p is also against previous Blizzard comments about moving away from stacking cooldowns. With 12% more dmg on zeal we'll, more than ever, want to stack GoaK/pots/AW/trinkets together with zeal.

While correct in terms of coin flips and statistics, it doesn't apply to Ret paladin design. We presently have a dice roll (Divine Purpose) and a wheel spin (AoW). To that you're adding a coin flip. Does the coin flip create less variance? Nope, because it's not more of the same. Judgement producing HP does not affect AoW in the slightest (well, unless the extra HP alters whether you TV or Exo on a given GCD). 1 extra HP does not impact a no-HP-cost TV/Inq, either. You're neither more, nor less, likely to proc DivPurp or AoW with this bonus.

The 'another coin flip' metaphor would be "Judgement has an x% chance to proc AoW" or "Judgement has an additional x% chance to proc Divine Purpose." It does neither, so it's just another semi-unrelated proc.

Let's say you consider TV and Exo to both be "our spikiest damage." Exo proc is not reduced by generating more HP (at most delayed a GCD if you earned a TV earlier) - no spike reduction there. Earning 1 more HP every ~16 seconds (50% * 8 second CD) reduces spike damage how? By my simple math that's an extra TV every 48 seconds. It's not exactly a wildly noticeable and unbalanced increase in spike damage, but it's absolutely not a reduction.

Our major spike damage period is under the effect of Zealotry and AW (again linked when you break 4 T12). If you even happen to have a GCD for Judge during this period (not always the case if DivPurp and AoW procs are favourable), the HP it could generate is simply lost when your next CS generates 3 and fills your HP bar. Again no increase in spike, but it doesn't decrease it either.

On the tier bonuses in general
I'm rather wondering if these are placeholder abilities, as the 4 piece tanking bonuses all confuse me. Why would the tanks be encouraged to pick up 4 piece for abilities which reduce overall raid damage (or increase healing in the case of DK), rather than select the absolute best items for their slots to maximize their own personal survivability. I know I'd rather have a tank who survived/was easier to heal than one who had a higher chance of exploding (raid wipe), but helped reduce raid damage on CD.

Additionally compare Prot 4 with Holy 4. Prot 4 increases frequency of Divine Guardian usage and increases its range to a whopping 100 yds enabling huge spread-out fights such as ones that vastly discourage proximity, while Holy 4 piece increases Holy Radiance healing which is only maximized if enough targets are close enough together, say in a stacked situation. These seem diametrically opposed to one another in utility. Either all fights encourage distance (Holy 4 becomes crap), all fights are stack-tastic (why add the distance rider on Prot 4 at all), or 50% of the time the Holy bonus is not of assistance (while all the other healing specs remain useful).

Finally, set bonus between the classes certainly looks unbalanced. Comparing a 50% chance to proc a TV to a flat 55% damage increase to Shadow Word: Death (10 sec CD spell) and 400 mastery for 15 seconds on use of Elemental Mastery (3 min CD) - none of the three seem even vaguely related. Such vastly different effects on such wildly different types of ability definitely appear to be something that would be extremely difficult to balance. The risk seems inordinate that a class would jump from middle of the pack to uber-supreme when completing 4 piece or the converse of a class slipping from above average to low when the raid completes 4 piece bonuses and theirs just doesn't measure up to the others.

This gets more into prot theorycraft, and may be outside the focus of this thread, but I think you're overestimating the effect of choosing to take 4-piece. Generally at least 2-3 pieces will be well-itemized, and the remaining pieces won't be terrible. The stats we don't want are hit/exp, and they don't dual-itemize those on tank gear. So at worst, you're taking a hit/mastery or hit/avoidance piece and reforging the hit to avoidance. You'd be making a very small survivability trade (~100 dodge or parry rating, at best) for an increase in raid survivability (more frequent raidwalls). It's a small enough that healers wouldn't be able to notice.

The benefit of the reduced cooldown will obviously depend on the encounter, but it's still a reasonable trade to consider. I'd argue that taking 4-piece has a smaller impact than a DPS druid shifting forms to use Tranquility, which happens all the time in progression.

The increased range does seem odd for most of the reasons you cited. I think it might be intentional though, which would suggest that we'll have some fairly spread-out fights in T13.

On-topic, the ret 2-piece confuses me simply because it suffers from the "Grand Crusader problem." They're tacking an additional effect on to a low-priority spell in an attempt to make it higher priority. In this case it's Holy Power on Judgement, which is clearly supposed to make you want to boost Judgement's priority. The problem is that it's a moderate chance on a long-cooldown spell that has significantly inferior damage to the alternatives (Exo, HoW). Getting to TV faster helps, of course, but because the filler magnitudes are almost as large as TV it's not a huge effect. So you're taking a guaranteed loss on individual casts by prioritizing J>Exo, and hoping that a 50% chance to get 1/3 of a TV is going to make up the difference. Without a buff to Judgement's damage, that seems unlikely.