With the advent of my paladin (alt, although I dont subscribe to the term alt) I've been doing some of the arbitrary reading recommended for a protection paladin class and ran into multiple discussions about how reckoning procs can cause an increased amount of parries from a mob (usually a boss), resulting in the swift demize of the paladin.
Now I am no wizard in theorycraft and in general I like to fly by the seat of my pants and a spot of reading here and there but if it actually is a problem that reckoning procs cause increased amount of parries that turn into a so-called dreaded insta-gib then I ask the simple question: How do warriors tank anything then?
The questions arises purely from the vantage point of the amount of parryable attacks each class is capable off. Warriors have a ton of abilities that can be parried, paladins have practically none since most damage is through 'passive' return damage or magic, like exorcism, seals, consecrate and shield returned damage.
None of these outside the standard paladin melee (auto)attack can be parried.
So unless reckoning speeds up your attack well beyond the attack rate of a warrior there's no reason to believe that haste induced parry gibs are actually a significant problem (or at least more of a warrior issue than a paladin one).
Much like any parry though it can increase spike damage. The more you hit a mob the more you increase the chance the mob will parry and thus the more damage you can receive over a percievable timespan.
In the meantime even a parry can be blocked, and since paladins were made to block every blocked parry means extra damage to the enemy and more importantly extra threat.
So it stands to reason that reckoning is not a bad thing, perhaps slightly dangerous in straight up single boss fights, but much more useful when the targets are numerous which they generally are.
But reckoning at 5/5 is expensive so I did some more looking around to see how much you could actually scrape off of the talent without losing the bulk of its effect.
And voila a bit of digging on maintankadin and I was faced with a nice little graph that shows reckoning uptime depending on the amount of targets. Turns out that there is virtually no difference between 4/5 and 5/5 reckoning in terms of uptime and 3/5 is really close to what you could hope to achieve as well.
2/5 however shows a significantly lower uptime and of course 1/5 is abysmall (but arguably better than nothing).
See the graphs here.
Looking at the most common 4-8 target range you could say that with 3/5 reckoning it would be up 50% of the time and with 4/5 reckoning about 65% of the time and only marginally more with 5/5.
So there's definitely call for dropping at least 1 point out of reckoning and if you don't mind an approximate 10% loss of reckoning uptime for your average group of mobs then you can easily go 3/5.
A godsend for those of us who want to take reckoning but can't find the points anymore to pick it up.