This is a great analysis! I really like that there are so many options for legendaries and that not every single one is BIS in some situation. Sometimes it's cool to play around with a legendary just because it's 'cool' and I think they did a great job of adding a bunch like that, just like this one!
It looks like an interesting legendary for Resto as the 4 second convoke is often not needed and you end up with a ton of over healing. So having a 1 minute cooldown for half the spells might actually but a lot better. Especially in dungeons. Also if you fancy using your convoke for moonkin and balance affinity, you haven’t blown it for 2 minutes.
It would be handy for Guardian too as their convoke is often a bit weak, but unfortunately they have extremely powerful legendaries and this would not even be a contender for them.
It really needs to be some kind of viable risk/reward to play, because the other restodruid raid legendary is boring as hell- just heal as per normal and sometimes you will proc some free rejuvs.This Lego needs to be worse than that if you don’t press it enough, somewhere comparable if you press it on CD, and better if you time the use skilfully, trading number or uses for timing of uses. A more predictable/reliable flourish would do that, and be amazing eg for mythic council, where holding it for the Dancing Fever to smooth out your cool downs would be perfect skillful use for better results.
Wonder if this will be good in m+ as well. If it casts Flourish this often, it might be pretty good... But then I'd be giving up Verdant Infusion. Any suggestions/ideas?
I suspect what's going on here is essentially a three-card deck getting reset/reshuffled every two cards.This kind of thing is a common implementation-dependant artifact of "stratification" (in jargon, the "successes" should appear distributed like a stratified uniform random sample). It can be programmed in a way that works "reasonably" well with any fraction out of any sized "deck", even if that doesn't end up with a whole number, like "1/10 of 64". So this might have been someone changing the probability from 0.5 to 0.33 without changing the stride from 2 to 3. With bigger "decks", that sort of incongruity would be insignificant, but a "1/3 of 2" might just behave like what's described.
Not sure if I love the phrasing of
There was still a lot more thought put into this druid legendary than nearly any of say the monk ones for example. So at least a developer put some effort into this one, sorry it wasn't as OP as desired. Still seems useful for resto to weave in more often for M+.
I have been using a totally different Lego than what I am supposed to be using. At first I tried the recommended Lego, but I didn't like how it worked. It didn't suit my needs (or style maybe?) so I tried others, until I found one that matched my healing style. I am still not sure if replacing it with this one would be worthy, but I think it might. I guess I will have to try it. I am just sorry that I couldn't test ptr as much as I wanted.
You deserve praise for making such a large effort in this theorycrafting. I feel that theorycrafting as a whole has diminished somewhat over the years, and this is a great example of how it should be done, i.e. with peer review.
Is the test data publicly available? This is great work and I would love to dig through the sequences to compare it to other potential deck systems like the deck of 3 that resets after 2 pulls that was suggested earlier
For everyone who hasnt understood it yet, it is because we lost 25% of duration on convoke with the legendary.These 25% less duration is what the poster found in their analysis, as you might GET a positive card, but that Spell would be in the missing 25% duration.The post goes for "70%" chance, but the actual chance is 75% as per duration we lost from the legendary.So indeed you could possibly trigger a positive deck, but don't actually cast the spell, because we don't cast a full Convoke and only 75% of one, which makes it so the system things you got the spell eventho it died in the 25% that we weren't allowed to cast.