Resto Shaman Gear Strategy

I ran a full sim yesterday using the Raid Healing script for resto shaman. The results are surprising. All other sources I can find indicate that the state weights should be as follows:

Leech = Avoidance
Intellect
Critical
Haste
Versatility = Mastery

The sims for Mythic + and the Raid Healing script keep pushing mastery as one of the top stats. Other than the argument that “our sims run different to others” why would it determine in a raid healing setting that Mastery would be top?. In our case, it is highly dependent on burst damage which isn’t very effective in a raid setting.

Here is the link to the sim. Maybe someone can explain in more detail how it comes to it’s conclusion.
https://www.askmrrobot.com/wow/simulator/report/e4d8104518e0431b9de832b239c057bf?weights=true&strategyId=fdc20d306e2048928bae

Thanks!

The gearing strategies automatically tune the fight to a point where it is difficult to keep everyone alive. In that situation, your allies’ health will get low enough to make mastery a powerful stat. I have no idea how these other people are determining which stats are best. The raid healing script we have right now does pretty heavy raid-wide damage from large boss abilities - and that is a very common boss mechanic/damage pattern.

Resto Shaman mastery is the ideal stat for a healer. To me, even from a napkin math, logic perspective, it is crazy not to get a lot of it if you are a resto shaman. You do more healing when you really need it the most. Looking at your results, crit is valued just as highly as mastery in most stat combos - which also makes sense since shaman have that mana return passive on crits. Shaman are very mediocre at topping off high health targets, and extremely good at topping off low health targets. Gearing to be even better at your strength just makes sense to me.

I’m also in the process of making a new raid healing script with some different damage patterns based on what I’m seeing for Uldir - but I really don’t expect the results to change significantly. You never know, though! My approach to healing theorycraft is to approach the problem with no bias. I model a realistic healing situation and then analyze the results. All the interactions are way too complicated to say what stats are best without a simulator, in my opinion.