“Those are correct. I am does HC with this gear.”
The wow website itself and warcraft logs both only show at best 7/10 LFR bosses killed.
Regardless lets talk about why simulating healers is (too) complex and very situational, if at all useful.
Lets prephase this by saying that simulating or creating a model of anything is possible, that does not however mean every model is either useful or a correct or predictive model of reality.
So lets start with the goal of healing a raid boss encounter in WoW, the most basic goal of healing a raid boss encounter in WoW is to contribute in such a way as to ensure the boss dies. How do we contribute to that as healers?
- We heal people to prevent them from dieing.
- We deal damage to help kill the boss
- We use utility to directly reduce incoming damage.
- We use utility or actions that do not fit into the above points, such as doing a mechanic, giving someone a movement speed buff etc
Point 4 is really hard to simulate (although technically not impossible) or evaluate from a numbers perspective so we’ll ignore that for now.
So what do we want to gain from simulating an encounter; Advise on how we can best contribute in the above mentioned points in the things we can control. That means; Gearing choices, talent and conduit choices, casting choices, consumable choices.
Now lets start with the few knowns we have;
- The effect(s) of our spells individually
- Our current or baseline stats
- Our current or baseline gearing choices and talent choices
What other data do we need to come to a meaningful simulation?
- The amount of allies we have and the way they will (or should) act
- The damage pattern of the encounter
- The duration of the encounter
- Other specific requirements of the encounter.
Now we start running into issues, how can we know what our raid composition is like and what our other healers are doing? How do we factor in our own capabilities as a player, what is the duration of the boss fight? What is the damage pattern? Is it a lot of spot healing evenly spread throughout the fight or is it a few big spikes of damage? Or maybe there is constant raid damage? A combination of all those.
Do we have periods where we need to move and can only rely on certain abilities? Maybe we don’t have to move at all?
This is the first and one of the biggest issues of simulating a boss as healers; We have to make guesses or assumptions about all these things, that means that if those assumptions don’t match reality (closely) our simulation is already worthless or severly diminished in value.
If the input data is incorrect then the simulation loses value
Now we get to the next issue: What metric do we use to estimate our contribution? For damage it is doable; generally a set amount of damage has to be dealt within certain timeframes (dps) and if we can optimize damage in these timeframes we are fine.
For healing we have to match the specific damage patterns and it gets a lot more complex, we can not simply rely on overall average hps; Although all healing contributes to that metric it does not tell us what the healing did or how useful it was, if we heal 5 people at 70% health that are not going to be damaged and are standing in a shaman’s healing rain our healing is essentially worthless, maybe we could have dpsed instead or maybe there was a target with a heavy dot that now died because we did not heal them. We followed the highest hps option but we did not succeed in our primary goal of helping kill the boss. This leads to the following conclusion:
Not all healing is created equal and HPS is not a definitive metric of our effectiveness in an environment
What other metric can we use? Certainly dps can be a part of the equation but that is not our main job as a healer. The average amount of people that (should) die because of the way we and our cohealers heal in unison with the effects of both the environment and the actions of our other allies ? The answer is not a very easy one.
Now we can try to simulate all these things, tune all the assumptions and data so it correctly fits our specific personal and situational needs and then when we have done that herculean task we still only get relevant data for that one specific encounter.
Even if we succeed in that we should then not use that specific simulation we have created to give general advise to someone else, because the input of our simulation will most likely not match their specific situation. And we should especially not provide a metric that is as misleading as HPS.
So how do we help other players and get meaningful information on how to act?
- Look at which spells are good in a simple situation based on mana efficiency and their healing throughput.
- Look at what our needs are for specific fights based on the composition of our raid and other environmental factors.
- Look at our abilities and things we can change to see how they will effect how we act in certain fights.
Most of these are trivial calculations some are harder, we do not need complex simulations because we have perfectly fine tools to help ourselves and others out: Logs and tools that calculate the effect our actions and choices had in those logs (wow analyzer).
With these tools and by looking at the wealth of data we have available we can make informed decisions about what is effective in the actual environments based on achieved results and can give advise tailored to our and other people their specific needs.
So next time someone asks for advice do not give them an arbitrary hps number based on arbitrary decisions in a healing simulator that might not fit their situation at all or is just incorrect, instead take a look at their logs, run them through wow analyzer and give them advise tailored to their specific needs.
So how do we give general advise to people? We use a combination of achieved results and parses and calculations of the effects our different tools have.