Key Metrics

[Frogreaver]

One thing I've seen many recent threads for is comparing classes. Most are damage related but not all. Some have been about survivability. Etc. I don't think we can ever get all the class differences down to a single number.

So I propose a new framework, 'Key Metrics' where we work together to define a set of robust and definable metrics that we can use to evaluate characters. These metrics should not be subjective in themselves. They should be meaningful to gameplay. They don't have to account for everything, but hopefully they cover alot of ground.

Potential Key Metrics
Damage (at will damage, Nova damage no prebuff, nova damage prebuff, daily damage with adventure day structured as short, normal, long).
HP/Recovery (potential per day, per encounter)
AC
Saves
Skills (breakdown by skill and show max available bonus per skill?)
Movement (move speed in a turn using all actions and bonus actions, move speed in a turn using just your bonus action, move speed in a turn using just your action, move speed in a turn while disengaging, irregular movement types?)
Incomparables (Rage damage resistance, stunning strike, etc)
Spells (Unsure how to create a metric for these? Possibly just max spell slot level and level of highest spell known?)

What other metrics do you think do a great job of encapsulating and differentiating character performance?

[animorte]

I would think that various forms of utility should certainly be included in the conversation combat and otherwise. Being able to lock down opponents to win the fight much easier. Other methods through which to redirect the opponents and avoid the fight altogether. I'm not sure how you would stat that as some are direct interference of someone's mind while others focus on battlefield control.

Then we have the things such as Mending, Create or Destroy Water, Goodberry, Sending, Tongues, Commune, Scrying... just to name a few. I'll admit some of them can certainly be more skill-relevant. Others might fall under some form of natural resilience while different choices account for tactical preparation.

[Frogreaver]

I would think that various forms of utility should certainly be included in the conversation combat and otherwise. Being able to lock down opponents to win the fight much easier. Other methods through which to redirect the opponents and avoid the fight altogether. I'm not sure how you would stat that as some are direct interference of someone's mind while others focus on battlefield control.

Then we have the things such as Mending, Create or Destroy Water, Goodberry, Sending, Tongues, Commune, Scrying... just to name a few. I'll admit some of them can certainly be more skill-relevant. Others might fall under some form of natural resilience while different choices account for tactical preparation.

First reaction is sure, but second reaction is that I don't think you can define a metric that takes those abilities into account. Unless you want a 'has goodberry flag', which really isn't the point of key metrics.

If you can't make an easily definable metric for it then it probably shouldn't be listed as a key metric - not saying any of that isn't important, but key metrics must be definable and preferably objectively so.

[tKUUNK]

You have a good list already here. Random thoughts in no order:

It's funny, we can probably leave total HP off the list of Key Metrics. It's nice to have a few extra hit points, but the ability to end encounters quickly & avoid taking damage are usually far more important than total HP, in my 5e experience. In some earlier versions of D&D, when there was a wider gap between the high & low HP classes, I'd call it a Key Metric.

Movement, I'm assuming, also includes things like teleportation & flight, yeah? Not just plain old land speed.

I was going to suggest adding "Out of Combat Utility" but this is probably best left as a combination of easier to define things like Skills & Magic. To keep it objective, like you said.

Another way to add nuance to damage comparison is single-target vs AoE.

Combat avoidance could be its own category. But again, a lot is wrapped into that.


My conclusion:
Every time someone says, "Great build", your next thought should be "great for what?". Side note, I think this is one reason why everybody (all the cool people anyway) enjoys reading LudicSavant's character concepts. They're usually great builds for a specific purpose or role, and Ludic is always clear-sighted (and entertaining) on the character design concept. That's the side of the game that can be lost when we're all wrapped up in straight DPR / spike damage comparisons.

Thanks for calling this out.

[tKUUNK]

oh yeah... I forgot THE Key Metric in 5e: Action Economy!

[animorte]

I was going to suggest adding "Out of Combat Utility" but this is probably best left as a combination of easier to define things like Skills & Magic. To keep it objective, like you said.

This is essentially what I was trying to say, but it isn't so easily definable, thus not key metrics.

[Leon]

The Key Metric is play what you want and feel is best for you not what a bunch of people on the internet think is.

[PhantomSoul]

oh yeah... I forgot THE Key Metric in 5e: Action Economy!

DEFINITELY this! Uses for Bonus Actions and Reactions can be quantifiable at the very least.


This is essentially what I was trying to say, but it isn't so easily definable, thus not key metrics.

For out of combat utility, there are at least some things that could be quantified: cantrips, rituals and skill/tool proficiencies+expertise seem like options. Sure, cantrips aren't all equal (and not all utility), you might not have the right ritual, but they probably still give some starting point that's reasonable while being definable and "objective".

[strangebloke]

The problem really comes down to how much people value things like skills, movement ability, damage, initiative, and defense. You'll see people who hyper-fixate on one of these things, while other people see it as worthless. Everyone knows damage is worthwhile, at least. In 5e you do generally have to reduce the enemy hitpoints to zero, and more damage lets you do that faster, but things like defense are often contingent on a lot of other decisions - most of them made by the DM.

This is something I realized while watching critical role. Matt Mercer calls for skill checks for everything. It's from my perspective absurd. But someone who played in his campaign would logically think rogues and bards are a lot better than someone who plays in my campaign. Similarly, if your main experience with DND is stuff like Storm King's Thunder you probably think of paladins as hilariously broken and ridiculous, whereas someone who plays a more attrition/exploration-focused game would think of paladins as really strong but not insane. And the reverse applies for rangers.

Similarly, things like banning GWM/PAM/CBE/SS as feat options really changes the focus of basically all martial builds.

[Frogreaver]

The problem really comes down to how much people value things like skills, movement ability, damage, initiative, and defense. You'll see people who hyper-fixate on one of these things, while other people see it as worthless.

Everyone knows damage is worthwhile, at least. In 5e you do generally have to reduce the enemy hitpoints to zero, and more damage lets you do that faster, but things like defense are often contingent on a lot of other decisions - most of them made by the DM.

This is something I realized while watching critical role. Matt Mercer calls for skill checks for everything. It's from my perspective absurd. But someone who played in his campaign would logically think rogues and bards are a lot better than someone who plays in my campaign. Similarly, if your main experience with DND is stuff like Storm King's Thunder you probably think of paladins as hilariously broken and ridiculous, whereas someone who plays a more attrition/exploration-focused game would think of paladins as really strong but not insane. And the reverse applies for rangers.

Similarly, things like banning GWM/PAM/CBE/SS as feat options really changes the focus of basically all martial builds.

It really sounds like you are trying to subjectively compare character X to character Y and determine which is overall better. That's not really what key metrics are for. They are more like a list of separate metrics along which we can compare characters.

Character X does A damage, has B hp/recovery and C movement speed. Character X' does A' damage, has B' hp/recovery and C' movement speed. We can compare Character X with X' along those metrics without needing to subjectively claim either is better overall.


oh yeah... I forgot THE Key Metric in 5e: Action Economy!

TWF serves as a great counter example.


You have a good list already here. Random thoughts in no order:

It's funny, we can probably leave total HP off the list of Key Metrics. It's nice to have a few extra hit points, but the ability to end encounters quickly & avoid taking damage are usually far more important than total HP, in my 5e experience. In some earlier versions of D&D, when there was a wider gap between the high & low HP classes, I'd call it a Key Metric.

As an example, I think HP/Recovery metric is the closest single data point we have to explaining why monks in the early game just feel weak.


Movement, I'm assuming, also includes things like teleportation & flight, yeah? Not just plain old land speed.

Yes, I was envisioning things like misty step impacting those metrics. Alternate move types would have a few separate metrics associated.


I was going to suggest adding "Out of Combat Utility" but this is probably best left as a combination of easier to define things like Skills & Magic. To keep it objective, like you said.


Another way to add nuance to damage comparison is single-target vs AoE.

Agreed


Combat avoidance could be its own category. But again, a lot is wrapped into that.

I'm not sure that's definable, at least without getting down to individual spell selection, which isn't the kind of breakdown that makes a great metric. Though potentially there are a few specific spells that are just so strong they need called out individually (Wish perhaps).

[Hael]

One thing I've seen many recent threads for is comparing classes. Most are damage related but not all. Some have been about survivability. Etc. I don't think we can ever get all the class differences down to a single number.


Websites like dndcombat that assign ELO ratings to specific builds is probably more what we would ultimately need. They dont have full functionality yet, but that's really what you would want.

Non interacting white room numbers are useful, but how they play together is ultimately more important and unfortunately hard to quantify. For instance anecdotally one of the big problems for martials late game in my experience, isn't their damage. But rather the fact that they rarely get to do that damage. Most of the time they failed a wisdom/dex save and are patiently waiting for a dispel magic, or they are waiting for someone to cast fly or their magic sword just got poofed and they cant do damage anymore..

Thats why features like paladin aura is ultimately one of the best dpr abilities in the entire game as it actually enables them to perform round after round.

[Amechra]

I honestly don't see the point of doing this instead of, say, setting up a small number of "benchmark encounters" and seeing how well different characters do in them.

[PhantomSoul]

oh yeah... I forgot THE Key Metric in 5e: Action Economy!

TWF serves as a great counter example.



Isn't it the best example, not counter-example? From all discussion, the BA cost is (one of) the big problems people bring up, and it's such a big cost that pretty much the one context post-level-4 that people highlight it as reasonable is for Rogues, and only if they missed their Action Attack, since then it's an Attack Action in value at the cost of the BA (and even then, the BA Attack is still not treated as being a given!).

[Waazraath]

One thing I've seen many recent threads for is comparing classes. Most are damage related but not all. Some have been about survivability. Etc. I don't think we can ever get all the class differences down to a single number.

So I propose a new framework, 'Key Metrics' where we work together to define a set of robust and definable metrics that we can use to evaluate characters. These metrics should not be subjective in themselves. They should be meaningful to gameplay. They don't have to account for everything, but hopefully they cover alot of ground.

Potential Key Metrics
Damage (at will damage, Nova damage no prebuff, nova damage prebuff, daily damage with adventure day structured as short, normal, long).
HP/Recovery (potential per day, per encounter)
AC
Saves
Skills (breakdown by skill and show max available bonus per skill?)
Movement (move speed in a turn using all actions and bonus actions, move speed in a turn using just your bonus action, move speed in a turn using just your action, move speed in a turn while disengaging, irregular movement types?)
Incomparables (Rage damage resistance, stunning strike, etc)
Spells (Unsure how to create a metric for these? Possibly just max spell slot level and level of highest spell known?)

What other metrics do you think do a great job of encapsulating and differentiating character performance?

For my own builds I (usually) use the following, some of which are more quantitative and others more qualitative):

Action economy: uses for bonus action, reaction, availability of other options through action surge, spells (e.g. time stop), minions, etc.
Offence: to hit, normal damage, damage with resources, AoE (if available)
Defense: hp, AC, saves, resistances, immunities
Mobility: movement speed, modes of movement (flying, swimming, climbing, etc.), teleportation effects
Utility: out of combat stuff the character can do (mostly scouting and social pillar) through skills, spells, background, class features etc.

Optional: Buffing/debuffing/battlefield control - depends a bit on how much a character has on whether it is a category in itself, or part of offence.