Proposal: how to compare the utility of two classes

[Shining Wrath]

Yes, this is partially in response to the Druid versus Martial Classes threads, but I think it has more general utility.

First, I think that the smart 5e player optimizes party, not character in isolation.

What I would suggest, then, is that you imagine the following:

Two parties, Party A and Party B, hate each other with an eternal and destructive passion for Reasons. If A & B ever encounter each other, they attack immediately without anyone considering the possibility of parley or retreat.

As it happens, A and B have exactly the same composition: same classes, same archetypes, same levels, same spell selections, same gear.

To compare two classes, subtract one person from A and B, and substitute a character of the classes under comparison, one into A, and one into B. Then construct scenarios which test the aspect of the classes you wish to compare.

By the scenarios you construct your readers shall know whether or not you intend to conduct a fair trial. For example, if comparing the Fighter and the Druid, placing the parties in a forest with each knowing the other is there favors the Druid; teleporting the parties with no warning into a 20' by 10' room with a 6' ceiling so everyone has to fight toe-to-toe favors the fighter (especially if you rule that the Druid can't wildshape into anything large without being forced to his / her knees).

So, to compare a Druid and a Martial class, I propose the following:

The other 3 members of the party are a Wizard (Diviner), Rogue (Assassin), and Cleric (War). The former fourth was a Warlock.

Everyone is level 20; everyone has +1 weapons and armor as appropriate. Spellcasters have scrolls. No one has any magic item specially designed to shut down a Druid or a Barbarian / Fighter / Monk / Paladin / Ranger.

Scenario the first: A & B meet each other in a town square. Neither knew the other was in town, because each Diviner cancelled out the other. The crowd parts, and the parties face each other at a distance of about 30'. No surprise (or, rather, equal surprise so that both A & B spend 6 seconds gaping at each other). The parties have been looking for information and doing some shopping all day long, and it's been at least 10 hours since they left their inns this morning; thus, any buffs of duration less than 12 hours have likely expired unless reapplied at lunch. No one has any weapons readied, it's not the sort of town where you stroll about with sword drawn.

Scenario the second: A & B meet each other in a dungeon. Again, they didn't know the other was in this dungeon. A opens a door and comes in behind B, who are examining some glyphs on a wall; the room is about 50' by 50' by 10'. At level 20, it may be assumed that someone in Party B is watching behind them at all times because if they were fools they'd be dead before level 20. The parties all have the usual "In a dungeon" buffs applied.

Scenario the third: A & B are searching a desert for the fabled Temple Of Phat Loot (a minor deity once worshiped by the Bling tribe). They may not meet each other at all; the real contest is finding the temple. The Temple is known to be hidden from above by divine-level illusions; that is, it can only be spotted by someone on foot (or flying very, very low). Did I mention the sandworms that are attracted to the vibration of footsteps?

[GiantOctopodes]

Except, this is actually completely worthless. D&D is not (thankfully) Battle Arena, Fantasy Edition. The Utility of a character, and how much they add to an overall party, has almost nothing to do with "if in combat with a party composed the same except for character X, how would they perform? Can't retreat or do anything that involves thinking, either, this is just beating each other over the head until something caves". It is much *more* about "how often can character / class X pull their weight?" or "How often is character X outperformed by character Y fulfilling an ancillary role, while character X is performing their primary role?" or "How often would a party with character X be better off if it was character Y instead?" or "how much does character X add to the total capabilities and effectiveness of the party, vs if it was character Y?" These are the questions I recommend asking when comparing the utility of two classes.

Now I know, those are incredibly subjective, campaign dependent, and reliant on the existing mechanics of the party. After all, ultimately the DM helps create the story, and a good and wise DM creates it around the capabilities of his players (which leads to some interesting mechanics where specializing outside of the abilities of the rest of the party may actually make the game harder for them vs if no one was good at something, but that's a different topic), and makes sure everyone has a chance to grab the spotlight now and again. That does not mean that I think there is not an answer for those questions.

I think it can be agreed that any evaluation should begin on the basis of the three pillars. Rather than evaluate based on three different combat scenarios (effectively relegating grades to 33% of the total experience), let's incorporate exploration and social interaction into these things. If we evaluate total party impact, it is really hard to know what they bring to a party without knowing the party composition. However, in the same way that Peyton Manning was in the MVP conversation the year that he didn't play based on how poorly his team did without him, or how the impact of air resistance is most easily observed by measuring comparable performance in a vacuum, we can likely measure the impact of a party member by how well a party might fare without them.

In that regard, I think it best to break abilities down in the following way:
Unique: Only possessed by that class
Semi-Unique: Only possessed by (a number of) classes, one of which must be present to obtain it
Universal: Something able to be gathered by anyone, with the right selections on background / feats.

If we look at it that way, and analyze the usefulness of the abilities that are unique or semi-unique, we may come to a pecking order of classes. I suspect, done that way, we will find that it is much less about druids vs martial classes, and more about druids vs clerics, paladins, or other casters, and identifying which brings the most to a party dynamic (shocking reveal, it's the druid), and from there we can probably create an "A team" of classes which fill all desired gaps, finding "suitable" then "non-optimal" replacements for each of them. I suspect that the ideal 4 person party involves a Druid, Rogue, Bard, and some 4th class, but that's not based on any proper analytical research, but rather my own prejudices and at a glance analysis of them (wherein Bards + magic rock the social aspect, Druids + Rogues just absolutely wreck the exploration side of it, and Rogue + tank + buffing + environmental control + summoning wraps up your needs in combat quite nicely, with the 4th person merely allowing someone to not have to fill multiple roles). If / when I have time, I'll try to start a proper analysis, so that others can hopefully add in, but for now I just wanted to say: I admire your goal, but I think you're asking the wrong questions.

[Giant2005]

It still doesn't work - there isn't any possible means of devising a standardized way of comparing classes due to each class's strengths being focused into different tasks.
For example, if the two parties only consisted of two people then the classes with higher single target damage or mitigation would benefit more than those that specialize in aoe effects. If the parties both consisted of 12 people, then the opposite would be true.
I have no idea who a 4 person party benefits the most without crunching the numbers but it is guaranteed that a 4 person party doesn't showcase the strengths of all of the different classes equally.

[Eslin]

Agreed with the above two posters. Honestly I think the best way to do it would be to have several scenarios that might come up and see what two different classes would do about them. Take a party of your choice (say, barbarian, bard and warlock) and add different fourth characters to compare them. Add a fighter to the party, see how much he contributes, then add a druid instead and see if she does better or worse.

Examples: we need to get to the elven lands very quickly and convince the queen to evacuate the coasts because the kraken will attack them soon, party has to journey to the bottom of the sea to retrieve a magic trident, party is attacked by 5 ogres and an oni, party needs to kill a spirit naga and go to the plane of air and convince a djinni to wish the spirit naga to die permanently, party has to find a way to kill a host of orcs 100 strong lead by a clever warchief before they destroy a village two days away.

[MaxWilson]

The substitution methodology is a good approach, but evaluating the substitutions is intractable--there are too many variables to benchmark. Consider how different the "druid vs. fighter" melee comparisons would be if it were 4 druids vs. 4 fighters: onion druid HP would become a non-factor. And that's only one variable (party size), completely neglecting all the other variables you allude to: task domain, party composition, party level (could be heterogenous or homogenous), magic item level, alert status/buffs, environment, engagement range if a combat scenario, and probably some more that I'm overlooking.

And even if you could evaluate two classes along every one of these variables, how would you weight the results to come up with which class is "better"?

I admire your approach to the problem, but I think the problem is both intractable and unimportant. The only concern when comparing classes is whether a class dominates or is dominated by other classes to the point of destroying the enjoyment of players at a table who don't take the dominant class--and that concern is only a concern for system designers (including homebrewers) and individual DMs. 5E is well-enough balanced that it's pretty much a non-issue in actual play. Class comparisons is a social (Internet-forum-focused) activity, not a role-playing activity.

[silveralen]

I suspect that the ideal 4 person party involves a Druid, Rogue, Bard, and some 4th class.

Honestly? it's probably just two bardsand two combat based characters. A bard can bring any utility that a druid or rogue can, especially with two to double up on expertise and cherry pick all the best utility spells. Toss in one class to focus on AoE and one class to focus on single targets. Two bards can handle any real out of combat issue between the two of them, especially since the other two classes will bring something to the party (as AoE is practically guaranteed to be some variation of caster), and in combat you have two specialists plus the bards to fill in. The bard list is also naturally supportive, with solid healing and buffs/debuffs. That party can solve pretty much any problem with ease.

[Eslin]

The substitution methodology is a good approach, but evaluating the substitutions is intractable--there are too many variables to benchmark. Consider how different the "druid vs. fighter" melee comparisons would be if it were 4 druids vs. 4 fighters: onion druid HP would become a non-factor. And that's only one variable (party size), completely neglecting all the other variables you allude to: task domain, party composition, party level (could be heterogenous or homogenous), magic item level, alert status/buffs, environment, engagement range if a combat scenario, and probably some more that I'm overlooking.

And even if you could evaluate two classes along every one of these variables, how would you weight the results to come up with which class is "better"?

I admire your approach to the problem, but I think the problem is both intractable and unimportant. The only concern when comparing classes is whether a class dominates or is dominated by other classes to the point of destroying the enjoyment of players at a table who don't take the dominant class--and that concern is only a concern for system designers (including homebrewers) and individual DMs. 5E is well-enough balanced that it's pretty much a non-issue in actual play. Class comparisons is a social (Internet-forum-focused) activity, not a role-playing activity.

Well, I can answer that - sometimes the balance is so skewed that it's obvious who wins, in 3/5 of those scenarios a caster is flat out more useful than a martial without question and in the other 2/5 I'm pretty sure I know what the question's answer is. I still think my tests are a decent way of doing it, it just seems time consuming.

[MaxWilson]

Well, I can answer that - sometimes the balance is so skewed that it's obvious who wins, in 3/5 of those scenarios a caster is flat out more useful than a martial without question and in the other 2/5 I'm pretty sure I know what the question's answer is. I still think my tests are a decent way of doing it, it just seems time consuming.

More power to you then. I hope you enjoy yourself. (No irony intended, I really do.)

[Eslin]

More power to you then. I hope you enjoy yourself. (No irony intended, I really do.)
I do, but I'm looking for ways to let martial players contribute more. I've ported ToB across, but they're still far less useful when combat's over - the party would be fine with all casters and unable to participate in the plot with all martials and they know it, I want the non casters to feel like they can contribute too beyond the occasional skill check.

[Rack]

It's almost impossible to judge the utility of skills. Take open lock. If the DM wants you to have an item you'll have it, if not then no open lock check will be enough to get your hands on it. So it goes for investigation, detect traps. Heck even combat skills are irrelevant in most games.

Utility comes mostly down to what feels good. You can't put any objective standard on that.

[Z3ro]

the party would be fine with all casters and unable to participate in the plot with all martials

This sounds like a table issue and not an edition issue. The parties I've played with that were just martials had no trouble moving plots forward.

[silveralen]

I do, but I'm looking for ways to let martial players contribute more. I've ported ToB across, but they're still far less useful when combat's over - the party would be fine with all casters and unable to participate in the plot with all martials and they know it, I want the non casters to feel like they can contribute too beyond the occasional skill check.

The occasional skill check? That's problem number one. What sort of challenges are you throwing at your group, and what sort of skills do your martial characters have?

ToB accomplishes nothing for 5e, the only power gaps it fixed in 3rd were ones this edition fixed from the start (full attack issues and martials without buffs compared to casters stacking buffs). So it is unsurprising it didn't alleviate any problems. Unless they had problems contributing in combat, which is so alien to my 5e experience I'd have to ask if you tailored the encounters to do so.

[Yorrin]

This sounds like a table issue and not an edition issue. The parties I've played with that were just martials had no trouble moving plots forward.

+1

The party I'm DMing right now has very little magical access and yet has done fine with the plot.

[Shining Wrath]

I am slightly bemused that there are people in this thread saying you can't really compare classes when I've seen them in other threads contributing to conversations comparing classes.

As for Eslin's problem with Martials contributing, there's a spell to do nearly everything, but not every caster has that spell prepared. Look at the skills non-casters tend to have, create situations where those are useful. Fly solves a lot of problems - unless there's a high wind. Or large fliers who dine on other fliers but are less likely to molest a climber. And so on....

[MaxWilson]

I am slightly bemused that there are people in this thread saying you can't really compare classes when I've seen them in other threads contributing to conversations comparing classes.

Since I've contributed to some of those threads, I feel free to address this point.

I've participated in the "Druid 20 vs. Fighter 20" thread because I think that scenario is constrained enough that you can analyze it. That doesn't really say anything about whether druids are "better" than fighters or not, as a class. I prefer 5E fighters frankly, although I'd like it if someone else took druid so he could cast Free Movement on me before I head into the Web spell to bash some heads.

That doesn't make fighters "better" than druids either, BTW. They're just different.

[Shining Wrath]

Since I've contributed to some of those threads, I feel free to address this point.

I've participated in the "Druid 20 vs. Fighter 20" thread because I think that scenario is constrained enough that you can analyze it. That doesn't really say anything about whether druids are "better" than fighters or not, as a class. I prefer 5E fighters frankly, although I'd like it if someone else took druid so he could cast Free Movement on me before I head into the Web spell to bash some heads.

That doesn't make fighters "better" than druids either, BTW. They're just different.

Yeah, I'm only slightly bemused, not completely aghast :smallsmile:

I'm trying to provide a framework, and notice I tried to constrain the scenarios to comparing one aspect of a build, not everything.

It's fair to ask which classes can solve which sorts of problems; no one class solves all problems.

[MaxWilson]

Yeah, I'm only slightly bemused, not completely aghast :smallsmile:

I'm trying to provide a framework, and notice I tried to constrain the scenarios to comparing one aspect of a build, not everything.

It's fair to ask which classes can solve which sorts of problems; no one class solves all problems.

Again, I like your general approach. You could say that my objection is not to substitution or even benchmarking; I just don't think class utility is reducible to a single scalar summary statistic.

YMMV.

[GiantOctopodes]

Ok, I've got a smidge of time, so let's break down some of the pillars, and see what bricks make them up.

Combat:
- Damage. Straight, raw, damage.
- Damage mitigation. Whether through improved AC, illusions of yourself, being out of range of your foe or whatever, you want to take less than full damage.
- Advantage and Disadvantage. Mechanically one of the most powerful and common RAW things to do.
- Crowd Control
- Summoning
- Environmental / Terrain control
- Buffing
- Debuffing
- Healing. Note that just HP damage is effectively, from another way of looking at it, damage mitigation. However, the ability to remove poisons, curses, and other long term debilitating effects is a different thing entirely.
- Mobility (not just for damage mitigation, but for catching fleeing foes, escaping if need be, and also to deal with exotic movement types, such as flying, swimming, burrowing, or extradimensional travelers)

Feel free to throw more at me, I'm just whipping up ideas at the moment.

Exploration:
- Perception. The ability to see what's going on.
- Travel. The ability to get from point A to point B, even if there is a cliff, river, or whatever else in the way.
- Resource Management. The ability to keep you and yours healthy, fed, and alive while exploring.
- Tracking. Some people don't want to be found, and sometimes you need to find them.
- Stealth. The ability to evade detection / tracking yourself.
- Traps. The ability to bypass or disarm them, or to create them.
- Environmental adaptation. The ability to survive storms, blizzards, sandstorms, whirlpools, huge falls, or whatever else your DM might throw your way.
- Opening Paths. Whether by picking locks, breaking down doors, or walking through walls, the ability to get where others don't want you to be.
- Storage. Finding a Dragon's Horde is less impressive when you can't lug it out of there.
- Knowledge. It's one thing to see something. It's another thing entirely to understand it.

Same applies here, of course.

Social interaction:
- Communication. Gotta talk to people, and not everyone speaks your language.
- Insight. By this I mean intelligence gathering, whether it's their mood, their thoughts, their alignment, their level, or whatever else.
- Deception. Sometimes you don't want to tell the truth.
- Coercion. The ability to make others do what you want.
- Leadership. The ability to bolster allies or lead others willingly into adversity.
- Emotional Control. To ease a tense room or to make it tense, as needed.
- Performance. The ability to keep others interested, and some might say, distracted.
- Appearance. Appearance can be everything, whether to set the right impression or the wrong one

As before, ideas are welcome.

I'm basically trying to get a reference for myself of the broad category of abilities or tasks players would be seeking out, so that I can begin to, as described, break classes down into their component parts and see where they can uniquely contribute. The better the idea I have as to what parties need, the more accurate the results will be, so contribution is definitely welcome.

[Vogonjeltz]

Scenario the first: A & B meet each other in a town square. Neither knew the other was in town, because each Diviner cancelled out the other. The crowd parts, and the parties face each other at a distance of about 30'. No surprise (or, rather, equal surprise so that both A & B spend 6 seconds gaping at each other). The parties have been looking for information and doing some shopping all day long, and it's been at least 10 hours since they left their inns this morning; thus, any buffs of duration less than 12 hours have likely expired unless reapplied at lunch. No one has any weapons readied, it's not the sort of town where you stroll about with sword drawn.

Scenario the second: A & B meet each other in a dungeon. Again, they didn't know the other was in this dungeon. A opens a door and comes in behind B, who are examining some glyphs on a wall; the room is about 50' by 50' by 10'. At level 20, it may be assumed that someone in Party B is watching behind them at all times because if they were fools they'd be dead before level 20. The parties all have the usual "In a dungeon" buffs applied.

Scenario the third: A & B are searching a desert for the fabled Temple Of Phat Loot (a minor deity once worshiped by the Bling tribe). They may not meet each other at all; the real contest is finding the temple. The Temple is known to be hidden from above by divine-level illusions; that is, it can only be spotted by someone on foot (or flying very, very low). Did I mention the sandworms that are attracted to the vibration of footsteps?

I mean, one of the questions we might ask is: What problems does each option present for the opposing team?

For example, none of those classes (wizard, rogue, cleric) have strength saving throw proficiency, the rogue doesn't have wisdom save proficiency, and neither mage nor cleric has dexterity save proficiency. So the Battlemaster Fighter can pretty much apply most any maneuver to them at will. DC 19 saves against which those classes are likely to have +0.

The Monk is a serious threat to all three for a similar reason (No Con saving throws means potential instant death).

I'm thinking at this point that to continue is silly because pvp is not intended and there are far too many combinations and variables involved (what builds for each? etc...). Changing one detail, or one action by either side, is too likely to upend who wins to make a reasonable guess. This also doesn't speak to their utility at all, just their combat options.

[archaeo]

It's fair to ask which classes can solve which sorts of problems; no one class solves all problems.

It sometimes feels like people ignore the degree to which the table should be in complete control of the kinds of problems they face. If I'm DMing for a group of players who want to play an all-martial party, I'm not going to throw up my hands and declare that they're incapable of meeting the challenges of my plot; I'm going to try and tell a story where their characters are the heroes (or villains, as the case may be), designing challenges and plot points that can be overcome with clever play.

Obviously, this becomes more difficult in a "mixed" party, but I don't see it onerous to design campaigns with the goal of providing time in the spotlight for each player. The DM is making up the whole shebang. If martial characters are unable to keep up with casters in solving problems and achieving goals, I would argue that that's far more the fault of the DM than the system.

But, a proviso: a lot of this misconception just comes from the presentation of the system itself, which spends a ton of page space on individual spells, all of which bounce around in a clever player's head. Other methods of problem solving tend to be suggested, with a robust skill check framework that the PHB certainly expects DMs to make liberal use of, but they aren't a 1/3 of the book's page count.