Caster / Martial Disparity in Pathfinder

This is a helper post to the main analysis of C/MD, going through an analysis of the Pathfinder RPG (First edition, not second one). You could see this as a template for analyzing other gaming systems. I also analyze some of the common objections to the existence of C/MD.

Structurally, I will go through the seven challenge areas I outlined in a previous post, and show that casters substitute martials on them, but the reverse is not the case.

As usual, if you are confused about any terms, make sure to check out the glossary, and if you are looking for other posts, check out the blog map.

1. Social encounters.

When it comes to social encounters, the undisputed king of Pathfinder spells is, of course, Charm Person. It instantly makes a person friendly to you regardless of their original attitude; chances of it working are much higher than for a diplomacy skill check, which has to hit a pretty steep DC; and, at higher levels, it even lasts longer than the effect of a normal diplomacy check.

Does the Charm Person solve every social encounter? No. There will be situations where casting a spell is inadvisable or impossible: for example, if you encounter a group of people, then casting a spell on their leader may provoke a fight. In some situations simply being friendly will not be sufficient: you may have to use lies or intimidation.

But Charm Person solves a lot of encounters. If I were to put a number on it, I would say that a good 20% of social encounters in published adventures can be immediately solved by it, and up to 40% can be with some planning and tactical arrangement. This does not just count the outwardly social encounters: for example, Charm Person is perfect for interrogating a captive after a fight. And it is a single spell: a trivial investment of resources. For example, a 5th level sorcerer would know between 6 and 11 spells total, and a 10th level one between 15 and 25.

Of course, there are other spells which can directly boost your social skills, expanding the repertoire of the casters even more. Having spells does not preclude investment into skills, either: social abilities of a caster with Charm Person and 5 ranks in Diplomacy are going to be strictly superior to social abilities of a martial with 5 ranks in Diplomacy, equivalent stats, and any other combination of abilities.

2. Logistics & Survival.

Casters are the undisputed kings of logistics. Let’s start with travel.

If you want to travel to a different plane, you need Plane Shift, there is no real substitution. If you have to travel a large distance in a hurry, you go for Teleport: no ship or vehicle will get you to your destination in a comparable time. Medium distance? Wind Walk, Phantom Steed or Mount will be as fast or faster than a horse. Need to go over difficult terrain or chasms? Overland Flight or Dimension Door have you covered.

How about carrying items? Shrink Item, Floating Disk and Ant Haul are perfect for this; if that fails, simply buy a horse, they are cheap. Or cast Mount.

On top of this, there are magical items which extend your carrying capacity. Besides the omnipresent extradimensional storage, there are also muleback cords.

How about sustenance? Create Water solves the water problem, and Goodberry or Create Food and Water solve the problem of nutrition. The only reason you would hunt is if you lacked one of the many options to create food out of thin air.

When it comes to navigation, the spells do not offer too much support. Know Direction is essentially a perfect compass, Lay of the land will give you a trove of essential information, and at higher levels, Find the Path solves the problem entirely; but for most of the game, you will be relying on player logic or on the Survival skill.

Overall, it is indisputable that casters fully substitute martials when it comes to logistics.

3. Information warfare.

Information warfare forms a spectrum depending on how direct the confrontation is, from what I term “information combat” to “investigations”. The terms should be self-explanatory.

In information combat, martials are mostly relegated to the stealth and perception skills: to hide oneself and to find those who are hidden. On the caster end, there are more options. The classic example is, of course, Invisibility, hiding a caster much more thoroughly than anything that is possible by a martial. In fact, the entire school of illusion spells can be renamed into “information warfare”.

Casters also have more sensory options. Spells such as Detect Magic, True Seeing, Echolocation, Detect Secret Doors, Detect Undead and Thoughtsense offer a wide variety of new sensoria. The only way to guard against them is with more magic, with spells like Mind Blank, Nondetection, Undetectable Alignment and Magic Aura. This entire sphere of confrontation that is absolutely essential for combat at higher levels is completely locked off from martial involvement.

When it comes to investigations, martials once again largely can only rely on their skills - such as using Diplomacy to gather information, Linguistics to read texts and their various knowledge skills for figuring out specific information. In principle, we could also imagine more varied approaches - such as going to a library to research a topic - but those are highly dependent on the setting, as well as being equally available to casters.

Casters have a much broader toolset. Contact Outer Plane, Divination, Scrying and Discern Location provide a massive trove of information that cannot be accessed by other means. Locate Object or Comprehend Languages could be substituted, but in practice simplify problems by an order of magnitude or more.

In other words, casters can do everything that martials can do, but there are things casters can do which martials cannot do.

4. Security & security penetration

Security penetration is what happens when you are trying to break into a place. Security is when you prevent other people from doing that.

The two classic security penetration skills are, of course, Disable Device and Stealth - used for lockpicking and sneaking respectively. I have already mentioned how casters can substitute the stealth skill, so it does not bear mentioning. Substituting Disable Device can largely be done with the Knock spell, but actually, that is the least of your problems.

One of the key concepts in security is the so-called attack surface. That is the set of all points or approaches that could be used to compromise a system. For example, if we look at the problem of entering a house, then the attack surface would include the following things:

• Front door

  • Lockpicking

  • Breaking down

  • Lifting off its hinges

• Windows

  • Lifting the latches

  • Breaking down

• Walls

  • Breaking down, but it is difficult

And so on. The problem of security can then be seen as minimizing the attack surface (i.e. removing potential ways for an attacker to get in) and reinforcing it (i.e. making things you cannot remove, such as the front door, harder to breach). The corresponding problem of security penetration is to find novel attack surface points, or to find a way to breach known ones.

This puts the problem of C/MD in security in a different light. Martials, by their nature, can only access the “physical” parts of a system. Casters, on the other hand, can also access the magical ones: the set of exposed attack points is thus leagues larger.

You only have to breach the attack surface in a single place, and it doesn’t matter which one. A caster may not be able to pick the lock on the front door, but if they can teleport inside, turn the handle on the other side of the door with telekinesis, summon a helper to open the door from the inside, disintegrate the door or the lock, transfigure a part of the wall into a new entrance or charm the owner into opening the door for them, then they have still solved the problem.

Because of this radical expansion of the attack surface, casters are, fundamentally, operating on a completely different level from martials when it comes to security problems.

5. Puzzle solving.

Puzzle solving is an area that is generally devoid of useful spells. At best, you may be allowed to roll an intelligence check in exchange for hints from the GM. Overall, it is an area where casters and martials are fairly equal - anyone can think of a clever solution and suggest it.

But therein lies the trap. If casters and martials are equal in puzzle solving, then martials are substitutable by definition. Furthermore, given the same puzzle solving ability, a caster will always have more information available to them by virtue of their dominance in the information collection sphere. Puzzle solving cannot meaningfully affect our estimation of how substitutable martials are by casters.

6. Uniquities.

Uniquities are an almost exclusive domain of the casters. This happens due to a simple physical asymmetry which I already brought up when discussing security - casters can interact with the real world and the magical world, whereas martials can interact only with the physical world. As a result, they are simply locked out of at least 50% of all weird and unusual stuff that could be happening - and in practice, much more, because plot relevant events will almost universally be magical in nature.

This asymmetry has been discussed more in the main post. In general, if something weird has to be done, you need a caster.

7. Combat.

Finally, we come to combat. This is one of the most arguable frames of analysis when it comes to C/MD, but I think it still holds up.

Image 1: Disparity in combat options.

If we take an eagle’s eye view of DnD combat, its core problem can be seen as disabling the enemy before they can disable your team. One of the simplest ways of disabling someone is to reduce their HP to zero. Martials do this fairly well, especially at lower levels. Casters are, generally, worse at direct damage. So where does the idea that they can be substituted in combat come from?

There are several layers to this. First of all, analyzing combat as a simple competition of HP is very simplistic. If we use a more complicated analytical framework - such as the Forge of Combat - then we can see that there are three primary jobs a player can do in combat. FoC labels them as “Hammer”, “Anvil” and “Arm”. Hammers are responsible for disabling foes, anvils for controlling where the battle is happening and on whose terms, and arms smooth over the deficiencies of individuals with various buffs.

Martials are decent at being hammers, and quite bad at being anvils or arms. Casters, on the other hand, can fill all three roles quite well. This already makes one side of the problem clear: martials are simply lacking in ways to fill critical roles on the team.

You may ask me: how could casters fill in a martial role? It would seem difficult to compete with their physical damage, no? Well, not really. Remember, “caster” isn’t just a wizard. Classes like Paladins, Rangers or Bards have good physical damage out of the box, and can supplement it with spells. Alternatively, a Druid can wildshape for long periods of time, and can easily match any fighter for damage output. None of this requires feats of high optimization; but if we wanted to look into those, then casters are capable of damage output martials can only dream of.

Secondly, we once again run into the attack surface problem I mentioned before. Martials, by and large, can only disable enemies by reducing their health. Casters, on the other hand, have much broader options (see Image 2), and can change them in response to their opposition. Even if a caster specializes in one type of combat spells, the cost of supplementing their attack portfolio with 1-2 alternative options is very low, compared to an equivalent for a martial.

Overall, this analysis shows that casters are superior to martials when it comes to encountering the main challenge areas present in the game. Nonetheless, there are still some potential holes  in my argument. Let’s fill them up, by going over some common objections.

Objection 1: Skills

In the previous analysis I’ve shown that Pathfinder casters are superior to martials assuming that they have an equivalent access to skills. But why should we assume that? It could be the case that martials have more skills overall than casters. Sure, in any given niche, a caster could outshine a martial; but if a martial has more niches, they may still be irreplaceable, right?

I empathize with people who want this to be the case, but unfortunately, it is just not true. In fact, casters have the exact same access to skills as martials.

To analyze this, I use classes from the skills table on the PFSRD; this can be seen as the list of the most popular classes in the game, ones that everyone is familiar with. Alchemist is assumed to be a caster.

If we look at average skill statistics for these classes, the following picture emerges:

Image 2: Table of average skill statistics for the main caster and martial classes. “Class skills” refers to how many “class skills” that class has. “Trained skills” refers to how many skills out of that set cannot be used untrained. Skill ranks per level is the raw number of skill ranks that class gains every level, before taking INT bonus into account. Adjusted skill ranks adds 3 to that number for Alchemist, Magus, Witch and Wizard: INT is their casting ability modifier, and so they can be safely assumed to have a higher stat value than other classes. In effect, I assume 10 INT for everyone except classes that get a use from INT. Calculations can be found here.

From this, we can see that casters have more class skills, on average, than martials; and the number of raw skill points they get per level only differs by about 20%. Adjusted number of skill points is, essentially, indistinguishable. In fact, if we compare the total skill modifier (+3 per invested class skill plus total skill ranks) we can see that the pictures are essentially indistinguishable:

Image 3: Average skill rank equivalent vs level, for casters and martials. Skill rank equivalent is defined as the total sum of skill points earned up to that level, plus 3 per invested class skill (aka a skill with at least 1 skill point, which increases the skill modifier by 3 for class skills). This is a more accurate measure than skill points because there is no mechanical difference between 3 additional skill ranks or having a stacking +3 modifier. Calculations can be found here.

This average picture conceals a lot of variation. Skill rank equivalents at level 20 vary all the way from 67 (sorcerer) to 223 (rogue). But the situation does not improve for martials when we look at specifics. Here is a choice selection of casting and martial classes:

Image 4: Skill rank equivalent vs level, for a selection of classes. Most of the analyzed classes omitted to make the picture more clear. Mathematics done here.

Here we can see that Fighter has essentially the same number of skills as a Sorcerer, while Monk is competing with Druid and Rogue with Bard. Wizard, the most famous casting class, is actually above average in terms of access to skills. It is thus clear that casters can substitute martials in the skill area just fine on their own. In fact, if you replace most martials in your party with wizards, your party’s access to skills has probably increased.

The only outlier is Rogue. It is true that Rogues have a reputation for being a “skill monkey” - a class with great access to skills. But Rogues can be easily substituted by Bards: their greater number of class skills (whooping 28 to rogue’s 21) compensates quite well for a somewhat smaller number of skill ranks, and on top of that, they also have 6 levels of spellcasting. There are only two class skills that a rogue has and a Bard doesn’t: Disable Device and Swim, with the latter being laughably easy to substitute. DD, on the other hand, can be picked up as a class skill with a trait, or invested into regardless of class status.

Ultimately, we can see that skill access does not defeat the C/MD.

Objection 2: Items

Fine, so skills are a no go, but what about items? If martials lack out-of-combat options, could they not make up the gap by relying on mundane items - ropes, pinions, crowbars, etc - or magic items - wands, scrolls, and so on?

The fundamental problem with this reasoning is that casters can buy items too. Beyond the first level, you can almost always assume that different characters will have the same wealth, because parties will split loot equitably. This results in an equal purchasing ability, and so equal access to items. Casters can also craft magic items, extending their access further.

Logically, the only situation where items might help them reduce the gap in capability would be if the martial abilities had better synergies with material wealth than those of casters. But in fact, it is the reverse. Martials (Rogue excluded) do not get Use Magic Device (skill used to utilize wands or scrolls from different classes) as a class skill; this means that they are strictly worse than casters at using all consumable magical items. On the other hand, when it comes to mundane items, the problems where they are helpful and where a caster cannot use them in the same manner simply aren’t very common. When they do occur, martials can often be substituted by summons, shapeshifting, or more indirect approaches.

Ultimately, items can, at best, reduce the relative size of the capability gap, but not eliminate it.

Objection 3: One Trick Ponies and the Schrödinger's Wizard

Often, once it is established that casters can substitute martials in a variety of situations, discussion moves to a new claim. Sure, maybe there are substituting spells in any given situation, but this only matters if the caster has anticipated the problem they are facing. Perhaps in practice a caster can only guess, and not with high accuracy. As a result, most of the time, C/MD would not come up.

This claim is referred to as the “Schrödinger's Wizard”. Like the popular conception of a well-known quantum physics thought experiment, a theoretical wizard always has the perfect spell available out of all the potential options.

This objection does not hold up to scrutiny. Many of the spells with the highest impact on C/MD can address many different challenges, depending on what comes up. For example, Overland Flight doesn’t just allow you to cross chasms: it also protects against a broad variety of ground threats, avoids difficult terrain, increases your speed, completely substitutes the climb and (usually) the swim skills and allows you to engage aerial threats in melee. Summon Monster doesn’t just give you access to a disposable source of damage, but also all the varied abilities and sensory modes of those monsters, and everything that implies. This means that a caster with a logical spell selection is going to be operating at high efficiency even when they didn’t predict the exact nature of any particular challenge.

On top of that, if the caster can predict what they will face, then they can simply prepare normally. This can easily happen in a variety of situations: there is little need to prepare social spells if the party expects to go through a lot of fighting tomorrow. If I were to guess, I would say that a good 20% of the time the party can predict what they will be doing tomorrow with a high accuracy, and another 40% of the time they can make reasonably precise guesses. If the party has access to divination spells, these percentages increase significantly.

Finally, casters do not just cast from their spell slots. They also have access to scrolls and wands: items that can effectively extend their casting ability beyond the normal limits. Such consumable items are a perfect tool for the spells that are used very infrequently. Keeping a scroll of Comprehend Languages in your backpack will guarantee that you will not be surprised by a chance encounter with sentient creatures.

The “one trick pony” claim is similar: that C/MD is only present for highly specialized casters that are lacking in other areas. This is likewise not the case: most of the spells require no particular effort or specialization to become effective, and do not prohibit the usage of other spells. All you have to do is pick them from the spell list, and they will perform their function admirably. In fact, this is often so easy, that a player can build an incredibly versatile and powerful caster inadvertently, by simply going down the list of spells available to their class and picking ones that make them go “Ooh, shiny”.

Objection 4: Leveling Misery

Another common mistake is to look at C/MD as a leveling artifact. The logic goes that yes, at high levels casters are superior; but at early levels they suck. Their mastery at higher levels is then simply a “reward” for how awful their experience was early on.

First of all, this argument is based on a factual falsehood. C/MD is present from the first level onward. It is initially less prevalent, and it is true that the problem increases significantly with levels; but it is always there.

But even if we grant the premise - that early on casters struggle to survive - the argument will still not hold water. The point of DnD is to maximize the utility of all players: therefore, everyone should be, at minimum, not miserable throughout the whole game. Making a player go through months of pure misery just so that later on they may feel cool isn’t game design: it’s just an attempt to justify hazing.

Objection 5: Attrition gauntlets

Final objection I will mention has to do with attrition. Casters, generally, have to use spell slots to do anything; martials, on the other hand, don’t seem to expend any kind of resource. Therefore martials could, theoretically, go on “forever”; casters, on the other hand, have a strict limit to their combat potential.

In practice, this isn’t a great objection. Martials do have a limited resource: health. When their hit points (plus healing) run out, they can’t go on any more than casters without spells. On the other hand, casters have options that do not rely on spell slots (e.g. Planar Binding or Wildshape). Scrolls extend the attrition limit even further. Even starting from the first level, casters have various cantrips which can be used indefinitely.

But even if we assume, for the sake of the argument, that the factual statement (that martials are better at handling attrition) is true, the argument still falls on its face. In modern DnD, most adventures do not assume you are doing the sort of multi-session-long dungeon delving that was emblematic of earlier editions; you are often free to decide on your own approaches to problems, and can pace yourself appropriately. Ability to resist extreme attrition is thus largely unnecessary, and what matters more is how effective you can be in your first 4-5 encounters, which is a caster specialty.

In fact, given how many encounters you need to properly exhaust a caster beyond the lowest levels, trying to drain them might be directly detrimental to your game. Not only will this lock you out of shorter adventuring stories, but also make your game extremely combat-heavy: something that your players may or may not enjoy.

To sum up: yes, in Pathfinder, C/MD is definitely present. We can confidently say that the Non-combat C/MD statement is true, and I lean in the direction of saying that Hard C/MD is true.

If you are looking for other posts on my blog, check out this list of all other posts, and if you enjoy what I write, you can subscribe to receive updates by email:

Additional reading

Dispelling the Myths: Caster-Martial Disparity