Using data to understand how best to defend against adversaries.
In diving into character creation in the Cosmere RPG, several questions came to mind about which skills and attributes are worth investing in for various aspects of gameplay. Obviously increasing ranks in things like light weaponry will make your light weapon attacks better or building on thievery will make you a better pickpocket.
But what about for defending against enemy attacks? Which defences are they targeting? And as I dug into this, I realised that resisting or evading conditions sometimes required a skill check, so which of these are most important for a survivalist? I also wondered which conditions I can expect to come up against the most.
Because of all these questions knocking around in my mind, I decided to run a bit of a data analysis on the adversaries found in the Stormlight World Guide. I’ve pieced this research into a load of findings in the article below.
Overview
- Physical defense tend to be needed more than cognitive or spiritual.
- Agility, athletics and discipline tend to be the most highly contested skills.
- Prone, restrained and afflicted are the most commonly caused afflictions by adversaries.
- Adversaries tend to have stronger physical defenses than cognitive or spiritual defenses.
Which defenses are adversaries targeting?

In reading the rules, obviously physical defense is often used against physical attacks and other physical effects. It’s also clear that skill tests made for skills that fall under the cognitive umbrella are made against a creature’s cognitive defense and the same for spiritual and physical skill tests.
However, NPCs rarely make these kinds of skill tests. These are almost always initiated by the players. So the question I had was how important are cognitive and spiritual defenses compared to physical defenses. Can we expect NPCs to cause attacks and effects that might require beating our other defense types?
To work this out, I went through all the adversaries listed in the Stormlight World Guide and identified every creature that makes attacks or other effects that might cause a skill check against physical, cognitive or spiritual defenses and added them up. This is to get a sense of how much these types of defenses might be called into action. Here’s what I found:
| Defense Needed | Tier 1 | Tier 2 | Tier 3 | Tier 4 | Total |
|---|---|---|---|---|---|
| Physical | 10 (100%) | 21 (100%) | 10 (100%) | 5 (100%) | 55 (100%) |
| Cognitive | 0 (0%) | 3 (14%) | 1 (10%) | 1 (20%) | 5 (9%) |
| Spiritual | 3 (30%) | 5 (24%) | 2 (20%) | 3 (60%) | 13 (24%) |
As expected, physical defenses are targeted much more often (by 100% of enemies). Cognitive and spiritual defenses are both targeted from time to time, but this is the case with much fewer enemies. It’s worth noting though, that spiritual defenses are invoked by more enemy types than cognitive defenses making this a potentially more useful defense to invest in.
As mentioned, outside of combat, NPCs rarely invoke tests against PC defenses. For example, if an NPC attempts to deceive a PC, they don’t tend to make a deception skill test. Normally the PC will ask if the character is lying and roll something like an insight test against the NPC’s spiritual defense to determine their truthfulness.
Examples are also given for less prescribed skill tests that might be made against a cognitive or spiritual defense. The book uses the example of a disarming attack that others could apply at the GM’s discretion. Realistically though, these are likely to be rare.
Based on this data, it’s clear that physical defense remains the priority for a character, with spiritual defense and then cognitive defense being the order of priority. This may change though as new adversaries are published (perhaps with the Mistborn setting).
Which skills are most commonly contested?

In this analysis, I also noticed that many skill tests from NPCs beyond standard attacks, didn’t just test against a creature’s defenses; often they were contested skill checks against a particular skill.
This led to my next question. Which skills are most commonly contested against by NPCs? Again, by trawling through the adversaries in the Stormlight World Guide, I’ve identified how often each skill is contested against by adversaries:
| Skills | Tier 1 | Tier 2 | Tier 3 | Tier 4 | Total |
|---|---|---|---|---|---|
| Agility | 3 (30%) | 6 (29%) | 5 (50%) | 3 (60%) | 17 (31%) |
| Athletics | 4 (40%) | 8 (38%) | 4 (40%) | 2 (40%) | 18 (33%) |
| Heavy Weaponry | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) |
| Light Weaponry | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) |
| Stealth | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) |
| Thievery | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) |
| Crafting | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) |
| Deduction | 0 (0%) | 1 (5%) | 0 (0%) | 0 (0%) | 1 (2%) |
| Discipline | 5 (50%) | 5 (24%) | 4 (40%) | 0 (0%) | 14 (25%) |
| Intimidation | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) |
| Lore | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) |
| Medicine | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) |
| Deception | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) |
| Insight | 0 (0%) | 1 (5%) | 0 (0%) | 0 (0%) | 1 (2%) |
| Leadership | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) |
| Perception | 1 (10%) | 0 (0%) | 0 (0%) | 0 (0%) | 1 (2%) |
| Persuasion | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) |
| Survival | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) |
As you can see, there are only 3 skills that are contested against by more than 1 adversary type with their abilities. They are agility, athletics and discipline.
These feel like the skills I’d have expected to be contested the most in combat. They seem to focus on how quickly you can avoid an effect, your ability to resist an effect and how much you can hold your nerve against a foe. It’s also worth noting that sometimes there’s a choice between contesting with agility or athletics, meaning investing in just one of these will be sufficient for prevailing in many of these physical skill tests.
These 3 skill tests range in prevalence between about a quarter and a third of adversaries. All the other skills aren’t really worth considering from a defensive point of view (though they obviously possess other benefits).
How many adversaries cause conditions?

While certain conditions can be caused by basic actions like the grapple action (which causes the restrained condition), most can only be caused by special effects. In the table below, I’ve indicated which conditions are caused by effects from which adversaries (focusing only on what’s in their stat block rather than whether they have the capability to grapple for example, as all creatures can do this).
I wanted to focus on what NPCs can do to PCs, so there are a few instances where conditions are positive and adversaries can cause those conditions to end. This is what I’ve calculated for some positive conditions, though this is rare.
| Skills | Tier 1 | Tier 2 | Tier 3 | Tier 4 | Total |
|---|---|---|---|---|---|
| Afflicted | 1 (10%) | 4 (19%) | 4 (40%) | 0 (0%) | 9 (16%) |
| Determined | 0 (0%) | 1 (5%) | 0 (0%) | 0 (0%) | 1 (2%) |
| Empowered | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) |
| Enhanced | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) |
| Exhausted | 1 (10%) | 0 (0%) | 0 (0%) | 1 (20%) | 2 (4%) |
| Focused | 0 (0%) | 2 (10%) | 0 (0%) | 0 (0%) | 2 (4%) |
| Immobilized | 1 (10%) | 0 (0%) | 2 (20%) | 1 (20%) | 4 (7%) |
| Prone | 6 (60%) | 7 (33%) | 4 (40%) | 1 (20%) | 18 (33%) |
| Restrained | 2 (20%) | 6 (29%) | 4 (40%) | 2 (40%) | 14 (25%) |
| Slowed | 0 (0%) | 2 (10%) | 1 (10%) | 1 (20%) | 4 (7%) |
| Surprised | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) |
| Unconscious | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) |
What’s immediately obvious is some conditions are caused more often than others. Prone and restrained are the most common, followed by afflicted. It’s worth noting that sometimes it’s the same creature that can cause multiple different conditions, so while it may look like a high proportion of tier 4 adversaries cause conditions, most of this comes from 1 adversary (the servant of Yelig-Nar).
As a general rule, avoiding being made prone or restrained is going to be helped by a high agility or athletics rank, which we know now are highly contested skills (probably because of the prevalence of prone and restrained conditions in adversary stat blocks).
Which defenses are adversaries worst at?

So what if we then want to be proactive rather than reactive? Is there a type of defense that adversaries tend to be worst at?
To work this out, I took an average of each defense type for all adversaries to see which one is best to target if that’s an option (speaking very roughly, and understanding there will definitely be individual differences here).
| Defense Needed | Tier 1 | Tier 2 | Tier 3 | Tier 4 | Overall |
|---|---|---|---|---|---|
| Physical | 13.6 | 16 | 18 | 18.2 | 15.7 |
| Cognitive | 12.7 | 14.9 | 17.3 | 18.2 | 14.9 |
| Spiritual | 13.2 | 15.8 | 16.9 | 17.4 | 15.2 |
From this data, we can see that physical defenses tend to be strongest among adversaries. This means that if you’re equally strong in the relevant skills, but have a choice between targeting a physical defense or a different defense, you’re more likely to be successful with a non-physical test. Of course, many defense uses are pigeon-holed so you aren’t going to be able to make a choice, but sometimes you have a choice where you can make an attack or effect that could target one defense or another and this is when this can become relevant.
Interestingly, at tiers 1-2, cognitive defenses tend to be weaker than spiritual defenses. Conversely, at tiers 3-4, spiritual defenses tend to be weaker than cognitive. And of course, it’s worth factoring in your individual enemies. Dustbringers and skybreakers have superior cognitive defenses to their physical ones, so targeting their physical defenses will have more success for example.
Limitations
It’s worth being aware that while this data can have it’s uses, it does have limitations too including:
- Some adversaries may occur more often in one campaign than another.
- Some GMs may use different abilities or actions more than others.
- It doesn’t account for how often or how good certain adversary abilities are that might invoke conditions or contested skill tests.
- It doesn’t account for how good a PC is at certain skills. It’s no good targeting cognitive defenses if you have poor intellect and willpower for example.
- There will likely be differences when you account for unique enemies found in a campaign book, homebrew enemies and enemies from future settings and books.
Conclusion
If you want to think defensively in your character build, then there are some learnings you can take:
- Physical defense will usually be your most important defense type as enemies just don’t invoke other defense types nearly as often.
- If you want to invest in skills that will enhance your defenses (particularly in combat), then agility, athletics and discipline are the best ways to go about that.
- You can expect conditions like prone, restrained and afflicted to be caused more by adversaries than other types of conditions.
- Generally, physical defenses on adversaries are stronger than cognitive and spiritual defenses.
