Not every encounter has the same challenge level. In general, there ought to be roughly three challenge levels:

• Medium
• High
• Extreme

Why not "low"? See above point about encounters not worth having.

But what do these categories mean? Well, in general, it's about risk: how much is at stake? And this is where traditional combats often fall apart: since the likelihood of PCs actually dying is so incredibly remote, there is little to no default risk.

Killing the PCs: Don't Bother Trying

Even if the PCs legitimately feel that their lives are threatened, these are battle-hardened warriors who, by definition, are willing and accustomed to risking their lives. Thus, if there isn't something else at risk besides the lives of the PCs, there isn't much risk at all.

PC's always win; it's a truism. Why? Simply put, killing the PCs ends the story; it prevents any more realization of all the effort put in by PC and DM alike to flesh out the story. It's simply not worth losing all that to punish a few poorly-chosen actions in combat. But there's something else going on here: namely, an assumption about the meaning of "win".

In most combats, the PCs are trying to stop their enemies from doing what they're doing--usually, it's something bad, that often involves killing them. But what exactly does the enemy considering "winning"? Killing them all? Destroying the world? If so, he's in for some disappointment, as his idea of "winning" is simply not compatible with the game. Thus, the PCs must win.

But what if "win" means something else? What if some invisible, guiding force (read: the DM) could help the bad guys achieve something for a change, even if killing the PCs is impossible? Their lives and freedom may be sacrosanct, but might they have something else to lose?

What Is At Risk

For any given encounter, something must be at risk. The PCs' time and physical well-being do not count...it must be something else. So what do PCs have that can be endangered? Here's a quick list of possibilities:

Mechanical Resources

• These are attributes given by the system, specifically those which do not recover in the ** Hero Points
  • Energy
  • Strain
  • Fatigue
• The best way to threaten these is to simply make encounters difficult. These attributes are depleted faster when the PCs try harder.
• This can also include things like potions, charged items, ammunition, etc. This is highly variable, but generally PCs will use more of these in difficult encounters.

Long-term Health

• Inflicting Hit Point damage is of no concern outside of the flow of combat; Hit Points regenerate quickly. However, injuries and afflictions do not. It is possible to inflict long-lasting harm on PCs simply by hitting very hard.
• Again, this mainly applies to very difficult encounters, particularly combats in which the enemy has high offensive power. The enemy's defense (read: HP) does not factor strongly into this.

Roleplaying Risks

• Assuming your PCs don't suck, their characters actually care about things other than mechanical rules. This gives you all sorts of things to put in danger, such as:
  • Noncombatant NPCs
    • Friends
    • Loved ones
    • Innocents (if the PCs care about that sort of thing)
  • Non-present resources
    • PCs' homes, business interests, etc
    • Treasure not carried
    • Property of innocent NPCs (again, if the PCs care)

Plot Risks

• You can establish that a critical plot point hinges on the PCs' success in a given encounter. This is somewhat effective, in that it at least puts something at risk other than the PCs' lives, but since it must occur to advance the plot, then it has the same problem as trying to kill the PCs: you can't win.
• Instead, try creating multiple plot paths, some more preferable to the PCs than others. Succeeding at an encounter might allow for an easier path to completing the episode, but failure is an option; if they fail, they'll have to do things the hard way.
• You can also create bonus objectives, which are not required to complete the episode as a whole, but grant some sort of meaningful plot reward. In these, too, failure is an option.
  • For example: the players might be tasked with destroying a rampaging demon who threatens the whole country. He happens to be heading for a small town. Destroying the demon is mandatory, while saving the town is optional; therefore, you could create an encounter to handle the "save the town" scenario. Perhaps the reason the demon is heading there is because he likes to drink souls, and only while drinking souls is he vulnerable to kill, so how could he be killed without assaulting the town? But it turns out that by attacking him from within his own home plane, he will be vulnerable, but also more dangerous, as there are other demons there. Failing on that vector still leaves the original vulnerability.
  • A less morally-obvious example: suppose the PCs are breaking someone out of prison. The only mandatory objective is to free their target; however, many other prisoners are there unjustly. Freeing their target only requires taking out one guard and grabbing his key, but freeing more will require many more skill tests, possible combats, and in general increase the risk greatly.
  • The trick here isn't simply to offer bonus content; it will still feel like mandatory objectives. In the prison example, perhaps the prison has safeguards against prison breaks--when one is detected, they will gas all the prisoners in that section. Thus, the PCs are actually risking the lives of the prisoners by attempting to free them, rather than simply doing more work than they otherwise had to.

Weapons out, damage will be inflicted. The victory condition: outlast your enemy.

Combat is well-understood, and has received all too much attention from systems past. Still, it is worth explaining in some detail here.