Friday, July 10, 2015

The Perils of Scaling

I am back with another contribution to the GMs' Roundtable of Doom.

This month's topic comes to us courtesy of +John Marvin:

How do you scale encounters for a smaller or larger group than you had planned on. Or than the published adventure planned on? What works, and what does not? Do different systems affect how you scale? And what about fish? They have scales.

Scaling encounters can be tricky.  In some systems, it is of vital importance.  The presumption of careful balancing of each encounter to be challenging, but not overpowering, requires careful rules for scaling with party size.  In Pathfinder and DnD 4e, for example, the strong presumption of encounter balance requires that DMs pay careful attention to the match between party power and encounter power.  Narrative systems are less careful about balance but careless encounter design can push the game into the lethal territory discussed in a previous GMs' Roundtable series.

So what does one do to scale encounters for different party sizes.  The answer depends, rather critically, on the specific system you use.  However, there are some general consideration to keep in mind across various systems.

1]  Small parties are not just short of hit points and damage output.  

Many systems for scaling are based on linear or nearly linear models of damage output and a party's capacity to absorb damage (often using consumables like potions or limited use healing spells, etc.).  In such a system, scaling involves providing enough monsters to challenge the damage output and hit point absorption of the party.

This approach has a serious limitation.  The capacity of the party to handle opposition is strongly non-linear.  Depending on the system, there is a threshold beneath which the party can not offer the breadth of abilities needed to put them on the track of the typical party power curve that serves as a basis of the scaling systems.  A small party may lack healing or only have a hybrid healer.  This party may lack a dedicated damage dealer (like a glass cannon spell caster or some other character type balanced with heavy damage dealing and personal vulnerability -- with the vulnerability often offset by a party tank).  The tendency for characters in small parties to cover multiple roles often leaves the the party behind the assumed power curve based on a balanced (larger) party.

Scaling for small parties involves targeting the encounters for the abilities that the party does have -- not what a balanced party is presumed to have.  At the logical extreme, scaling for a single player involves providing challenges for what the player can do.  If the player does not have trap-finding abilities, you would not want to send that player into a trap-filled tomb.  With a small party, the same rule applies.  While you can provide some challenges that require novel strategies or highlight the limitations of the party, you should stay away from generic challenges that serve only to frustrate the party or risk a total party kill (TPK).

2]  Large parties are not just full of hit points and damage output

The scaling power curve also confronts problems with larger parties.  Frankly, scaling systems have never worked very well.  There are well-known problems with the DnD 3/3.5 CR system.  I have not played a lot of Pathfinder, but I suspect there are similar limitations.  The DnD 4E scaling system was more robust but not perfect.  In fact, the robustness of the scaling system (and the ability to predict the lethality of an encounter) was a common source of complaints.  The predictability that supports the scaling system also requires very strict balancing across character abilities -- leading many to complain about the similarity of the character options.

Even in these systems that have emphasized balance and scaling systems, the models break down with large parties and high level parties (sometimes even moderate level parties).   There is a discussion thread through recent episodes of The Tome Show podcast's coverage of recent DnD 5e releases have involved the frailty of the encounter design math.  Prepared parties are often defeating threats well past what they are supposed to.  This is often a product of specific interactions of powers the players or sets of players have.  In DnD 4e, this was often a product of stun lock.  In other systems, this can be complex interactions of accuracy improving powers, critical hit range increases, etc.

The complication become a big problem with high level parties.  The complications are present, though, in large parties as well.  As you increase the size of the party, you increase the number of abilities and the potential for these complex interactions.  This is what a lot of players are looking for in games.  It is not necessarily a sign of a broken or weak system.  However, the presence of complex interactions and strategies implies that any system to describe the typical power scale based on party size will be limited.

3]  Scaling in different game types

The issue of scaling depends a lot on the game.  I don't mean this only based on the abilities of characters in the game (such as stuns, critical hit systems, etc.).  More critically, different systems place different importance on combat encounter balance at all.  In a system where combat encounter balance is critical, the issue of scaling is important.  In a system where combat encounter balance is less critical, scaling is less of an issue.

Consider DnD 4E.  This was a system built with character and party balance as a fundamental design element.  There were strong statements about the necessary size of parties and the roles that the party members filled.  If you had a party without a leader or a striker, you had a series balance problem for encounters.  Even with this focus, the scaling system had many problems (solo bosses, stun lock, etc.).

Compare this to a narrative-based system.  In some of these systems, notably those based on the Apocalypse World system, the threats to the party scale almost exactly with the party size.  The fewer actions the party has, the fewer the opportunities for the opposition to act.  In other systems, like the Cypher System games, the system is not based on strict balance but expects players to use the flexibility in the system to handle a variety of encounters with powerful one-shot items (like cyphers).  These systems focus more on stories that transcend specific encounters and thus worry less about the balance of these encounters.  A single powerful cypher can render most encounters moot.  You can constantly try to frustrate the use of powerful, encounter-ending cyphers or you can write stories that don't rely on any single encounter.

This means that your efforts to scale encounters will depend on the nature of the system.  If you are in a combat encounter intensive system, you will need to work careful to modify the hit thresholds (like armor class) and party resources (hit points, healing, etc.) to ensure that parties are not more frail or more powerful than you want for the encounter.  If you are running a more narrative system, you may want to embrace the lack of balance and just let that single crazy item abruptly end an encounter now and then.  This is certainly a common approach in film, television, and novels.

Of course, this flexibility comes at a price.  As discussed previously in this GMs' Roundtable series, some people want a game wherein they can compare their experiences with other players in other groups.  People want to say they beat the Tomb of Horrors or escaped the Ghost Tower of Inverness and have other people know, exactly, what sort of challenge that entailed.  Scaling adventures may change the challenge associated with any adventure.  It renders such comparisons invalid -- which is not a problem for me, but may be for some people.  This is why classic published DnD adventures provided a party size and a level range.  Playing with scaling is, and always has been, an inexact science.

It is interesting to see this debate (or related debates) evolve over the past couple of decades.  DnD 3E was praised, and then criticized (not always by the same people, admittedly), for being very specific and carefully balanced to make encounter difficulty predictable with the CR system.  In previous versions, encounter balance was highly subjective and depended on DMs having some experience with encounter design.  3E provided a more formal system for balancing encounters -- though for a specific range of party sizes.  Pretty quickly, people realized that the system was only an approximation.  4E tried harder to make scaling predictable.  This was, in part, to make encounter design easier and faster for DMs and less dependent on DM experience.  This led to criticism that the system was too predictable, powers were too similar, and the balance came at the price of diversity in creature and player design.  5E moved back to a system where encounter design require more work and thought from the DM and party balance was de-emphasized -- explicitly by the designers.  The result is a system that requires more careful design by DM, putting more emphasis on the expertise of the DM.

DnD (and system design generally) may become something like the choosing a pope.  A common idiom about papal selection is that a fat pope follows a skinny pope.  Maybe a carefully balanced system that makes encounter design easy for GMs will be followed by a system that leans more on GM judgment.

You can submit a question for a future roundtable to

See other responses to the questions at

John Clayton -- Files and Records
Evan Franke -- A Sage among His Book
John Marvin -- Dread Unicorn Games
Marc Plourde -- Inspiration Strike
Peter Smits -- Planeatery Express
Lex Starwalker -- Starwalker Studios
Burn Everything Gaming
Allan Kellogg -- Mythus
Tom Harrison -- Brace of Pistols

No comments:

Post a Comment