Structure of a Combat Round

Sometimes, opponents in close proximity to each other decide they want to start fighting until one of them dies.

The relevance of this picture is entirely up to you.

Combat is the most structured part of Phoenix, as it is in most tabletop games.  This is a necessary feature, because GMs have to pretend they are being fair when they kill off PCs.

The structure of a combat round is a time where design theories like gamism, disassociated mechanics, and rules-oriented design take first priority.  Once this framework is established, drama-oriented design can take over to fill in the meat and potatoes part of combat.

Here's the simple structure of combat in Phoenix.

Design Goals

1. Combat must move quickly.  The most important metric for this is the amount of time a player sits on their ass between their turns.  Players in D&D 3.5e clocked in at up to 45 minutes between turns sometimes.  That's flat-out retarded and I blame the system.
2. Combat must be interesting - remember Rule #1.  It's easy to achieve quick combat if I just have everyone make one roll and then state the outcome of combat - but that's not what I'm looking for.  I want a round-by-round flow of activities.  I want characters to have choices on their turns, and for those choices to have tactical costs and benefits.
3. Order of action must be determined separately every round, independently for each PC and major opponent (at a minimum), with modifiers for faster/slower characters.

With these structural goals in mind, here is the result:  three phases for each combat round.

Phase 1:  Initiative
Initiative is determined each and every round, as per Goal #3.  There are plenty of stumbling blocks to making this happen, chief amongst them being how easy it is to violate Goal #1.  To meet both of these goals, initiative shouldn't take more than a few moments to figure out.  That means little or no math, and no ties to worry about.

I'll post more detail on the Phoenix initiative system in the near future, but dice-based mechanics that met all the parts of Goal #3 while not violating Goal #1 were not meeting my expectations.  I explored some alternatives and devised a card-based method instead.  The only problem with cards is usually the part of Goal #3 that talks about faster/slower characters, but I do have that factored in and it all functions very quickly.

Phase 2:  Actions
This is the part where every participant actually does something.  In Phoenix, characters receive one action per round.  They can do one thing with it.

Common Actions:  Attack, Defend, Delay, Move, Sprint, Support, Suspend, Use.

Look, the problem I have with systems that allow multiple actions is the combinatorics of it all.  To pick on later editions of D&D again, if I have 10 things I can do with minor actions, 10 things I can do with move actions, and 10 things I can do with standard actions - all small numbers - I've suddenly got 10x10x10 or 1,000 possible combinations of things I can do on a turn.  Does anyone seriously wonder why some folk have to sit around and think for a while about what to do on their turn?  Combinations are good for chess-like combats with thoughtful, precise moves; they are not conducive to fast-paced combat.

Yes, I am aware that some players/groups move quite quickly in modern D&D.  But the system doesn't encourage it enough for my tastes.  Only crappy software architects use "they should know better" as an excuse for not getting users down the shortest workflow possible.  So what kind of game designer would use that excuse?

Each character taking one action means the important metric of time-between-players is faster.  If a character wants to move and attack, then they can move this round and attack next round.  This round they just focus on their movement and don't worry about what sort of attack they want to do.  Next round they just have to focus on who to attack, and how.  In between, other characters get to do stuff.

Phase 3:  Maintenance
Changing order of activity each round means that, in order to be fair, some things need to be pulled out of sequence.  Regeneration, for instance.  If you go last in Round 1 and first in Round 2, do you regenerate twice before anyone gets to hit you?  What about ongoing damage effects?  None of that makes sense and it defies dramatic expectation.

In Phoenix, both of these are examples of Maintenance activities.  Players and GMs can maintain all of their characters simultaneously, since order of activity is irrelevant.  There are some other things that can be done in Maintenance, like switching weapons, but Maintenance activities are always self-affecting, and usually just matters of calculation.

After Maintenance, the next round's Phase 1 starts right away.

That's it!